Party-society linkages and democratic representation in Latin America.

Roberts, Kenneth M.

Abstract. Although it is often said that there is a crisis of party systems and political representation in Latin America, it may be more accurate to assert that a transition is underway from some types of partisan representation to others. This transition is best understood by exploring party-society linkages and their transformation over time. Five types of party-society linkage are identified, along with the social and economic changes that have undermined encapsulating and programmatic linkages while buttressing those based on political brokerage, personal appeal, and political marketing. The net effect of these changes is to produce political parties that are more detached from organized social constituencies and more individualized and contingent in their patterns of affiliation.

Resume. Bien que l'on ait souvent l'occasion d'affirmer que le systeme de partis et la representation politique en Amerique latine est en crise, il serait plus exact de dire que certains types de representation sont en voie de transition. L'etude des liaisons partis-societe et de leur transformation au fil du temps permet de mieux comprendre ce phenomene. On peut ainsi degager cinq types de liaisons ainsi que les changements economiques et sociaux qui ont affaibli celles qui ont trait aux programmes et aux formules tout en renforcant celles qui sont axees sur les tractations politiques, l'attrait personnel et le marketing politique. Ces changements ont directement contribue a la creation de partis politiques moins lies a des electorats socialement organises, mais bien plus individualises et contingents quant a leur mode d'affiliation.

Introduction

As elected officials replaced military dictators across most of Latin America in the 1980s, it seemed as if they had barely taken the oath of office before scholars began to lament the "crisis of political representation" that plagued the region's fledgling democratic regimes. (1) The manifestations of this crisis were legion, including voter abstention, declining partisan identification, electoral volatility, the rise of independent candidates and "anti-politics" outsiders, and the demise of traditional political parties as well as secondary associations like labour unions. (2) Since political parties are widely recognized to be the primary intermediary institutions between citizens and the state under democratic regimes, (3) they were often held accountable for this alleged crisis of representation. Although some scholars claimed to find evidence of increasing party system institutionalization in the region, (4) there was a growing sentiment over the course of the 1990s that party systems were failing in a number of countries and that their deficiencies posed formidable problems for the quality and stability of democracy in the region. (5)

Is there a crisis of political representation in Latin America? If so, to what extent is it attributable to the failings of political parties? While intuitively appealing, given the turmoil that exists in many Latin American party systems, the notion of a crisis of representation is overly simplistic. Political representation, after all, can assume a number of different forms, and parties have performed a variety of social and political functions. A crisis in a particular mode of political representation or a certain type of party organization does not necessarily imply a more generalized crisis of representation. What may actually be occurring is a shift away from one mode of representation to another, with distinctive institutional features and societal linkages.

To understand such a process of transformation requires a more systematic analysis of the different ways in which parties represent citizens in democratic contexts--that is, how they articulate societal interests, aggregate groups and individuals, and mobilize support in a competitive electoral arena. In short, it requires that parties be analyzed in the social milieu in which they compete. This departs from much of the contemporary literature on political parties in Latin America, which has placed greater emphasis on the institutional environs of partisan competition. Certainly, the design of electoral rules and the structure of legislative-executive relations can shape the contours of party systems, including the number of parties in any particular national system and the nature of the competition among them. (6) But institutional approaches do not tell us enough about how different kinds of parties relate to social actors, nor do they explain how these relationships are modified as parties' structural foundations are reconfigured by the forces of social and economic change. If parties are in crisis, it is because they are detached from their social moorings, and such detachment can only be understood by means of an assessment of party-society linkages and their evolution over time.

This paper attempts such an assessment of Latin American party systems. It argues that economic, social, and technological changes have altered the political functions performed by party organizations and transformed their societal linkages. The paper develops a typology of linkage patterns and suggests that modes of representation based on durable encapsulating and programmatic linkages between parties and social groups have been displaced by a variety of more contingent and individualized linkages. In a social landscape that is increasingly structured by market relationships, the mass media, and a globalized consumer culture, ideological and group-based patterns of political representation that emerged during the import-substitution industrialization (ISI) era of development have entered into crisis, while marketing-based modes of representation that are tailored to individual preferences have risen to prominence. More than a crisis or failure of representation, what has occurred is a fundamental transformation of the character of political representation. This transformation has produced a new, more fluid political matrix that corresponds to a social landscape dominated by market individualism and a generalized depoliticization of society.

Parties and Democracy in Latin America

Political parties are easy to vilify. By definition, they represent only part of the body politic, which they inevitably lead into conflict with other parts. They are self-interested actors who pursue political power while cloaking their particular interests behind the veil of the public good. When overly strong, they are assailed for monopolizing the state, subordinating civil society, and concentrating authority in self-reproducing bureaucratic enclaves. (7) When overly weak, they are criticized for inadequately representing societal interests, allowing de facto economic or military powers to dominate the political process, failing to establish a programmatic and legislative agenda, undermining the prospects for policy reform, and failing to hold elected officials accountable to their constituents. (8) In public opinion surveys, parties typically rank as the least respected of the major institutions in Latin American societies. It is hardly surprising, then, that independent mavericks can often make a political career out of attacking the party-based political establishment or partyarchy.

Nevertheless, no modern democratic regime has ever functioned without political parties. The closest approximation was probably Peru in the 1990s, where neither governing nor opposition parties played a significant role in the political process. (9) The degeneration of Peruvian democracy under the autocratic rule of Alberto Fujimori, however, hardly augurs well for the prospects of creating a party-less democracy. Parties remain central actors for articulating and aggregating societal interests, structuring electoral and programmatic alternatives, coordinating legislation, recruiting candidates for public office, and holding them accountable once they have been elected, all functions that are vital to the healthy operation of a democratic regime.

Indeed, the Latin American experience over the past two decades suggests that democracy is endangered where there is severe instability in party systems. Table I provides comparative data on electoral discontinuity in Latin American party systems during the most recent wave of democratization between 1978 and 1999, measured as the net shift in vote shares for all of a nation's political parties between the first and last national elections during this time period. The table demonstrates that there are striking differences in the stability of party systems across the region. Whereas countries like Honduras and Costa Rica experienced virtually no change in relative vote shares during this time period, other nations like Peru, Ecuador, and Venezuela witnessed the collapse of the party systems that began the period and the emergence of an entirely new set of (often non-partisan) electoral referents. Notably, these latter three cases are the only countries in the region that experienced a definitive rupture in their democratic regime during the 1990s, (10) and in all three cases the democratic breakdown followed the demise of the party system. This does not necessarily imply that the collapse of party systems is a direct cause of democratic breakdown; it is possible that party system demise is merely a symptom of a broader crisis of democratic institutions that is attributable to other factors, such as economic mismanagement or egregious political corruption. Nevertheless, the association suggests that turmoil in party systems--or a generalized incapacity of parties to articulate societal interests, address outstanding problems, and reproduce their electoral support--is a powerful indicator of a crisis in democratic governance.

In short, despite the proliferation of new social movements, non-governmental organizations, and other non-partisan agents of interest articulation, parties remain vital intermediaries between citizens and the state under democratic regimes. There is no consensus, however, on how parties perform their intermediary functions or what balance they maintain between their societal and statist orientations. In part, this is attributable to variation in patterns of intermediation across space and time. Classic sociological approaches treated European parties as direct expressions of social cleavages and the group-based interests and identities that they spawned. (11) In Latin America and other developing regions, however, party systems were less clearly grounded in such structural divisions. (12) More recently, as social cleavages and the bonds linking parties to social groups have loosened in Europe and elsewhere, scholars have portrayed parties as collusive governing cartels that are entrenched in state institutions and dependent on state resources. (13) In this latter approach, parties' linkages to society are understood to be shallow and tenuous.

Clearly, party-society linkages can vary across nations, (14) across different parties within the same national political system, or over time for a given party as social, economic, and technological changes alter the landscape on which it competes. As linkages change, so does the character of political representation. A conceptual mapping of different linkage patterns is thus necessary to understand how changes in party-society linkages have transformed political representation in contemporary Latin America.

Reconceptualizing Party-Society Linkages

Party systems are generally categorized according to their number of competing units (or parties) and/or their ideological makeup. (15) More recently, Mainwaring and Scully have developed a typology of Latin American party systems based on their level of institutionalization. (16) But to understand the alleged crisis of political representation in Latin America, categorization based on the nature of party-society linkages has a number of advantages, as it makes it possible to identify the different ways in which parties mobilize support and how these evolve in response to societal change. Previous efforts to conceptualize different patterns of party-society linkage have contributed greatly to scholarly understanding of party systems, (17) yet they do not exhaust the range of possible linkages or fully specify the dimensions along which linkage patterns vary. The discussion that follows builds on previous work in order to extend its range of application, particularly within the Latin American region.

Party-society linkages vary along several important dimensions, including their level of association and their degree of contingency. Association can occur at the individual level, when a citizen decides to join or vote for a party organization, or at a collective level, when citizens acquire a formal or de facto association with a party by virtue of their membership in a party-affiliated social group or organization, such as a labour union, peasant federation, or ethnic group. Likewise, linkages vary widely in their permanence or contingency. For some individuals or groups, association with a party is soldered by deep-seated ideological commitments, political identities, and organizational bonds. Linkages based on such durable loyalties carry over from one electoral cycle to another--indeed, they may be passed on from one family generation to another--and they are rarely severed in the absence of serious political trauma. For other individuals or groups, however, linkages may be temporary and conditional, as they are forged during specific (generally electoral) political conjunctures and may not withstand a poor governing performance by the party, a shift in its policy stance, the rise of new competitors, or a change in the issue agenda. Where such contingent linkages prevail, voters do not approach a new electoral cycle as loyalists with pre-determined partisan preferences; instead, they are free agents who shop around for a party or candidate that strikes a responsive chord in the prevailing political conjuncture.

These dimensions are readily apparent in the typology of linkage patterns developed below. The typology identifies five ideal-type modes of linkage that are logically distinct and independently sufficient to mobilize support in the electoral arena. The five modes of linkage are jointly exhaustive, but they are not mutually exclusive; that is, every party is assumed to employ at least one of these linkage patterns to mobilize political support, but reliance on one does not preclude the utilization of others. The existence of multiple or cumulative linkages is an indicator of deeper, and presumably stronger, bonds between parties and citizens.

Political Brokerage and Patron-Clientelism

Arguably the oldest and most pervasive mode of party-society linkage in Latin America is patron-clientelism, or what I will call political brokerage, following the lead of Arturo Valenzuela. (18) Brokerage entails an exchange of selective material benefits for political support; as Lawson argues, it creates linkages based on the provision of rewards and favours for political loyalty. (19) Parties based on such linkages--identified by Kaufman as "machine parties" (20)--are generally loosely organized alliances of political notables and their patronage networks or machines. Local power brokers distribute particularistic benefits--such as public employment, a government contract, or paved streets--in order to co-opt electoral support, but their patronage machines generally do not require extensive grassroots organization or participation by party loyalists. Brokerage-based parties are thus vertically rather than horizontally organized: they cut across class distinctions, and they are usually comprised of hierarchical chains of patrons, brokers, and clients rather than strong mass organizations. The constituencies of such parties are socially heterogeneous and relatively individualized, as affiliation is based on personal ties or family and kinship connections rather than membership or identification with larger social groups. (21) Little is asked or expected of party supporters beyond the voting booth, and parties' limited presence in society outside electoral campaigns is sustained by the activities of patrons and brokers. Political loyalty is reproduced by periodic material exchanges and personal bonds rather than ideological commitments. Although brokerage practices can exist in virtually any political party, they generally provide the dominant linkage mechanism in centrist and conservative parties with roots in the oligarchic era prior to the onset of mass politics, such as the Colombian Liberals and Conservatives, the Paraguayan Colorados, and the Uruguayan Blancos and Colorados.

Encapsulating Linkages

Encapsulating linkages have two distinctive features: mass-based organizational structures and participatory modes of affiliation. In short, parties with encapsulating linkages, or what Lawson calls "participatory" linkages, (22) incorporate the masses directly into the political process beyond the act of voting. This entails the construction of a party organization with local branches or grassroots units that provide members, or militantes, with permanent opportunities for political activism. These party organs are often supplemented by close bonds to mass secondary associations of workers or peasants, creating collective modes of association among groups defined by pertinent social cleavages or identities. For this reason, Kaufman speaks of "group-based" party systems where organized social blocs comprise the primary constituencies of parties. (23) The political activism of militantes carries the expectation that they help the party penetrate and organize civil society. Even where such parties possess centralized and hierarchical leadership structures, there is an important horizontal, bottom-up dimension to their organizational models, and they rely greatly on the political labour of committed non-professional loyalists.

As the term signifies, encapsulating linkages create powerful bonds between militantes and their parties. Militantes are enveloped within a web of social and organizational networks that bring them into regular contact with the party and other loyalists. Parties thus help to integrate society and socialize citizens to political life. Indeed, they may provide an array of social services that create channels for participation and reproduce political loyalties, including health or dental clinics, childcare facilities, educational programs, youth and women's groups, cultural activities, and sports clubs. Levitsky's depiction of Peronist local branches, or unidades basicas (UBs), provides a paradigmatic example of the encapsulating activities and services performed by mass party organizations:

 UBs play a central organizational role during electoral processes,
 signing up members and providing activists to paint walls,
 put up posters and mobilize voters for rallies. Between elections,
 many UBs continue to play an important role in neighbourhood
 life, serving as critical points of access to city and
 provincial governments. In addition to distributing food, medicines,
 and clothing, attending to local residents' problems with
 municipal government and providing social services such as
 legal and medical assistance, school tutoring, and even free
 haircuts, many UBs administer government social programs
 and attend to neighbourhood infrastructural problems such as
 sewage, street lights and road surfaces. Many UBs also serve
 as cultural centers, organizing sports activities for young people,
 vacation trips for the elderly and parties for neighborhood
 birthdays. (24)

The encapsulating linkages forged by such activities tend to be highly durable and resistant to many kinds of political shocks. While these linkages are occasionally found within centrist parties like the Chilean Christian Democrats and the Argentine Radicals, they are most common among mass-based populist or leftist parties with close ties to labour movements and historical roots in processes of popular mobilization, including the Peronist party in Argentina, the American Popular Revolutionary Alliance (APRA) in Peru, the Democratic Action party (AD) in Venezuela, the Institutional Revolutionary Party (PRI) in Mexico, the Workers' Party (PT) in Brazil, the Sandinistas in Nicaragua, and the Socialist and Communist parties in Chile. As such, encapsulating linkages are exclusive to the post-oligarchic era of mass politics in Latin America.

Programmatic Linkages

Programmatic linkages exist where citizens develop loyalties to a party that are based on ideological or general programmatic commitments. Such linkages require that parties adopt ideological positions that are reasonably consistent, coherent, and differentiated from those of their competitors. Likewise, they require that citizens develop relatively well-defined belief systems and programmatic preferences. In the European experience, ideological differences were historically grounded in social cleavages, and were often superimposed upon encapsulating linkages; in particular, socialist parties articulated ideological positions that appealed directly to the collective interests and identities of their core working class constituencies, and they built mass organizations to encapsulate their supporters. This pattern was much weaker in Latin America. Outside Chile, the working class was typically mobilized politically by ideologically amorphous and socially heterogeneous populist parties rather than class-based parties of the left. (25) Consequently, leftist partisanship in most of Latin America has relied on individual rather than collective ideological commitments and is often divorced from mass-based encapsulating linkages. Centrist, populist, and conservative parties, on the other hand, have generally eschewed explicit ideological definition, leaving programmatic linkages to play a secondary role in their appeals for popular support. (26)

Personalistic Linkages and Charismatic Bonds

There is a long tradition in Latin America politics of parties serving as little more than electoral vehicles for prominent personalities. Supporters are attracted to the party not by ideological commitments or organizational identities, but by the leadership qualities of the dominant personality. Although personalistic linkages do not necessarily imply charismatic leadership, they are strongest where "charismatic bonds" exist between a leader who demonstrates special gifts and popular masses who deposit their confidence in a messiah figure to direct a process of radical change or resolve a national crisis. (27) Since personalism is generally in conflict with institutionalization, parties based on personalistic linkages are often loosely organized or even ephemeral in their existence, as can be seen in the "parties" that accompanied leaders like Velasco Ibarra in Ecuador, Collor in Brazil, and Fujimori in Peru. Even a movement as formidable and resilient as Peronism has been renowned for its lack of organizational institutionalization. (28) In some cases, however, personalistic parties may routinize their authority, as seen in the Mexican PRI after Cardenas, or even develop disciplined organizational structures, as with APRA under Haya de la Torre. Although affiliation with personalistic parties is often highly individualized, collective association may exist when a charismatic figure mobilizes support among organized social blocs, as seen in the Peronist absorption of the Argentine labour movement.

Marketing Linkages

Whereas brokerage, encapsulation, ideology, and charisma are all capable of forging durable bonds between citizens and parties, marketing linkages are by definition contingent and temporary. Such linkages are generally formed in specific electoral conjunctures as parties appeal to uncommitted voters on the basis of a particular policy stance, recent performance in office, the relative capabilities of particular candidates, or the negative attributes of competitors. These linkages generate conditional support rather than political loyalty, as citizens do not forge lasting organizational bonds or identities. Instead, they approach each electoral cycle as a one-shot process, and they reassess competing parties' issue positions, track records, and candidate offerings before deciding whether to stay with the same party or change their vote. This forces parties to continuously polish their public image by reframing the political agenda, modifying their issue positions to accommodate public opinion, defending their track record, reinventing their candidates, and highlighting the failings of their opponents.

Parties that rely on marketing linkages tend to be streamlined and professionalized in their organizational structures, as captured in Panebianco's concept of the professional-electoral party. (29) They do not possess strong grassroots branches or affiliated mass organizations; support is thus individualized rather than collective. Indeed, such parties have little need for grassroots militantes to penetrate civil society and mobilize loyal constituents. In an era of modern communications and survey research techniques, they can appeal to voters via the mass media and tap popular sentiments through public opinion polls and focus groups. Rather than bear the costs of a permanent mass organization composed of grassroots activists, parties are teams of professional politicians advised by hired technical specialists in the fields of campaign advertising, media communications, and survey research. These teams are activated during electoral cycles, but they have little presence in society in the interim period between elections.

Clearly, these five modes of linkages are not mutually exclusive. Brokerage is present in virtually any party that has had access to public office, and in an era of mass communications any party that is a serious contender for state power must develop marketing linkages. In Europe, there is a long tradition of combining encapsulating and programmatic linkages, especially on the political left. In Latin America, conservative and centrist parties have relied heavily on brokerage and personality, while populist parties like the Peronists in Argentina and Democratic Action in Venezuela have sometimes combined personalistic, encapsulating, and brokerage linkages. Leftist parties like the Workers' Party in Brazil and the Communist Party in Chile, on the other hand, have combined programmatic and encapsulating linkages. To further complicate the picture, linkage patterns are fluid and adaptable, and thus subject to reconfiguration as political entrepreneurs respond to changes in social, economic, and technological contexts. The Chilean Socialist Party, for example, which relied historically on programmatic and encapsulating linkages, has increasingly been transformed into a professional-electoral party that depends on marketing and personalistic linkages. (30)

Indeed, recent patterns of social and economic change have undermined several traditional modes of party-society linkage while reinforcing other, less institutionalized forms. The dislocation that results accounts for much of the perception of a crisis of political representation in the region. To place this crisis in the proper perspective requires a more systematic assessment of the transformation of party-society linkages and the factors that lie behind it.

The Transformation of Party-Society Linkages in Contemporary Latin America

In his classic comparative study of party organizations, Duverger suggested that mass-based, encapsulating organizational forms provided socialist working-class parties with a comparative electoral advantage in modern mass democracies that was sure to stimulate imitation by centrist and conservative parties. (31) Eckstein challenged this notion of a "contagion from the Left," claiming that mass party organizations were too expensive to maintain and functionally unnecessary for generating electoral support in an era of mass media communications and professionalized political campaigns. (32) Eckstein argued that loosely organized, cadre-based parties oriented toward the middle class and business interests such as those in the US had pioneered in the development of modern media campaign techniques, providing them with a comparative advantage over European-style mass parties in contemporary democracies.

Eckstein's notion of an organizational "contagion from the right" clearly comes closer to capturing the evolutionary dynamics of party-society linkages in the recent Latin American experience. Three core trends within this contagion are readily discernible: (1) a shift from collective to individualized modes of association; (2) a shift from mass to cadre or elite-based organizational forms; and (3) a shift from fixed and durable bonds to more fluid and contingent forms of support. Taken together, these three trends entail a shift toward looser party-society bonds, less institutionalized forms of political representation, and a professionalization of the political arena.

At the heart of these trends has been the severe erosion of both encapsulating and programmatic linkages, the only two forms of linkage that historically encouraged participatory forms of party militancy. These linkages were based on fixed organizational or ideological bonds, creating durable political loyalties and, at least in the case of encapsulation, collective modes of association and mass-based organizational structures. Nevertheless, these linkages have been pummeled by socioeconomic, political, and technological changes over the past several decades. Encapsulating linkages were forged during the era of import substitution industrialization (ISI) in the middle of the twentieth century, when rapid urbanization and industrialization created a new working class that was unattached politically and amenable to organization in both economic and political spheres. The expansion of the state's regulatory and welfare responsibilities generated powerful incentives for political mobilization, and as capitalism spread through the countryside, eroding traditional patron-client bonds between landlords and peasants, even the peasantry became a potential target for mobilizing appeals around land reform issues. Parties typically mediated the corporatist relationships that states established with labour and/or peasant organizations in order to control the articulation of lower-class demands and structure their political participation. This allowed parties to establish organizational bonds and distribute material benefits to encapsulated social constituencies, which were quickly transformed into durable bastions of electoral support.

Likewise, ideological linkages were significant during the ISI era, when distinct market, state capitalist, and socialist development models confronted each other in the political arena. Although many populist, centrist, and conservative parties lacked explicit ideological definition, there was sufficient variance in programmatic positions across the party spectrum to give voters meaningful choice in many (though not all) national electoral arenas. (33) Consequently, citizens with ideological preferences had incentives to align with the party that best articulated their beliefs and values.

Both encapsulating and programmatic linkages have been hard-pressed to withstand the maelstrom of political and economic changes that have swept across Latin America over the past two decades. Programmatic linkages have been greatly weakened by the collapse of both socialist and state capitalist development models in Latin America and the dramatic narrowing of policy debate, as seen in the diffusion of the so-called "Washington consensus" for market liberalism.(34) The debt crisis of the 1980s eroded the fiscal foundations of ISI and the state capitalist development models followed by populist parties, while the burgeoning crisis and eventual collapse of communism, combined with the growing defensiveness of European welfare states in an era of market globalization, undermined the appeal of both Marxist and social democratic alternatives. As global financial pressures narrowed the maneuvering space and sovereign economic powers of national governments, parties that held fast to populist or statist programs and had the misfortune to shoulder governmental responsibilities during the 1980s, such as APRA in Peru and Bolivia's leftist coalition under Hernan Siles, were politically devastated by their mismanagement of fiscal crises and their association with hyperinflation. Leaders from other parties with statist traditions--including the Mexican PRI, the Argentine Peronists, the Venezuelan AD, the Bolivian MNR, the Costa Rican PLN, the Brazilian Social Democrats, and the Chilean Socialists--embraced the core of the Washington consensus, and although most of these parties (excepting the AD) have remained electorally competitive or even thrived, their contemporary appeal has little to do with the articulation of programmatic or ideological alternatives.

Paradoxically, although the political right has been ideologically rejuvenated by the spread of neo-liberalism, conservative parties have played a minor role in the implementation of market reforms and the diffusion of the new programmatic consensus. Deep market reforms have been imposed by a variety of military dictators, civilian autocrats, nonpartisan technocrats, and erstwhile populist parties, but they have rarely been initiated by pro-business conservative parties. As such, the reform process is often entrusted to parties and leaders who act less out of conviction than out of a pragmatic concession to political and economic pressures. Many neo-liberal technocrats and business elite prefer non-partisan channels of political influence over the costly and time-consuming tasks of party building, and conservative parties continue to rely heavily on brokerage and marketing linkages as opposed to explicit ideological appeals.

Where populist or centre-left parties were saddled with the responsibility of implementing neo-liberal reforms, some fraying of encapsulating linkages to labour organizations invariably occurred. The deregulation or "flexibilization" of labour markets weakened mechanisms of corporatist control and eroded workers' rights and benefits that had been hard-fought gains of previous rounds of populist mobilization. Likewise, reforms that privatized state-owned industries, streamlined public administration, and opened national economies to foreign competition created unemployment and dampened wages, straining the relationships between labour movements and their partisan allies. Organized labour, however, was dramatically weakened by the collapse of ISI and the shift to neo-liberalism, especially in those countries where strong labour movements had emerged during the middle of the twentieth century. The informal sector of the workforce expanded rapidly, as did the number of workers employed in temporary contract or non-contract positions. These workers were notoriously difficult to organize, as their economic activities were too widely dispersed, irregular, and functionally heterogeneous to facilitate collective action or the construction of strong class identities.

Trade union density thus plummeted across the region as labour organizations--increasingly the domain of relatively privileged, permanent workers in the public sector and the very largest private firms--strained to articulate the interests of a more fluid and fragmented labour force. A parallel process of social atomization was underway in the countryside, where a limited parcelization of property, the flight of rural workers to urban areas, temporary labour practices, and restrictions on collective bargaining diffused peasant land claims and undermined collective action. By the 1990s large-scale peasant mobilizations for land had largely ceased outside of Brazil and Ecuador, and market-oriented claims related to wages, prices, credits, and agricultural inputs did not spawn rural civil societies with organized mass constituencies that could be bound to political parties. In short, the social and economic landscape of the neo-liberal era discouraged the formation of organized blocs of lower class citizens in both urban and rural areas who could be encapsulated by party organizations. And although a plethora of new social movements and popular economic organizations emerged as class-based forms of collective action declined,(35) they were typically wary of restrictive (or manipulative) ties to political parties and highly defensive of their political autonomy. Often, they preferred to acquire technical and material support from non-governmental organizations (NGOs) or municipal authorities rather than parties or states. Civil societies in contemporary Latin America are thus more diverse but also more fragmented and depoliticized than those of the populist era, with less density in large-scale secondary associations and a growing detachment from parties and states. As more social actors develop ties to NGOs or carve out their own market niches, the old model of class actors who press claims on the state that are mediated by national party organizations has been increasingly displaced by more fluid, decentralized, and issue-specific "associative networks" of private groups and local public officials that operate at the margins of party systems. (36)

But if civil societies are less structurally disposed to provide readymade collective constituencies for party encapsulation, it must be recognized that contemporary parties have also devoted far less energy to the tasks of penetrating and organizing civil society. Indeed, party leaderships have de-emphasized encapsulating linkages in both the civic and internal partisan arenas, clearly reflecting their calculations that mass organization is no longer a prerequisite for electoral success. Candidates for public office can enter voters' living rooms via television or campaign mailings and appeal for support directly without a mass organization of committed activists to inform, proselytize, and mobilize the electorate. Candidates like Fujimori and Collor have even won national elections without any significant organization at all. And if mass organization is no longer necessary to mobilize electoral appeal, neither is it required to articulate political and economic demands or provide political entrepreneurs with input regarding popular sentiments. Public opinion surveys and focus group studies allow candidates to measure the salience of different issues, identify the policy preferences of targeted social groups, and gauge the effectiveness of different strategies for framing issues and addressing problems. A relatively small team of professional fundraisers, pollsters, media consultants, direct mail specialists, and campaign managers can thus displace mass organizations from their traditional functional responsibilities in the electoral arena, much like new social movements and NGOs have taken over many of parties' interest-articulation and service-provision functions in civil society. Furthermore, hired professional campaign teams do not require investment in a permanent party bureaucracy or branch offices, and in contrast to mass party organizations they pose few restrictions on the strategic autonomy of party leaders to adapt their programmatic positions or coalitional alignments in response to changing political and economic circumstances. (37)

Consequently, there has been a notable shift from programmatic and encapsulating linkages to media-based marketing linkages, which place a premium on the technical skills developed by professionalized campaign teams. This can be seen in the rise of new centre-left professional -electoral parties, such as the Chilean Party for Democracy (PPD) and FREPASO in Argentina, which were spawned by historic socialist or populist parties but have weak grassroots structures, minimal ties to organized labour, and powerful appeal among the unorganized but increasingly cosmopolitan middle classes. These parties typically generate electoral support by projecting attractive, if ideologically amorphous, images of modernity, moderation, and civic republicanism through the mass media.

Personalistic linkages and charismatic bonds have also thrived in contemporary Latin America. Given the demise of organizational and ideological identities, personal appeal has increasingly provided the adhesive force to aggregate disparate interests in the electoral arena, and television provides a powerful medium for the projection of individualized candidate images. Considerable attention has thus been focused on the rise of new forms of populist leadership that appeal directly to atomized mass electorates while bypassing or subordinating representative institutions like parties and labour unions. In contrast to classical populist figures like Peron, Vargas, Cardenas, and Haya de la Torre, who built the mass social and political organizations that dominated the ISI era, the new breed of populists eschew mass organization and extra-electoral forms of political mobilization, and in some cases they have even imposed neo-liberal structural adjustments.(38) Nevertheless, personalistic leaders like Fujimori, Menem, Collor, Bucaram, and Chavez have mobilized powerful support among the lower classes, even if their autocratic tendencies and lack of institutionalized bases have undermined the stability of their rule.

Recent trends have been more nuanced with respect to brokerage and clientelistic linkages. Patron-clientelism has deep roots in Latin America's rural social structure and the oligarchic party systems that it spawned, and it thrives in social contexts that provide direct personal ties between elite and subaltern sectors. Although brokerage has certainly influenced urban political dynamics in Latin America,(39) as it did historically in the US, it is probably less efficient in larger and more impersonal electoral environments; as Geddes points out, it should be more cost-effective for politicians to supply public goods than private (patronage-based) goods for a dense, impersonal mass electorate. (40) Furthermore, in theory the neo-liberal model should drastically reduce the opportunities and resources for private rent-seeking and political patronage. (41) As states slash public employment, privatize social services, deregulate the economy, and leave distributive outcomes to the mercy of the marketplace, parties should have fewer means to distribute selective material rewards in exchange for political loyalty.

In practice, however, brokerage appears to be alive and well in contemporary Latin American party systems. The privatization of public enterprises can provide executives with windfall resources and opportunities to bestow favors, while the new social policy emphasis on targeted poverty relief spending can easily be manipulated to the advantage of incumbents. (42) In contexts of heightened individual economic insecurity and diminished state supply of public goods, the political dividends that derive from the partisan provision of selective benefits may actually be accentuated. As Valenzuela argues, brokerage is most prevalent where scarcity precludes the universal provision of public goods and encourages a resort to particularistic allocative criteria. (43) Consequently, scholars have noted that parties like the Mexican PRI and the Argentine Peronists that historically developed encapsulating corporatist linkages to mass organizations have become increasingly dependent on the selective distribution of patronage after implementing neo-liberal reforms. (44) Even in Chile, where neo-liberal reforms have advanced the furthest and parties are among the most professionalized in Latin America, a Christian Democratic leader recently lamented that the party "doesn't have militantes, only clients," leaving it without a "common project." (45) The streamlined state of neo-liberalism may thus impose limits on patronage distributions, but inventive politicians appear capable of adapting political brokerage to the exigencies of a new socioeconomic era.

Implications for Democratic Representation

I have argued that encapsulating and programmatic linkages between parties and society have been gravely weakened, while marketing and personalistic linkages have risen in prominence and brokerage appears to be holding firm. These linkage trends have several important implications for democratic representation in Latin America. First, political representation is increasingly individualized rather than collective; that is, party affiliation is being driven more by individual preferences and choices rather than membership in a particular social bloc or organizational collectivity. Marketing, personalistic, and brokerage linkages are all based on partisan appeals to individuals rather than social collectivities, and they do not require the construction of mass organizations. This trend makes it especially difficult for lower class citizens to articulate their interests effectively in the political arena, because they have few political resources other than their strength in numbers, and their voice has historically been highly dependent on their capacity for collective action and expression.

A second, related implication of linkage trends is that they are generally associated with cadre or elite-based party organizations rather than mass bureaucratic organizations with active grassroots structures. Successful parties still appeal to a mass electoral constituency, but this does not make them mass parties, as they may possess no permanent organizational structure below the leadership cadres. Parties thus have sympathizers and voters, but relatively few members, activists, or militantes. Marketing, personalistic, and brokerage linkages demand little or nothing of party supporters beyond the ballot box, and they provide few structures or opportunities to encourage extra-electoral forms of participation. As such, party organizations revolve, respectively, around the professionals, the caudillos, and the brokers or patrones, with minimal input from the bottom-up. This trend reflects a generalized depoliticization of social life and a narrowing or specialization of the political sphere. As Lechner argues, (46) individuals in socially differentiated market societies have multiple channels of cultural expression that compete with the political sphere and undermine the bases of large-scale collective solidarities that fuel political activism.

Finally, these trends point to a shift from fixed, durable bonds to more fluid and contingent patterns of support. The individual preferences and choices that undergird brokerage and marketing linkages are generally more fluid and instrumental than ideological convictions or membership in a social bloc. Likewise, the personalistic linkages forged by contemporary populist figures appear far more fragile than those of classical populists, since they have not been accompanied by serious efforts to institutionalize support. These linkage patterns thus allow relatively high levels of mobility for individual voters, who may switch allegiances from one election to the next as parties compete to project a better image, frame the issues more effectively, package more attractive candidates, or offer more enticing selective benefits. Although brokerage in Colombia, Uruguay, Costa Rica, and Honduras has generated strong partisan identities and remarkably stable patterns of electoral competition, (47) in other countries it has provided more instrumental, contingent, and volatile forms of support. Furthermore, brokerage is a two-edged sword: it can create material dependencies that generate durable partisan loyalties, but it can also degenerate into blatant forms of corruption and political favoritism, alienating voters and leaving party systems vulnerable to the rise of anti-establishment political outsiders.

Marketing and personalistic linkages also present a number of significant vulnerabilities. Marketing linkages are very flexible, but they leave parties relatively detached from society, with weak organizational bonds, shallow social roots, and ephemeral political loyalties. Parties' professionalized leadership teams can easily become an entrenched, collusive political caste that uses state resources to reproduce its authority and exclude outside challengers from the political arena. Like patronage-based machine parties, such "cartel parties"(48) are vulnerable to voter backlash and challenges from anti-establishment figures, as the Venezuelan case so graphically demonstrates.

Personalistic linkages, on the other hand, create vulnerabilities that are related to their lack of institutionalization. By their very nature, such linkages do little to hold leaders accountable to their constituents, and autocratic, unpredictable political behavior is the frequent result. In Peru, for example, Fujimori's independence was both his greatest political resource and, in the end, the Achilles heel that undermined his regime. Lacking an institutionalized base of support, Fujimori turned to military and intelligence officials with minimal democratic credentials to prop up his authority, then proclaimed a presidential coup that allowed him to dissolve a congress controlled by his opponents. His principal advisors later provoked a political scandal by forging a million signatures to facilitate the registration of Fujimori's fourth "party" vehicle, and then bribed opposition parliamentarians to switch sides and thus create an unelected official majority in the congress. These blatant manipulations of the democratic process eventually detonated a political crisis that drove Fujimori from office.

Ultimately, the three primary linkage mechanisms reduce parties to their most basic, self-referential political function: electing candidates from their ranks into public office. As such, parties have been stripped of a range of other sociopolitical functions that historically served the public interest and enhanced their legitimacy in the eyes of the citizenry. Parties in contemporary Latin America are less active in representing social groups in the political arena, organizing civil society, articulating ideological and programmatic alternatives, socializing citizens in democratic values, supplying public goods, and creating grassroots channels for popular participation and social integration. Stripped of such edifying functions, teams of self-interested office seekers are hard-pressed to maintain their legitimacy, and they become increasingly dependent on instrumental assessments of their performance to maintain popular support. When political corruption or economic hardship render such assessments negative, parties become easy prey for anti-establishment political entrepreneurs who blame them for an assortment of collective ills and appeal to an alienated electorate. At such moments a crisis of political representation becomes palpable, but it is generally preceded by a more complex, long-term transformation in the character of party-society linkages. Any effort to improve the quality and stability of democratic governance in Latin America must start with a recognition of these changes and how they have altered the character of political representation.

Notes

(1.) Torcuato S. Di Tella, ed., Crisis de representatividad y sistemas de partidos politicos (Buenos Aires: Grupo Editor Latinoamericano, 1998); Frances Hagopian, "After Regime Change: Authoritarian Legacies, Political Representation, and the Democratic Future of South America," World Politics 45 (April 1993): 464-500; Pedro Nikken, America Latina : La democracia de partidos en crisis (San Jose, Costa Rica: Instituto Interamericano de Derechos Humanos, 1992); Steve Stein and Carlos Monge, La crisis del estado patrimonial en el Peru (Lima: Instituto de Estudios Peruanos and the University of Miami, 1988).

(2.) Carina Perelli, Sonia Picado S., and Daniel Zovatto, eds., Partidos y clase politica en America Latina en los 90 (San Jose, Costa Rica: Instituto Interamericano de Derechos Humamos, 1995).

(3.) Giovanni Sartori, Parties and Party Systems: A Framework for Analysis (Cambridge: Cambridge University Press, 1976).

(4.) Robert H. Dix, "Democratization and the Institutionalization of Latin American Parties," Comparative Political Studies 24 (January 1992): 488-511.

(5.) Frances Hagopian, "Democracy and Political Representation in Latin America in the 1990s: Pause, Reorganization or Decline?" in Felipe Aguero and Jeffrey Stark, eds., Fault Lines of Democracy in Post-Transition Latin America (Miami: University of Miami North-South Center Press, 1998); Scott P. Mainwaring, Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil (Stanford, CA: Stanford University Press, 1999); Kenneth M. Roberts and Erik Wibbels, "Party Systems and Electoral Volatility in Latin America: A Test of Economic, Institutional, and Structural Explanations," American Political Science Review 93, 3 (September 1999): 575-590.

(6.) Maurice Duverger, Political Parties: Their Organization and Activity in the Modern State (London: Methuen and Co. Ltd., 1964); Mark P. Jones, Electoral Laws and the Survival of Presidential Democracies (Notre Dame, IN: University of Notre Dame Press, 1995); Arend Lijphart, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies 1945-1990 (Oxford: Oxford University Press, 1994); Matthew Shugart and John Carey, Presidents and Assemblies: Constitutional Design and Electoral Dynamics (Cambridge: Cambridge University Press, 1992).

(7.) Michael Coppedge, Strong Parties and Lame Ducks: Presidential Partyarchy and Factionalism in Venezuela (Stanford, CA: Stanford University Press, 1994).

(8.) Mainwaring, Rethinking Party Systems; Scott Mainwaring and Timothy R. Scully, eds., Building Democratic Institutions: Party Systems in Latin America (Stanford, CA: Stanford University Press, 1995).

(9.) In general, the opposition to President Fujimori revolved around prominent individuals rather than party institutions, including in the electoral arena. The succession of official "parties" sponsored by Fujimori were essentially shadow organizations that provided a label for his electoral candidacies and those of his loyalists, although they did play a role in coordinating support for the president's legislative agenda.

(10.) Alberto Fujimori launched a presidential coup or auto-golpe to popular acclaim in Peru in 1992. In Ecuador, military officers sympathetic to a mass protest movement toppled the elected president in January 1999, although they quickly turned power over to the constitutional vice-president under considerable international pressure. In Venezuela, newly elected President Hugo Chavez adopted extra-constitutional plebiscitary measures in 1999 to elect a constituent assembly, displace the elected congress, and draft a new constitution that proclaimed an end to the old republic.

(11.) Seymour Martin Lipset and Stein Rokkan, eds., Party Systems and Voter Alignments: Cross-National Perspectives (New York: Free Press, 1967); Stefano Bartolini and Peter Mair, Identity, Competition, and Electoral Availability: The Stabilization of European Electorates 1885-1985 (Cambridge: Cambridge University Press, 1990); Mark N. Franklin, Thomas T. Mackie, and Henry Valen, Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries (Cambridge: Cambridge University Press, 1992).

(12.) Robert H. Dix, "Cleavage Structures and Party Systems in Latin America," Comparative Politics 22 (October 1989): 23-37.

(13.) Richard S. Katz, "Party Organizations and Finance," in Lawrence LeDuc, Richard G. Niemi, and Pippa Norris, eds., Comparing Democracies: Elections and Voting in Global Perspective (Thousand Oaks, CA: Sage Publications, 1996); Peter Mair, Party System Change: Approaches and Interpretations (Oxford: Clarendon, 1997).

(14.) Kay Lawson, ed., Political Parties and Linkage: A Comparative Perspective (New Haven, CT: Yale University Press, 1980).

(15.) Giovanni Sartori, Parties and Party Systems (Cambridge: Cambridge University Press, 1976).

(16.) Mainwaring and Scully, Building Democratic Institutions.

(17.) See especially Lawson, Political Parties and Linkage, and Herbert Kitschelt, Zdenka Mansfeldova, Radoslaw Markowski, and Gabor Toka, Post-Communist Party Systems: Competition, Representation, and Inter-Party Cooperation (Cambridge: Cambridge University Press, 1999).

(18.) The brokerage concept is more generalizable than that of patron-clientelism, which is most properly used in reference to highly asymmetrical and personalized exchange relationships between lower class individuals and traditional social elite. Brokerage can refer as well to exchanges of material benefits for political support that are more voluntary and less asymmetrical in nature, as in many urban party machines. For an elaboration, see Arturo Valenzuela, Political Brokers in Chile: Local Government in a Centralized Polity (Durham, NC: Duke University Press, 1977), especially Chapter 7.

(19.) Kay Lawson, "Political Parties and Linkage," in Lawson, ed., Political Parties and Linkage.

(20.) Robert R. Kaufman, "Corporatism, Clientelism, and Partisan Conflict: A Study of Seven Latin American Countries," in James M. Malloy, ed., Authoritarianism and Corporatism in Latin America (Pittsburgh: University of Pittsburgh Press, 1977).

(21.) In the US experience, urban party machines often developed brokerage ties with ethnic immigrant groups at the turn of the century, giving them some foundation in social cleavages. In Latin America, however, the constituencies of brokerage or machine parties have been poorly differentiated by class or ethnic cleavages, leaving individual or, at times, community bonds to determine the direction of partisan loyalties.

(22.) Lawson, "Political Parties and Linkage."

(23.) Kaufman, "Corporatism, Clientelism, and Partisan Conflict."

(24.) Steven Levitsky, "Crisis, Party Adaptation, and Regime Stability in Argentina: The Case of Peronism, 1989-1995," Party Politics 4, 4 (1998): 458.

(25.) Ruth Berins Collier and David Collier, Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America (Princeton: Princeton University Press, 1991); Dix, "Cleavage Structures and Party Systems in Latin America."

(26.) Some exceptions to this generalization would include the centrist Christian Democratic Party in Chile during its formative years and the right-wing Independent Democratic Union (UDI) in Chile, both of which articulated explicit ideological positions.

(27.) Douglas Madsen and Peter G. Snow, The Charismatic Bond: Political Behavior in Time of Crisis (Cambridge, MA: Harvard University Press, 1991).

(28.) James McGuire, Peronism without Peron: Unions, Parties, and Democracy in Argentina (Stanford, CA: Stanford University Press, 1997); Levitsky, "Crisis, Party Adaptation, and Regime Stability in Argentina."

(29.) Angelo Panebianco, Political Parties: Organization and Power (Cambridge: Cambridge University Press, 1988), pp. 262-267.

(30.) Kenneth M. Roberts, Deepening Democracy? The Modern Left and Social Movements in Chile and Peru (Stanford, CA: Stanford University Press, 1998).

(31.) Maurice Duverger, Political Parties (London: Methuen and Company Ltd., 1964), p. 25.

(32.) Leon D. Eckstein, Political Parties in Western Democracies (New Brunswick, NJ: Transaction Books, 1980), pp. 233-260.

(33.) Programmatic choices were the best defined and most varied in Chile, where clearly differentiated conservative, centrist, and Marxist alternatives each captured around a third of the electorate during the 1933-1973 democratic regime. In nations where traditional oligarchic parties dominated the electoral arena, such as Colombia and Uruguay, programmatic alternatives were much less significant. Nations with strong populist parties, such as Venezuela and Peru, generally occupied an intermediate position.

(34.) John Williamson, Latin American Adjustment: How Much Has Happened? (Washington, DC: Institute for International Economics, 1990).

(35.) Arturo Escobar and Sonia E. Alvarez, eds., The Making of Social Movements in Latin America: Identity, Strategy, and Democracy (Boulder, CO: Westview Press. 1992); Philip D. Oxhorn, Organizing Civil Society: The Popular Sectors and the Struggle for Democracy in Chile (University Park, PA: Pennsylvania State University Press, 1995); David Slater, ed., New Social Movements and the State in Latin America (Amsterdam: CEDLA, 1985).

(36.) Douglas Chalmers, Scott B. Martin, and Kerianne Piester, "Associative Networks: New Structures of Representation for the Popular Sectors?" in Douglas A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, and Monique Segarra, eds., The New Politics of Inequality in Latin America: Rethinking Participation and Representation (Oxford: Oxford University Press, 1997).

(37.) Herbert Kitschelt, The Transformation of European Social Democracy (Cambridge: Cambridge University Press, 1994).

(38.) Kenneth M. Roberts, "Neo-liberalism and the Transformation of Populism in Latin America: The Peruvian Case," World Politics 48 (October 1995): 82-116; Kurt Weyland, "Neo-populism and Neo-liberalism in Latin America: Unexpected Affinities," Studies in Comparative International Development 32 (Fall 1996): 3-31.

(39.) Susan C. Stokes, Cultures in Conflict: Social Movements and the State in Peru (Berkeley: University of California Press, 1995).

(40.) Barbara Geddes, "A Game Theoretic Model of Reform in Latin American Democracies," American Political Science Review 85, 2 (June 1991): 373.

(41.) Barbara Geddes, Politician's Dilemma: Building State Capacity in Latin America (Berkeley: University of California Press, 1994).

(42.) Wayne A. Cornelius, Ann L. Craig, and Jonathan Fox, eds., Transforming State-Society Relations in Mexico: The National Solidarity Strategy (San Diego: Center for US-Mexican Studies, 1994); Carol Graham and Cheikh Kane, "Opportunistic Government or Sustaining Reform? Electoral Trends and Public-Expenditure Patterns in Peru, 1990-1995," Latin American Research Review 33, 1 (1998): 67-104; Norbert R. Schady, "The Political Economy of Expenditures by the Peruvian Social Fund (FONCODES), 1991-95," American Political Science Review 94, 2 (June 2000): 289-304.

(43.) Valenzuela, Political Brokers in Chile, p. 154.

(44.) Denise Dresser, Neo-populist Solutions to Neo-liberal Problems: Mexico's National Solidarity Program (San Diego: Center for US-Mexican Studies, 1991); Steven Levitsky, From Laborism to Liberalism: Institutionalization and Labor-Based Party Adaptation in Argentina (1983-1997) (Unpublished doctoral dissertation, Harvard University, 1999).

(45.) From El Mercurio, 2 June 1999, p. C2.

(46.) Norbert Lechner, "The Transformation of Politics," in Aguero and Stark, Fault Lines of Democracy.

(47.) Roberts and Wibbels, "Party Systems and Electoral Volatility in Latin America." (48.) Katz, "Party Organizations and Finance."

KENNETH M. ROBERTS

University of New Mexico

Table 1

Net Electoral Discontinuity in Latin American Party Systems,
1978-1999 *

 Presidential Legislative
Country Elections Elections

 % %
Peru 93.7 85.8
Ecuador 88.8 77.5
Venezuela 82.2 51.1
Nicaragua 59.9 55.6
Brazil 70.2 40.8
Bolivia 48.8 39.2
Dominican Republic 38.1 35.9
Mexico 29.4 34.4
Panama 31.7 21.2
Argentina 26.8 25.2
Uruguay 23.6 22.3
Colombia 31.0 12.0
Chile 22.2 16.6
Paraguay 21.3 13.2
Costa Rica 6.4 6.9
Honduras 2.3 2.8

 Average
Country Discontinuity

 %
Peru 89.8
Ecuador 82.8
Venezuela 66.7
Nicaragua 57.8
Brazil 55.5
Bolivia 44.0
Dominican Republic 37.0
Mexico 31.9
Panama 26.5
Argentina 26.0
Uruguay 23.0
Colombia 21.5
Chile 19.4
Paraguay 17.3
Costa Rica 6.7
Honduras 2.6

* Calculated as the net shift in vote shares for all the
 party system from the first to the last election during
 the period from 1978 to 1999.