Abstract
From 2000 to 2010, Bolivia and Peru underwent similar processes of political decentralization toward the meso level of the government. Three elections later in Peru and two in Bolivia, the ability of national political parties to articulate interests differs markedly between the two countries. Peru tends toward fragmentation with national parties incapable of participating or successfully competing in subnational elections, while in Bolivia, the Movimiento al Socialismo (MAS) – and other parties to a lesser extent – are increasingly capable of participating and winning subnational offices. This paper argues that, despite having undergone very similar institutional reforms, the difference between the cases can largely be explained by two “society-side” variables: the caliber of the political ideas in debate and political social density. The substantive quality of ideas in debate and a greater political social density have been crucial to the Bolivian trend, while their absence has lessened the possibility of anything similar occurring in Peru. In general terms, the article sheds light on the social conditions that favor party-building in a context of decentralization reform.
Introduction
Decentralization reform in Latin America began in the 1980s; however, it was not until the end of the nineties that these reforms became a major concern for political scientists. Along with other angles, decentralization has been studied as an independent variable to explain the dynamics of political competition and, more precisely, the consequences of reform for party politics (Grindle 2000; Sabatini 2003; Harbers 2010, among others). Essentially, the case has been made that, ironically, “many of the parties that pushed for decentralization are simply not well-equipped organizationally to deal with its political consequences” (Sabatini 2003: 149). In this article 1 I analyze the contrasting experiences of Bolivian and Peruvian national political parties as they attempted to adapt to similar decentralization processes launched in both countries in the 2000s decade. Through this analysis, I seek to examine in greater depth the complex dynamics linking decentralization reforms and party weakness. How well have they performed since the decentralization reform was set in motion? Have decentralization reforms affected them in a homogeneous way? Do they succeed in aggregating interests from multiple levels of government? Finally, if parties display divergent trends when faced with very similar institutional reforms, what accounts for the differing outcomes?
I would like to thank friends and colleagues that helped me to improve this article with their comments: Manuel Balán, Graciela Ducatenzeiler, Philip Oxhorn, Steven Levitsky, Françoise Montambault, Salvador Romero, Eduardo Dargent, Tatiana Acevedo, María Inés Vásquez and three great anonymous reviewers. Of course, all mistakes and inaccuracies are my sole responsibility. Amy Beeson helped me to correct my ruinous English and Daniel Encinas collected some of the data, many thanks to both. The research was conducted with the support of the Trudeau Foundation.
During the first decade of 2000s Bolivia and Peru went through similar processes of political decentralization, creating comparable meso levels of elected authorities between the national and local governments. Three elections later in Peru (2002, 2006 and 2010) and two in Bolivia (2005, 2009), the performance of Peruvian parties at the subnational levels is distinctly poor, while in Bolivia the Movimiento al Socialismo (MAS) – and other parties to a lesser extent – seem to have taken advantage of the new levels of electoral competition, even improving their performance between the two elections. To illustrate the contrast using the two main parties in each country: in the 2002 regional election the Partido Aprista Peruano (known as APRA) won 12 regions out of 25, but only one region in 2010; on the other hand, MAS won three regions out of nine Bolivia's 2005 regional election, but increased to six four years later. Hence, in Bolivia there is a trend toward political parties articulating the political system through levels of government, while Peru tends toward fragmentation. I will argue that in a context of “revival of subnational politics” (Falleti 2010) parties must aggregate interests not only at the national level (as it is implied in the “party nationalization” literature) but also between and among levels. I will show that Bolivian parties, especially MAS, are performing this function better than Peruvian ones, despite a very similar context of institutional reform.
Why do some party systems adapt more effectively to decentralization reforms than others? What are the social and political conditions that allow national political parties to adapt to new, more complex forms of political competition? And what are the conditions for the emergence of new national political parties in this new institutional context? Applying these questions to my primary cases: what explains that in Bolivia certain political parties are improving their capacity to aggregate interests from multiple levels of government despite the decentralization reform? And why do all national political parties in Peru perform poorly at the subnational level, faced with a very similar institutional reform as their Bolivian counterparts? What explains these divergent outcomes?
I argue that the crucial difference lies in the social realm. While John Crabtree has recently argued that
what appears to be missing in the Peruvian context is the existence of a party – such as the MAS in Bolivia – capable of articulating these various different interests in the language that is most effective (Crabtree 2010: 378),
in this article I argue that the difference between the two countries lies less in the presence or absence of such a party than in the social conditions for the emergence of such a party. I highlight two specific factors. First, I focus on the caliber of ideological disputes in both countries. I posit that the more intense the political debate in a society, the better the chances of building political parties that represent at the sub-national levels the national ideas in conflict. Second, I argue that a high “political social density” (this is, the existence of a politicized network of civic organizations) provides greater opportunities for parties to build on that social density. Hence, the main argument put forth in this article is that, even though decentralization has an undeniable impact on political parties, this effect is mediated by two society-side variables that must be considered in an account of the consequences of institutional reforms on party politics.
The article proceeds as follows. The following section reviews the literature on the impact of decentralization on parties and argues for the importance of assessing political parties’ abilities to integrate interests from multiple levels of government. Then, I develop two society-side variables that should be integrated in the institutional analyses linking decentralization and political parties. In the third section, I apply this analytical framework to the two cases under study. Finally, the conclusion points out that the harsh political conflict of recent years in Bolivia has helped institutionalize political organizations, while the absence of such conflict in Peru has reduced the necessity of parties. In this sense, Bolivia appears to be politically unified by conflict while Peru's politics have been fragmented by consensus. More generally I suggest that institutional accounts of party-building must take into consideration the social and historical settings in which institutional reforms are implemented.
Variables and Theory
Decentralization and its Consequences
Since the end of the 1990s, many studies on Latin American politics have addressed decentralization processes. 2 Decentralization reforms became an important independent variable in the context of a broader “revival of subnational politics” (Falleti 2010). Among other dependent variables, the impact of decentralization was assessed on democratic quality at the subnational units (O'Neill 2006), the responsiveness and good governance of local governments (Faguet 2009), the consolidation of subnational authoritarian enclaves (Gibson 2005), and the power achieved by the subnational units after decentralization (Falleti 2010). Other works dealt with the impact of decentralization on Latin American political systems, and especially on party politics. In this article I engage with this specific literature.
I define decentralization as “a reform (or series of reforms) that increases political power through the election of sub-national officials where they have been previously appointed and that also accords some level of autonomous fiscal power to those elected officials” (O'Neill 2006: 174).
From this literature, despite differences in how each work conceptualizes decentralization and in the focus on different causal mechanisms that link decentralization and political party variables, it is possible to extract a core proposition. Namely, that by and large, decentralization reforms have the unexpected consequence of weakening national political parties by fragmenting the institutional framework of political competition.
A first group of studies analyzes party systems as the dependent variable, supporting a mostly negative impact of decentralization reforms. This literature is related to the work of Chibber and Kollman (2004) who argue that national party systems are prone to emerge when economic and political power resides in the national government: “party aggregation should be easier under conditions of centralization” (Chibber and Kollman 2004: 79). Contrario sensu, under conditions of decentralization, the creation of national party systems should be harder. This is what has been stated for some Latin American cases in recent years. Harbers (2010), in a statistical investigation of 16 Latin American countries, finds that political and fiscal decentralization inhibit the development of national party systems since those reforms disperse resources and responsibilities, providing incentives for the emergence of competing subnational party-systems. 3 A similar argument was made about the case of Costa Rica, where decentralization seems to generate, among other unintended consequences, “party system fragmentation” (Ryan 2004). In the same vein, based on the analysis of contemporary patterns in Argentina's political system, it has been stated that although we do not know why, “we know that decentralization conspires against party aggregation” (Leiras 2006: 37). Finally, recently, Dargent and Muñoz (2011) have suggested that the Colombian party system was directly harmed by electoral and decentralization reforms since they drastically reduced the resources that parties used to hand out at the subnational level.
Harbers (2010) uses the term “nationalized party-system” in the sense of Mainwaring and Jones (2003).
Another group of scholars argues that decentralization reforms have an impact on individual parties. Focusing on Latin America as a whole, Sabatini (2003: 149) states that parties that pushed for decentralization did not have the capacity to deal with the consequences of that reform. Tanaka (2005: 54-76) proposes a similar argument in his comparative study of institutional reforms in the Andean countries, where he finds that reforms aiming to “open” political systems – among them decentralization reforms – have had the unintended consequence of weakening national political parties. In Peru, the decentralization reforms increased pressure on already-weak parties by pushing them to participate in additional subnational levels (Vergara 2009). In Venezuela, decentralization destroyed the ability of party leaders to dominate their bases by removing their capacity to nominate party members for office, consequently weakening national parties (Penfold-Becerra 2004; Lalander 2004). Others argue that national parties are indirectly weakened by decentralization, which encourages the emergence of subnational parties to compete with them at the meso level (see Bracanti 2007).
From this literature which suggests that decentralization conspires against national parties (whether as party systems or mere parties), we can hypothesize, at the very least, that the overall pattern for national political parties competing in a context of decentralization reform would be one of weakening. However, as I illustrate in the next section, the experiences of Bolivian and Peruvian parties do not easily fit into this pattern. Why in the last decade have all Peruvian parties suffered in their performance at subnational levels while the Bolivian MAS – and to a lesser extent other parties – have improved theirs?
This question is both theoretically and normatively relevant. If an important block of studies is pointing out the difficult conditions that some institutional reforms are inflicting on a crucial actor of democratic systems (parties), we must both assess that impact on party politics and explore ways in which we can mitigate such negative and unintended consequences.
Parties and their Ability to Aggregate Interests from Multiple Levels of Government
In a widely influential study, Mainwaring and Scully point out that “institutionalized parties must be territorially comprehensive” (Mainwaring and Scully 1995: 15-17). Given that a primary function of parties is to aggregate interests (Almond and Coleman 1960: 38; Linz and Stepan 1996: 17), Mainwaring and Scully emphasize the importance of national parties aggregating interests throughout a whole country. In the absence of a party system that effectively aggregates interests, government stability can be threatened by social conflict which proliferates, lacking a representative channel through which social cleavages are mediated (Coleman 1995; Mainwaring and Jones 2003).
The question remains, which interests should parties ideally aggregate and at what level(s)? According to classical theory, a party should aggregate interests from classes, individuals and/or pressure groups. In new decentralized contexts, parties accustomed to dealing with centralized structures face new challenges; in addition to social interests spread horizontally over a national territory, parties must aggregate interests vertically from multiple levels of government. Thus, in the context of Latin American decentralization, it is not enough that parties are “territorially comprehensive”; they must also be vertically consistent, or consistent between levels. This is the main challenge that decentralization poses for parties henceforward.
Where and how can we observe parties’ capacity to aggregate national interests? Most studies focus on the legislature. Studies on nationalization of parties and on “party aggregation” consider that a party – or party system – is nationalized or not depending on the homogeneity of the distribution of its vote over the national territory as measured in the lower house (Mainwaring and Jones 2003; Payne, Zovatto, and Mateo Díaz 2006). That is, the party's national reach is measured over the whole national territory. However, in increasingly decentralized political systems, the national reach of a party depends not only on its presence in the legislature but also on its ability to aggregate interests arising from multiple levels of government. A party could be highly nationalized on a horizontal dimension (national legislature) but not on a vertical one (between levels of government), which would still pose a challenge for the articulation of the political system. To fill this gap in the literature, I propose to analyze the performance of national parties at multiple levels of government. By performance I mean the ability of national parties both to participate in elections at multiple levels of government and to win offices at these levels. Where parties perform well at the national level but poorly at the subnational levels, the polity will suffer from political fragmentation as each level of government will be left to political organizations without ties to other levels. Conversely, if parties perform well on multiple levels of government these organizations will tend to aggregate interests that arise on each level, helping to articulate the political system. That is, rather than analyzing only horizontal aggregation, I emphasize the importance of observing political parties’ capacity to aggregate interests vertically. As I will show, Bolivian and Peruvian parties do not equally succeed in this latter task. After undergoing very similar decentralization reforms, the divergent outcomes suggest that the effects of these reforms are mediated by society-side variables.
Two Society-side Variables
Although institutional reforms such as decentralization have been privileged in the study of party aggregation in the Latin American context, these variables do not impact polities and parties in a unique way. A shift in focus from institutional variables to social ones may be warranted in order to explain the differences observed in the outcomes. Institutional reforms are implemented in social contexts that can lead to unexpected and often undesirable consequences (see Grindle 2000; Weyland 2009). In order to tackle this issue, I propose to bring the society side of the equation back into our accounts of political parties; after all, the venerable line of research launched by Lipset and Rokkan (1967) stated that political parties were the political “translation” of sociocultural conflicts. If sociocultural conflicts in a society are the raw material for building political parties, it could be fruitful to look into societies to complete institutional explanations; not because social differences are destined to be translated directly in political ones, but because “the nature and magnitude of the socio-structural basis of the conflict must be viewed as basic conditions facilitating – to a greater or lesser extent – the translation” (Bartolini 2000: 19). With this in mind, I propose to analyze two society-side variables that can help explain different outcomes in party aggregation, despite similar institutional reforms: the quality of political ideas in debate and political social density.
Political Ideas in Debate
In social theory, the debate on political “ideas” and political “interests” is among the most frequently revisited questions. This article, rather than intending to establish whether ideas or interests are what really matter in politics – an inherently epistemological issue 4 – simply integrates political ideas into an account of party politics. Conflicts in societies are not just about interests; they are made out of “interests and/or weltanschauung” in Bartolini's (2000: 13) terms. In this line of thinking, ideas matter in political conflict and, thus, in the institutionalization of party politics.
For a discussion on “ideas” and “interest”, see Wendt (2000: 92-138).
Why are ideas important? Although politicians will often seek to disguise interests as ideas, ideas can be a powerful source of linkage between political groups and their members. According to Kitschelt (2000: 872), under specific circumstances, programmatic linkages, as well as clientelistic ones, enable the organization and institutionalization of relations of democratic accountability between politicians and citizens. Similarly, Berman (2006: 11) notes in her study on social democracy in Europe that “the development of parties cannot be understood without a focus on ideology.” Especially, parties “with a market-liberal and a Marxian Socialist ideology probably have the strongest bent toward programmatic competition” (Kitschelt 2000: 864-865). In other words, the salience of the left-right cleavage is important in articulating political competition. Thus, it can be hypothesized that the greater the programmatic clash on the market-economy issue, the greater the chances of institutionalizing political parties with ideological salience. If other strong ideological differences exist in addition, the chances for party-building are strengthened even further. Ideas have the capacity to establish an “imagined empathy” between political leaders and citizens vis-à-vis a political party 5 , which is also useful in creating a clear antipathy toward adversaries. In the same vein, very recently, Hanson (2010) has suggested that political ideologies are especially significant in the creation of political parties in contexts of chaos and change when ideologies provide a framework for rational individuals “to embrace a long run strategy of cooperation with other converts” (Hanson 2010: xiv).
I adapt the concept of “imagined empathy” from the Hunt's analysis on the human rights emergence in the eighteenth century (Hunt 2007: 32).
Even though I would not go so far as to echo Hanson's catchy formula, “no ideologies, no parties” (Hanson 2010: xv), I build on that literature focusing on ideas and parties to suggest that the stronger the programmatic differences that shape public debates (especially if the economic issue is salient among them), the better the chances of building cohesive political parties. Ideas debated daily in a national public sphere can give way to the issue salience that nurtures party identities. 6 Contrario sensu, a polity lacking substantive public debate about programmatic issues faces lessened possibilities for building parties. Although this literature focuses primarily on European countries, it has already been stated that the market economy issue can strengthen party building in Latin America, as Hagopian, Gervasoni, and Moraes (2009) demonstrates in the case of the institutionalization of the Brazilian Partido dos Trabalhadores (PT), perhaps the most successful new political party in Latin America. As will be argued, a main difference between the Bolivian and Peruvian cases lies on this idealistic realm; during the last decade the Bolivian public debate was dominated by very important ideological differences, especially a sharp dispute on “neoliberalism”, while in Peru this has not been the case. However, as Michael Mann (2004: 12) put it, “[w]ithout power organizations, ideas cannot actually do anything.” Let us turn now to the analysis of political social density, which also helps the emergence of cohesive political parties.
I bring in the idea of everyday public debates from Zuckerman's complaint about the literature on political cleavages that used to focus only on elections to gauge the cleavages; according to Zuckerman (1975: 242) other mundane, less obvious acts also shape political conflict.
Political Social Density
Since the modern democratic regime was first described, theorists have highlighted the importance of civic life. From Tocqueville to Putnam engagement in civic organizations has recurrently been pointed out as vital to a healthy democracy (Tocqueville 1981; Putnam 2000). Departing from literature that stresses civic engagement in the creation of subjective democratic assets, such as trust (on social capital and trust see Rothstein 2005), this section focuses on civic associations with political content that can be used by parties to reach a territory and, in some cases, may give way to intertwined parties and associations. Thus, the analysis does not deal with associations that provide moral assets to polities, but with associations that explicitly provide resources and recruit individuals to political parties.
Political parties have traditionally benefited from such a political social density. Linkages between political associations and the new political parties were important assets in the early moments of party-building in Europe. The interconnectedness of parties and civic associations with political content were so important that Duverger (1967) created the category “indirect parties” to refer to parties with the strongest linkages with political associations or corporate interests. Some examples of this experience include the Catholic associations that became gradually involved in politics giving way to the Christian Democratic Parties (Kalyvas 1996: chapter 2), the agrarian associations that formed the base for agrarian parties, and unions and labor organizations which were closely associated with the emergence of mass socialist parties (Bartolini 2000: chapter 6).
The relationship between such politically oriented associations and the emergence of political parties has also been important outside the European context; so important, actually, that Huntington stated that outside the European context, the linkages between parties and social organizations, unions and peasant associations were a requisite for strong parties (Huntington 1968: 410). These linkages were common in the process of party-building in Latin America during the twentieth century (Collier and Collier 1991). Peronismo emerged deeply entwined with the labor movement in Argentina (Ben Plotkin 2010); then, during the years of repression and the long exile of Perón (1955–1973) the Partido Justicialista (PJ) survived in the Unions to an extent that many of the Union leaders were almost as powerful as Perón (Ducatenzeiler 1980). In Bolivia, the Movimiento Nacionalista Revolucionario (MNR), which led the Bolivian National Revolution in 1952 and remained in office for twelve years, used its connections to the labor movement in order to establish a stable and larger coalition of government (Dunkerley 1984). Finally, the Peruvian APRA deployed a so-called “functional structure” that it used to link the party to different social sectors, such as students’ associations, professional groups and, of course, unions:
el aprismo se convirtió en el más importante vehículo de institucionalización del movimiento obrero, acentuándose – casi hasta la identificación – la relación entre el sindicalismo y esa organización partidaria (Pareja 1980: 115).
Hence, historically, unions and other social groups have provided different sorts of resources and affiliates to the political parties which enable them to be the primary connectors of state and society.
This relationship between political social density and party empowerment has recently been restated in Latin America. In the last years, the Argentinean PJ has renovated itself through an explicit alliance of the Kirchners with different unions and social movements. 7 In Brazil, the PT gathered strength from its relationship with strong grassroots and labor movements (Hunter 2010). Finally, as I will show later, MAS and to a lesser extent other Bolivian parties, are benefiting from their linkages to unions and other grassroots movements. Thus, “political social density” is an undeniable asset for party-building. Party-building is also supported by social density that facilitates the functions of reaching a territory, aggregating interests, diffusing ideas, and fostering a link between state offices and grassroots movements and citizens.
Elected president in 2003 with a mere 22 percent of the vote, Nestor Kirchner started in office as a weak president. In order to strengthen his presidency he reforged the alliance of the PJ with unions. This successful alliance has lasted and it is evident when, for instance, the powerful Secretario General of the Confederación General de Trabajadores, Hugo Moyano, says, referring to Cristina Kirchner, “La Presidenta es la jefa del Movimiento” (La Nación, November 1st, 2010); or when the chief of cabinet says “Esa CGT que hoy conduce Hugo Moyano es nuestra columna vertebral” (La Nación, November 2nd, 2010).
The following section links this theoretical literature on decentralization, party institutionalization and the society-side variables using the Bolivian and Peruvian cases during the 2000s decade.
Bolivia and Peru: Decentralization, Parties and the Two Society-side Variables
The aim in this section is twofold. First, it describes the very similar institutional reform implemented in Bolivia and Peru during the 2000s (decentralization to the meso level) and assesses the ability of political parties in Bolivia and Peru to aggregate interests from multiple levels of political competition. As will be shown, this ability diverges considerably between the two countries. Second, it suggests that the divergent outcomes might be explained by exploring the two society-side variables conceptualized in the previous section; these variables are important to integrate into the institutional analyses that link decentralization reforms and party building.
The 2000s Decade: Reforms and Context
The year 2000 was a watershed for the political systems of both Peru and Bolivia. In Peru, Fujimori fled to Japan, precipitating the collapse of the authoritarian regime and giving way, albeit belatedly, to a transition to democracy (McClintock 2006). Meanwhile in Bolivia, the “water war” in Cochabamba made clear that the democracia pactada (Mayorga 1992) inaugurated in 1985 had crucially weakened. In both countries, the millennium dawned on a traditional party system in shambles and weak presidents who launched decentralization reform to a meso level more out of political necessity than as a rational restructuration of the state. 8 Both Bolivia and Peru were among the few countries in Latin America lacking meso level governments (Falleti 2010: 27). Both in Peru and Bolivia, the state has traditionally been divided into administrative units called departamentos (nine in Bolivia and 25 in Peru). Historically, they represented a mere administrative-territorial division rather than a distinct political-electoral level. The regional authority was a prefecto departamental appointed by the Executive. In both countries this scheme was amended during the 2000s to make each departmento an official región with its own directly elected regional government. Thus, rather than the creation of new regions, both countries gave political life to the departmentos that used to be mere administrative units. To date there have been two regional elections in Bolivia (2005 and 2010) and three regional elections in Peru (2002, 2006 and 2010). How are the political parties in each country dealing with the institutional reforms that would be expected to damage their capacities according to the literature already revised? I am concerned with the national parties’ capacity to aggregate interests from multiple levels of government: Does the political system tend to have subnational demands unrelated to national political organizations (fragmentation) or, alternatively, are both subnational and national demands articulated through national political platforms? 9 Although the low number of elections held so far at this level requires caution in interpreting these trends in the political parties’ performance, there are some patterns that merit consideration.
For the Peruvian party collapse see among others Crabtree 2010; Levitsky and Cameron 2003. For the Bolivian party system collapse see among others Romero 2010; Mayorga 2005.
I consider a national party what the law (Bolivian or Peruvian) consider it is a national party. Unfortunately, the mere brand names that pass for political parties in both countries leads me to privilege the main parties in the in-depth analysis considering as “main” parties the ones with representation in the national Congress during the last decade (and in the Asamblea Constituyente for the Bolivian case).
After the ten-year rule of Alberto Fujimori, free elections brought about a return of long-absent Peruvian political parties to the headlines. To the surprise of all, they performed relatively well: APRA reached the runoff in the presidential election of 2001, with Lourdes Flores from the traditional Partido Popular Cristiano (PPC) close behind. The traditional parties also obtained auspicious and unexpected results in the parliamentary arena. After these results some hailed, perhaps a little too soon, the “rebirth” of Peruvian political parties (Kenney 2003).
During this same “democratic spring” the first regional elections were held in 2002. National parties achieved relative success on the new level of competition. In particular, APRA, the oldest and most organized Peruvian political party, and the conservative Unidad Nacional (UN) reached remarkable rates of participation: APRA had nominations in 23 out of the 25 regions, and the UN in 22. 10 The only national party that won several gobiernos regionales was APRA, taking 12 of the 25 contested. In total, the national political parties won 17. In a country where political parties had disappeared for a decade, such results were not a bad sign; it seemed that the opening of this new subnational political space had revitalized the forgotten political parties. On the other hand, although the national parties performed reasonably well, this new level of competition had also given way to meso level political organizations, called movimientos regionales or regional movements. These regional movements had 62 candidates in the 25 regions and won in eight of them. 11
The source for all polling data is the National Office of Electoral Processes in Peru (ONPE).
The law on political parties of Peru considers three types of political organization: national political parties, regional movements and local political organizations. National political parties can participate in all levels of government, regional movements in the Gobernaturas and municipal levels and local organization can only do so at the municipal level.
The next two elections (2006 and 2010) saw a dramatic decline in the performance of national parties at the meso level. The Peruvian political system became a very fragmented one characterized by levels of government, each with its own political and electoral movements. National parties do not win subnational offices, and worse yet, they hardly participate. APRA remains the only national party that consistently participates at the subnational level. However, after winning twelve regional governments in 2002, APRA got only two in 2006 and one in 2010. This calamitous trajectory – even more calamitous since the APRA was in power with Alan García between 2006 and 2011 – is confirmed in terms of votes: In 2002 APRA gathered 22 percent of the national vote in the meso level election and in 2010 it reduced to 10.2 percent. UN also underwent a severe decrease in its rates of participation, going from 22 candidacies in 2002 to eight and seven in 2006 and 2010, respectively. Even the new Partido Nacionalista Peruano (PNP) of Ollanta Humala has proven itself unable to reach subnational levels. After contesting every region in 2006, in the 2010 election, PNP competed in only a single region. The rest of the Peruvian political parties confirm the trend: No other national party but APRA that had participated in the previous regional elections reached more than 3 percent of the vote in a regional election. The other side of this trend is the rapid flowering of meso level political organizations or Movimientos Regionales. In 2002 they had 62 nominations; in 2006 77; and in 2010 a total of 151. In 2002 they won eight regional presidencies; four years later they more than doubled that, earning 21; and in 2010, they held steady at 19. In the last eight years an average of 80 percent of the regional governments has been in the hands of strictly meso level organizations with almost no connections to the national level. 12 National parties are confined to the national level of government having all but given up sub national competition. Since regional elections were introduced in Peru, political parties have ostensibly failed in their function of aggregating vertical interests from multiple levels of government. 13
This outcome after three elections has led some to say regional parties may be one of the last hopes for party renovation in Peru (de Gramont 2010).
The same trend toward fragmentation is evident if we take the municipal level; in the local elections of 2010, out of 195 provinces APRA won nine, Acción Popular seven, and UN and PNP none.
Bolivia
In Bolivia, as in Peru, when the meso level electoral game made its rentrée to the institutional framework of political competition, the traditional party system was collapsing after nearly two decades of “democracia pactada.” On 18 December 2005, after a troubled period in Bolivian politics, presidential elections and the first regional elections were held simultaneously. In presidential elections, Evo Morales and MAS prevailed with a historic 53.7 percent of votes. 14 The election became a moment of realignment (Romero 2007) that confirmed the existence of a territorial political cleavage between the east and west of the country (Centellas 2009; Laserna 2010). In 2005 Bolivia elected Prefectos in its nine departments. MAS had the highest participation rate at this level; it was also the only organization that presented candidates in all nine regions. Poder Democrático y Social (PODEMOS), the political organization that had Jorge Quiroga as its presidential candidate (who eventually finished second), participated in six regions, the Bolivian Unidad Nacional (UN) also participated in six departments, and the traditional MNR and MIR did so in four and three departments, respectively. This indicates that, in terms of ability to participate, the opening of the meso level had a similar effect as occurred in Peru in its first regional election in 2002: One national political party participated in all the regions while the rest could only do so in some of them. The results were mixed. Although MAS won by far the most votes (942,913 resp. 32.86 percent – compared to 577,711 for PODEMOS resp. 20.13 percent), both parties got three Prefectures. MAS secured three in the west of the country and PODEMOS three in the east. Hence, the same political groups that participated at the national level also performed well at the subnational one, helping aggregate interests from multiple levels of political competition in national platforms.
All electoral data from the Bolivian National Electoral Court (CNE).
In April of 2010 the second regional elections were held in order to elect Gobernadores (formerly Prefectos). The Bolivian subnational political landscape changed, though not dramatically. In terms of participation, MAS once again presented candidates in all the Bolivian departments and doubled its number of Gobernadores with six (in Peru, APRA did the opposite). In terms of votes, its consolidation at regional level was also outstanding: In 2005 it obtained just under a million votes and in 2010 1,870,652 (50.10 percent). The results of both 2005 and 2010 elections suggest the presence of a hegemonic actor at the national and subnational level: MAS. Although MAS is the clear leader in articulating multiple levels of government, other parties also perform better than their Peruvian counterparts, helping articulate the political system. The MNR continues to participate at the subnational levels, although unsuccessfully; UN, although very dependent on its national leader Samuel Doria, performs at national and subnational levels with similar success; and the last elections have witnessed the emergence of Movimimiento sin Miedo (MSM) which is trying to build a national structure from below. MSM participated in six of the meso levels contests and at the municipal level it presented candidates in the nine Bolivian Departamentos and won big municipalities such as La Paz and Oruro in the 2010 election. 15 In this trend toward aggregation, it is also important to note that in the three eastern regions where MAS failed to win the governorship in 2010 (Santa Cruz, Tarija and Beni), the power remained in the hands of an incumbent that was re-elected, which implies a continuity of political opposition in these eastern districts against the MAS government. 16
An important point vis-à-vis the MSM is that its charismatic leader, Juan del Granado, did not attempt to be re-elected as mayor of La Paz; in 2010 MSM won La Paz without its leader what indicates a wish to institutionalize the party.
As in Peru, these trends are confirmed by the local level. In 2004 municipal election, MAS won 112 municipalities (out of 337) and in 2010 obtained 225 municipalities (68 percent), including many in the eastern departamentos.
In conclusion, although the small number of meso level elections in each country prevents us from making drastic conclusions, a divergent trend is perceived between the two in terms of aggregation of interests after the introduction of decentralization reforms. Why this difference?
Before getting into the main argument of this article, some alternative explanations to the puzzle must be analyzed. First, some authors point to the crucial Evo Morales’ leadership in building a party such as MAS (Rousseau 2010). Although his leadership is undeniable, Peru's APRA also benefited from the leadership of President Alan García, by far the most important politician in Peru. Hence, strong leadership alone seems not to ensure party building. Another group of scholars point to the presence of neoliberal reforms as an important catalyst for parties such as MAS or MSM (Hylton and Thomson 2007). However, both Bolivia and Peru are among the countries that suffered “aggressive” neoliberal reforms (Corrales 2003), making it difficult to draw any conclusions directly based on this variable. Finally, some suggest, in an alternative institutionalist explanation to the decentralization one already explored at length, that the simultaneity of elections could have a positive impact on party building (see Leiras 2006: 101). However, the first subnational election in Bolivia occurred simultaneously with the national, while the second was not, and MAS still improved its performance. In Peru, the best and the worst performance by parties both occurred during non simultaneous elections, undermining this possible explanation.
Two Society-side Variables
Public Sphere and Ideas in Debate
In the recent years Bolivia has undergone a cycle of fundamental ideological conflicts. As I suggested in the first part of the article, the higher the degree of political confrontation, the greater need that exists for political parties that can engage in those debates. In Bolivia, one of the main issues up for dispute during the last decade has been the economic model, with a narrative of resistance to neoliberalism taking center stage (on the narratives of neoliberalism in Bolivia see Haarstad and Anderson 2009). Kitschelt (2000) and Hagopian, Gervasoni, and Moraes (2009) pointed out the importance of this particular issue salience for party building and the Bolivian case seems to confirm their hypothesis. Public national debates on such crucial issues help bring subnational units into the national sphere. The sharp dichotomy between a free market versus a state-planned economy (implicit in the narrative about neoliberalism) polarizes debate on subnational levels and naturally connects the national parties with those levels; e.g. as everyday national debates gain importance at the subnational level, the presence of MAS in each one becomes rational. The condemning narrative about “neoliberalism” that was in the air in Bolivia since the end of the 1990s was key in articulating an “imagined empathy” that increases the likelihood of party building.
In addition to the economic regime, the political regime has also been subject to debate in Bolivia. At least three other major polarizing issues have characterized the Bolivian public debate over the last decade: the centralization versus decentralization issue, the democracy versus authoritarianism issue, and a communitarian versus liberal approaches to citizenship (see Urenda 2009; Romero 2010; Postero 2010). Each of these fundamental debates is relevant at the national and subnational levels making it easier for organizations and interests at the national and subnational levels to articulate rather than fragment. In the face of such crucial debates, the people's voice has shone through many times in recent years. Bolivians have been summoned to the polls several times to vote on specific, sensitive issues such as the scheme of exploitation of hydrocarbons or the system of regional autonomies. On the issue of autonomies alone, Bolivians have voted in three referendums in the past five years (2006, 2008 and 2009). They have voted in another poll on energy policy (2004) and another to support or dismiss their regional authorities and the President Evo Morales (2008). As Bolivian political issues were put to the people's direct choice, there was also a transference of “la polarización y el conflicto que caracteriza actualmente las relaciones entre los partidos y los líderes políticos” (Laserna 2010: 32). Additionally, Bolivians went through a long constitutional reform process from 2007 to 2009, in which every aspect of Bolivian life has been subject to discussion. 17 The Constitutional Assembly concluded with a referendum to approve the new constitution. The magnitude of the discussion brought citizens, social organizations, political parties and a “free and vibrant” national press (Lehoucq 2008: 122) into the debate. Local and regional arenas became an extension of the national dispute. As reported by a Santa Cruz intellectual, counter intuitively, the sharp crisis of the 2000s and the cleavage that divides the country in east and west have been “un momento bolivianizante”. 18 To put it otherwise, the divisions have built a common public arena of debate of which parties can benefit.
The changes have been so deep that not even the country's name has remained the same: It went from “Republic of Bolivia” to “Plurinational State of Bolivia”.
Claudia Peña, personal interview, Santa Cruz de la Sierra, November 2009.
In Peru, on the other hand, the national arena is characterized by a striking absence of ideological debates that could link the national and subnational arenas:
la esfera pública en el Perú está muy fragmentada, cada departamento tiene sus propios temas y el nivel local también, ninguna de las discusiones de nivel nacional se reproduce en los niveles sub-nacionales. 19
Adriana León, Peruvian media researcher from the Instituto Prensa y Sociedad, personal interview, June 2011.
The biggest ideological confrontation in recent years occurred in the context of 2006 general election when the candidate Ollanta Humala challenged some of the prevailing ideas on democracy and free market economy. His insistence on “nationalizing strategic enterprises,” the explicit endorsement of his candidacy by Venezuelan President Hugo Chavez and a context in which many Latin American countries undertook a “left turn”, catalyzed a very brief discussion about the Peruvian economic model. However, since this election in which Humala did not win, no substantive political discussion has surfaced. No polarizing debates have generated considerable waves in public opinion, political leaders or social organizations. 20 If there are no national debates with issue salience, no discussion of ideas can take place in the national arena, much less trickle to the subnational levels. In the absence of national issue salience, subnational levels tend to focus on local and non-ideological concerns. National parties have little to offer on the subnational levels; while regional and local movements are better suited to create agendas that are locally relevant in the absence of national debates. Hence, the contrast between Peru and Bolivia allows us to confirm that the deeper the ideological dichotomies at the national level, the greater the involvement of subnational levels in issues of national importance. If ideological debates are important, they also decentralize, facilitating articulation or at least preventing fragmentation.
While I prepared this article Peru was in the middle of the presidential campaign and this was again the case, there were no important ideological debates, and even the once-upon-a-time radical Humala showed a very moderate economical and political discourse in this campaign.
Political Social Density
In Bolivia, popular participation in local governments, community organizations and political organizations has traditionally been very high (see Moreno Morales 2008). As Gray Molina argues in a compelling thesis, in Bolivia there exists a modus vivendi between society and state such that neither has been imposed on the other. Thus, “mobilization, rebellion, and organized contestation are structural characteristics of state and institutional development in Bolivian society” (Gray Molina 2008: 124). This organized and mobilized society has consistently played an important role in the country's political life. During the national revolution of 1952, the leading party's (MNR) leaders quickly realized that they could not govern if the government did not include the labor union Central Obrera Boliviana (COB). Soon they also saw the need to incorporate Bolivian peasants who were organized in unions into their coalition (for the National Revolution see Malloy 1971; Dunkerley 1984). Although the MNR government tried, it was never able to co-opt labor and peasant organizations (Lavaud 1991). These mobilized organizations managed to survive the end of the revolutionary regime in 1964. Similarly, during the long years of authoritarianism (1964–1982), the miners’ movement could not be eradicated despite brutal crackdowns against them. Peasant organizations were able to remain a key part of the Bolivian political framework when the military regimes of the 1960s relied on them to govern in an alliance known as the Military-Peasant Pact (Dunkerley 1984: 132-134).
The most recent piece in the interweaving histories of social density and political regimes in Bolivia has been the success of the MAS and its relationship with the coca growers’ unions, indigenous groups, and urban associations (Van Cott 2005; Madrid 2008; Anria 2010). The existence of such a complex political social density allows nascent political groups to overlap with organizational structures and to benefit from them in political competition. In the case of MAS, “these organizational ties provided the MAS with legitimacy in many indigenous areas as well as a network of activists and supporters” (Madrid 2008: 486). Cocaleros, a dense network of peasant unions that started when thousands of politicized former miners arrived to the trópicos from the bankrupted mine industry of Bolivia, are the main social base for the MAS (Klein 2003: 246-248). In urban settings, MAS also benefited from the Bolivian political social density. As Anria states,
organizations representing artisans, microenterprise, pensioners, cooperative miners, and other urban sectors perceived the alliance with MAS as a unique opportunity to achieve parliamentary representation (Anria 2010: 119).
In El Alto, the key city surrounding La Paz, MAS established a link with powerful associations such as Federación de Juntas Vecinales or the Central Obrera Regional-El Alto. MAS infiltrated those associations that “allow political parties to extend their influence and control throughout the territory and to recruit leaders that mobilize large numbers of people” (Anria 2010: 119). Although to characterize the Bolivian Executive as the “Gobierno de los Movimientos Sociales encabezado por el Presidente Evo” (García Linera 2011) is an exaggeration, what is undeniable is that the Bolivian political social density has helped in empowering and institutionalizing the MAS. Since the social organizations that have helped to enhance MAS are widely expanded and have strong local presence, it is not surprising that MAS has succeeded in aggregating interests from multiple levels of political competition.
Political social density also serves to facilitate the arrival of individuals and grassroots organizations to the national level of political representation. A cursory look at the Bolivian congress confirms this statement. According to a survey of Latin American parliamentary elites conducted by the University of Salamanca in 2006, the percentage of Bolivian congressmen that came from unions was 21.4 percent, while 17.6 percent came from a citizen, peasant or miner organization. Also, 74 percent of Bolivian congressmen in 2002 were previously municipal councilors. 21 Finally, we must not forget that many of the political leaders of Bolivia do not cease to lead civil social organizations while holding their political positions. The best example is Evo Morales himself, who is not only president of Bolivia but also the executive secretary of the six federations of coca growers of the Chapare region. Though this social density has been most often attributed to western Bolivia, in the east there is also evidence of important social density, even if it has not (yet) translated to a national political project. In each eastern departamento, the so-called comités de defensa departamentales have a rallying power that – though diminished in the last couple of years – would still be the envy of most Latin American political parties. On several occasions, the eastern departments have gathered enormous protests against Morales’ policies, exhibiting a great amount of legitimacy in their departamentos (Eaton 2007; 2011). Unlike the west, where social density is more union-based, in the eastern departments it is built on the productive private sectors and its leaders (Seleme 2008). Thus, regional organizations in the east are a second component of political social density in Bolivia. In summary, in Bolivia's political life it has traditionally been hard to distinguish between state and civil society, or, as Van Cott has argued, there exists a “fusion of state and civil society” (Van Cott 2008: 184).
Another recent study shows consistent results. A full 84 percent of the congressmen elected in 2005 had experience as community leaders, 44 percent were union leaders and 18 percent were leaders of social movements. Mendoza Botelho (2010).
If robust Bolivian social density has historical roots, so does the Peruvian weakness. According to Mauricio Mulder at the time he was Secretario General of APRA:
[El Apra] perdió la columna vertebral “funcional”, y el partido se quedó con la organización territorial como única columna sólida. […] Sufrimos una crisis de representación como consecuencia del golpe fuerte que nos dio el gobierno Velasco [The military authoritarian leader of Peru, 1968-1975]. Velasco al destrozar nuestras organizaciones funcionales, que las teníamos hasta ese momento muy fuertes en el ámbito universitario, sindical y gremial, nos hizo polvo y no pudimos recuperarnos ni siquiera en época del primer gobierno del APRA []. 22
Personal interview, Mauricio Mulder, APRA's Secretario General, August 2009.
The contrast with the Bolivian case is striking. MAS has benefited from precisely those social organizations that APRA lost. Moreover, not only has APRA lost contact with these social groups, but also political social density in Peru has weakened drastically on the whole during recent decades. The definition of political social density is useful: it is not that Peru has a small number of civic associations (Portocarrero et al. 2001). Rather, a qualitative observation shows that those organizations are not as politicized as in Bolivia. For example, Peruvian urban organizations tend to be instrumental to obtain minimum needs, then, “cuando tales necesidades mínimas de seguridad y urbanización son cubiertas, la organización tiende a debilitarse” (Portocarrero et al. 2006: 144). So, in contrast to Bolivia, civic organizations are instrumental and not politicized, which tends to make the link between parties and those associations difficult.
Many studies have found that Peruvian civil society weakened during the 1980s and 1990s (Yashar 2005; Burt 2009). The combination of the influence of the armed revolutionary group Sendero Luminoso, the antisubversive response, and aggressive neoliberal reforms delivered a heavy blow to the social sphere, which prevented the development of autonomous social movements (Levine and Romero 2006). For example, in Peru rates of unionization diminished constantly in the last three decades. Although figures for these rates are not absolutely reliable since there is no standard way of measuring, they do shed light to this issue. Unionized white-collar workers were 55.8 percent in 1981, 46.8 percent in 1991, and 10.2 percent in 2001. Blue-collar workers followed the same trend (see Verdera 2007: 228). By the end of 2000s the rate of unionization in Peru had fallen to a mere 4 percent. 23 In Bolivia, instead, according to a study,
Peruvian Ministry of Labor, online: <www.mintra.gob.pe/archivos/file/estadisticas/anuario/indice_anuario_2010.pdf>.
La tasa de sindicalización - medida como afiliados como porcentaje de los asalariados - no ha variado sustantivamente entre 1989 y 1999: de 25,7 % y 25,1 %, respectivamente (Kruse 2001: 168).
In Peru, according to Burt, fearful of being considered terrorists, potential Peruvian activists of the 1980s and 1990s took refuge in the private forum (Burt 2009). Acceptable civic participation had to be “despolitizada y ‘no perniciosa’ de los ciudadanos” (Panfichi 2007: 24).
This inhibition of social movements and their consequent lack of impact on national politics can be assessed by looking at the political origins of Peruvian congressmen. Unlike Bolivian congressmen who largely came from organized civil society, the same survey conducted by the University of Salamanca on parliamentary elites in Latin America shows that in Peru (taking the congressmen elected in 2006) only 6.4 percent of congressmen came from unions and just 3.2 percent from neighborhood movements. This contrast with Bolivia is useful to highlight the importance of low political social density, which in Peru prevents the engagement of grassroots leaders with nationwide organizations and debates, leading to the fragmentation of politics. The interests of the various levels are simply stymied at their respective level.
In conclusion, low political social density in Peru diminishes the capacity of Peruvian parties to build on this kind of social density. Parties lack social footholds from which to aggregate interests from subnational levels of government.
Conclusion
This article puts forth three main considerations. First, it suggests that in a context of decentralization, the literature related to party system institutionalization and party nationalization must observe the national parties’ capacity to perform at the different levels of government and not just at the national arena. Parties can achieve homogeneous vote percentages at the national level when comparing legislative and executive elections (horizontally) while performing poorly at the subnational levels (vertically). The parties’ primary function of aggregating interests must include those interests coming from the recently empowered subnational units.
Second, a group of works on Latin American politics has highlighted that decentralization conspires against political parties. Examining the capacity of parties to aggregate interests vertically, I assessed this hypothesis in the cases of Bolivia and Peru, which underwent similar and simultaneous decentralization reforms toward the meso level of government. Although the small number of elections held to date prevents a definitive conclusion, this article shows that in Bolivia, contrary to what one might assume from the firmness of the east/west cleavage, we observe a trend in which MAS – and to a lesser extent other parties – are aggregating interests from both national and subnational levels of political competition. In Peru, on the other hand, where there is no similar cleavage structuring political life, the political system lacks national political parties that can articulate the various levels of government. In summary, in Bolivia we see a trend towards aggregation of interests; and in Peru, a trend towards fragmentation. The Bolivian political system seems united by discord while the Peruvian one is divided by consensus.
In a more theoretical vein, I showed that similar institutional reforms can generate very different effects depending on the setting in which they are implemented. Countries are not test tubes where similar interventions necessarily produce similar results. Hence, while reforms in both countries were similar, the divergent outcomes are due to two society-side variables. First, there are “political ideas”: The caliber of the ideological debate in the public arena influences the possibility of having national political organizations that aggregate interests from both national and subnational levels of political competition. The more crucial the issues at stake in the public debate, the greater the chance of having comprehensive political parties that give order to them. Second, there is “political social density”: The existence of politicized social organizations to mobilize ideas, resources and individuals across levels of government plays an important role in explaining divergent outcomes. Each variable can moderate, block or invigorate similar institutional reforms, generating unexpected and opposed results. In other words, although there is a causal link between institutional reform and the outcome under analysis, I argue that this relationship is mediated by societal factors.
As taught by the old political sociology, major social conflicts are the raw material for political parties and party systems (Lipset and Rokkan 1967; Bartolini 2000; Caramani 2004). A theoretical framework that neglects the importance of social conditions for the emergence of parties will incompletely depict political competition. We must look at countries’ politics beyond electoral results and formal institutions. The relevance of the ideas in the everyday public debate and the involvement of social organizations in politics enable the articulation of the political system through parties that can aggregate interests. Usually, those dimensions are rooted in particular historical trajectories, suggesting that political scientists must make room for historical analysis of the countries where institutional reforms are implemented and elections held. An outcome triggered by institutional reforms is not the result of the reform itself but of the interaction of the reform with the social and historical setting where it is launched. As political scientists we are still looking for what James C. Scott (1998) once called the “missing link” between state reforms and political outcomes.