The Emergence and Performance of Indigenous Peoples’ Parties in South America
Comparative Political Studies Volume 39 Number 6 August 2006 709-732 © 2006 Sage Publications 10.1177/0010414005285036 http://cps.sagepub.com hosted at http://online.sagepub.com
A Subnational Statistical Analysis Roberta Rice University of New Mexico, Albuquerque
Donna Lee Van Cott Tulane University, New Orleans, Louisiana
One of the most significant and unexpected developments in Latin America during the past 10 years is the emergence of parties organized around indigenous identity. The authors use subnational data from six South American countries (Bolivia, Chile, Colombia, Ecuador, Peru, and Venezuela) to examine the factors responsible for the variation in the emergence and performance of indigenous peoples’ political parties in the region. Using a pooled cross-sectional twin snapshot analysis, the authors find that although indigenous party formation is the result of favorable institutional, demographic, and political conditions, such as permissive electoral rules, optimal indigenous population size, and a regional diffusion effect with respect to indigenous activism, enhanced electoral performance of these parties is determined by structural and political conditions, including higher rates of poverty and less salient class-based identities, in addition to the favorable conditions mentioned above. Keywords:
indigenous, political party, ethnic parties
Authors’ Note: We thank Mark P. Jones, José V. Molina M., David Samuels, and María Escobar-Lemmon for help obtaining access to data, Kenneth M. Roberts for advice and comments on a previous draft, Kimberly Nolan for statistical assistance, and Andrew Gunnoe for research assistance. Rice’s research was supported by a Latin American and Iberian Institute (LAII) PhD Fellowship, a Graduate Research Project Grant from the Graduate and Professional Student Association, and a Dylan Balch-Lindsay Memorial Research Award from the Department of Political Science of the University of New Mexico. Van Cott’s research was funded by a 1997-1998 Fulbright dissertation fellowship, and two professional development awards from the University of Tennessee, Knoxville.
709
710
Comparative Political Studies
P
olitical parties are crucial institutions for ensuring the representation of society in democratic political systems. In most Latin American countries, however, the quality of representation secured through parties is poor. Many parties lack linkages to society, and many social cleavages are underrepresented or poorly represented by parties (Roberts, 2001, p. 17; 2002a, 2002b). Traditional links between parties and constituents eroded in the 1990s as economic austerity depleted the coffers of clientelism, class identities became less meaningful, and many parties eschewed coherent ideological and programmatic appeals in favor of personal charisma and vaguely patriotic rhetoric. As a result, the region’s party systems became increasingly fragmented and incoherent, and many long-dominant parties collapsed (e.g., Bolivia’s Democratic Nationalist Action, Peru’s Popular Action and Popular Christian Party, and Venezuela’s Democratic Action and Social Christian Party; Coppedge, 1997b; Levitsky, 1999; Levitsky & Cameron, 2001; Mainwaring, 1999; Mainwaring & Scully, 1995; Roberts, 1998, 2002a, 2002b; Roberts & Wibbels, 1999; Tanaka, 1998). Amid this widespread party system decomposition, one of the most significant and unexpected developments in the past 10 years is the emergence of electorally viable political parties organized around indigenous identity.1 In Latin America, indigenous identity experienced a revival in the 1980s and early 1990s, as regional, national, and transnational indigenous social movements gained considerable public support and achieved impressive policy results, including the constitutional codification of special rights. In Bolivia, Colombia, Ecuador, Guyana, Nicaragua, and Venezuela, indigenous political parties have gained national legislative representation—often in their first electoral outing—while those in Argentina, Chile, and Peru are struggling for space at the local and regional level. The emergence and performance of indigenous parties indicate the political and organizational maturity of a population that had been excluded from politics or unable to overcome the efforts of nonindigenous elites to dominate and manipulate its political behavior. Most of the new indigenous parties are deeply rooted in dense networks of social organizations that channel the interests and preferences of indigenous peoples to political party leaders, who are relatively more responsive to them than traditional political party elites. Political scientists consider these qualities to be highly desirable (Mainwaring & Scully, 1995; Roberts, 2001, 2002a, 2002b). Thus, the central questions addressed 1. Indigenous peoples are the descendents of the original inhabitants of Latin America. They claim or maintain ties to traditional collectively held territories and continue to practice and seek to preserve customs, languages, and forms of social organization with preColumbian roots.
Rice, Van Cott / Indigenous Peoples’ Political Parties
711
by the current study are, why are indigenous parties emerging now in Latin America? And second, what factors explain the electoral performance of these parties when most other efforts to link voters to political parties are failing? In this article, we use what we term a pooled cross-sectional twin snapshot analysis to test various hypotheses that attempt to explain the unexpected emergence and performance of ethnic parties in Latin America in the 1990s. We contend that party emergence and electoral performance are influenced by institutional, demographic, and diffusion factors, and that party performance is further influenced by structural and political conditions. Specifically, we argue that the emergence of indigenous parties is the product of favorable underlying conditions, such as permissive electoral rules, relatively large indigenous population size,2 a regional diffusion effect, and changes in political conditions during the 1990s, although the relatively impressive electoral performance of these parties is the result of the aforementioned favorable conditions, and structural and political conditions, including higher rates of poverty and less salient class-based identities in society, as reflected in the vote share for leftist parties. The current study breaks new ground geographically and methodologically. Existing scholarship on new party formation has been limited largely to the stable democratic systems of developed countries (Hug, 2001; Kitschelt, 1989; Lipset & Rokkan, 1967; Schattschneider, 1960) and does not travel well to newly emerged democracies. Furthermore, much of the literature on ethnic parties has ignored Latin America, with the exception of Guyana, where the focus is on ethnic groups of African and Asian descent (Horowitz, 1985, pp. 311-318). Contemporary scholars of ethnic parties also have tended to rely on qualitative, comparative analyses, with a strong emphasis on the impact of democratic political institutions (Chhibber, 1999; Horowitz, 1985; Van Cott, 2005). Few statistical analyses of the emergence and performance of ethnic parties have been produced.3 The article begins with a discussion of the recent politicization of ethnic cleavages in Latin America and the attempt by indigenous movements in 2. Measures of indigenous population size are notoriously unreliable and tend to vary in methodology across countries. In some cases estimates vary wildly, ranging in Ecuador, for example, from 7%, according to the 2001 census (Secretarîa Técnica del Frente Social, 2001), to more than 40%, according to indigenous leaders. Moreover, using subnational data throughout the current study ignores variation at lower levels. Nevertheless, our subnational level of analysis reduces measurement error across countries and better captures within-country variation than would a large-N cross-national study. 3. Notable exceptions include Chandra (2004), and Mozaffar, Scarritt, and Galaich (2003), on India and Africa, respectively. Neither of these excellent studies, however, addresses directly the question of ethnic party emergence.
712
Comparative Political Studies
certain national contexts to form their own electoral vehicles. We then use subnational (i.e., state/department/provincial) data from six contiguous South American countries (Bolivia, Chile, Colombia, Ecuador, Peru, and Venezuela) to test the impact of institutional, structural, and demographic variables on the emergence and performance of indigenous peoples’ parties using a large-N statistical analysis. The combination of within- and between-nation comparisons of contiguous subnational units enhances the environment for theoretical generalizability while enabling us to control for cultural, historical, and political commonalities (Snyder, 2001, pp. 96-97). Subnational-level analysis is particularly appropriate for the study of this topic because most indigenous parties emerged in particular provinces, departments, or states where indigenous populations are concentrated, and their organizations relatively mature and effective. Furthermore, early electoral campaigns have tended to draw on networks of volunteers and grassroots, person-to-person contacts. Even those indigenous parties that manage to attain national-level success tend to draw strength from particular regions, while remaining relatively weak in others and spending relatively little on mass media advertising. This pattern of emergence differs from that of many modern Latin American political parties, which tend to depend more on television advertising and the popularity and fame of particular politicians to promote support. Therefore, in addition to giving us a larger number of cases than otherwise would be possible, using the subnational level of analysis enables us to focus on the real laboratory for the creation and performance of new indigenous parties: the department, province, or state. The article concludes with suggestions for future research on this new political phenomenon.
Indigenous Parties in Latin America Prior to the emergence of electorally viable indigenous parties, indigenous peoples in South America exhibited three types of party loyalty. Under conditions of scarcity or the hegemony of single parties, indigenous communities entered into patron-client relations with traditional parties. This model is typified by the relationship between Bolivia’s National Revolutionary Movement (MNR) and the highland indigenous population between 1950 and 1990, and by relations between Democratic Action (AD) in Venezuela (Van Cott, 2005). Traditional parties lost their hold on the indigenous electorate with the advance of more competitive elections in the 1990s, the loss
Rice, Van Cott / Indigenous Peoples’ Political Parties
713
of state revenue to be channeled into patronage, and the availability of more attractive, viable electoral options. Leftist political parties were the first to directly appeal to indigenous voters by addressing their needs as an exploited economic class. Since the 1930s, communist and socialist parties targeted indigenous movements for mobilization in peasant organizations and as key voting blocs in rural areas. This pattern emerged in Bolivia, Chile, Ecuador, and Peru, where strong ties still exist between the Left and the indigenous movement. However, relations between leftist parties and indigenous groups deteriorated in the 1990s as the former lost their electoral strength, and the latter grew tired of the racial discrimination and insensitivity toward cultural demands typical of these alliances. Finally, populist parties have attracted indigenous voters by offering economic changes that appeal to their class interests, and by attacking—at least rhetorically—the dominance of traditional elites. However, the mestizo leaders of populist parties typically disappoint indigenous voters when in office. The embrace and subsequent rejection of Ecuadorian president Lucio Gutiérrez by indigenous voters is typical of this relationship. Despite the ethnic diversity of the region, prior to the 1990s there were few attempts to form parties around indigenous, Afro-Latin American, or other ethnic identities, and none of these attempts resulted in enduring electoral vehicles (Stavenhagen, 1992, p. 434).4 Defying the expectations of the social cleavages’ approach to understanding party systems (Harmel & Robertson, 1985, p. 503; Horowitz, 1985, pp. 291-293; Lipset & Rokkan, 1967), deep ethnic cleavages in Latin America failed to produce an enduring set of corresponding political parties—that is, parties that survived and earned seats in more than two consecutive elections. The literature on indigenous movements and democratic political representation offers a number of competing explanations for the recent politicization of ethnic cleavages in contemporary Latin America. Yashar (1998) suggested that the combination of political liberalization, neoliberal economic reforms, and preexisting social networks served to trigger the politicization of indigenous identity in the region. Democratization created the political opportunity for indigenous peoples to mobilize in response to market reforms that had severed their corporatist ties 4. A number of indigenous electoral vehicles competed in Bolivia in the late 1970s and 1980s. None attracted more than 2% of the vote in national elections. See Van Cott (2005). Chilean Indians formed ephemeral vehicles that competed with some success in 1945, 1953, and 1989 (Albó, 1996). In Brazil, an Afro-Brazilian party competed in the 1930s (Htun, 2004, pp. 64-65).
714
Comparative Political Studies
to the state and undermined their material well-being. Brysk (2000), however, suggested that the process of globalization fostered the rise of indigenous rights movements in Latin America by enabling indigenous peoples to advance their cause by appealing to international norms, laws, and organizations. Albó (2002) and Bengoa (2000) argued that domestic and external factors produced the politicization of ethnic cleavages in the region, including urban migration, the failure of the developmentalist state, and liberal international norms. Although this growing body of scholarship offers a variety of explanatory variables for the politicization of ethnic identity, with a few notable exceptions (see Beck & Mijeski, 2001; Collins, 2001; Van Cott, 2003, 2005), it has largely failed to address how this new cleavage does, or does not, project into the electoral arena. This omission in the literature is curious given the unexpected emergence and successful performance of indigenous parties in the region since 1990. We define successful performance as the ability to win office at any level of government in consecutive elections. Although this is a modest definition of success, and it is too early to predict their long-term success or duration, compared to the earlier performance of indigenous parties in the few places they existed (mainly Bolivia and Chile), the experience of the 1990s and early 2000s is indeed a marked improvement. Successful indigenous parties first emerged in 1990 in Colombia, an unlikely case of indigenous party emergence, given the minuscule size (less than 3%) of its indigenous population. In that year, the Department of Cauca produced two indigenous electoral vehicles that gained seats in the 1991 National Constituent Assembly and performed unexpectedly well in the elections that followed. One of the parties continues to compete well at all levels of government. Another indigenous vehicle, the Indigenous Social Alliance (ASI), emerged in 1991. The ASI has consistently won elections at the local, regional, and national level and has established a presence in 11 of the country’s 32 departments. The surprising success of indigenous parties in Colombia inspired indigenous social movements in Bolivia and Ecuador to form new indigenous electoral vehicles in 1995 and 1996, respectively. By 2002, both were challenging their country’s largest parties for the presidency and had sent significant delegations to the national legislature.5 In Venezuela, the indigenous organization of the state of Amazonas formed a party (the United Multiethnic Party of Amazonas [PUAMA]) to contest 5. In Ecuador a coalition uniting the indigenous-movement-based party Pachakutik Movement of Pluricultural Unity and a new populist and/or personalist vehicle won the presidential election. In Bolivia, the Political Instrument for the Sovereignty of the Peoples, finished less than two percentage points behind the winner. See Van Cott (2005).
Rice, Van Cott / Indigenous Peoples’ Political Parties
715
municipal and statewide elections in 1997. It went on to elect a representative to the 1999 National Constituent Assembly. In 2000, PUAMA captured the governorship of Amazonas and sent a national representative to the legislature. In Guyana, indigenous organizations launched their first political party to compete in the 2001 national elections. They overcame attempts to disqualify their candidates and a successful move to reduce by one half the share of seats allotted to indigenous-majority districts, finishing third and capturing two seats in Parliament, while sweeping elections in indigenousmajority Amazonian districts (Joseph, 2001a, 2001b). In Peru and Chile, strong legacies of class-based partisan organizing by the Marxist Left continue to politically overshadow ethnic cleavages, thereby impeding the formation of indigenous-based parties. Existing indigenous movements are limited in scope, weak in outreach, and marginalized from political debates. Regional attempts at ethnic party formation in Peru have been unsuccessful. For instance, difficulties in collecting sufficient signatures to register the political party Indigenous Movement of the Peruvian Amazon (MIAP), forced party members to form alliances with registered parties to compete in elections as of 1998 rather than run under their own party label (Van Cott, 2005). In the 2002 regional and municipal elections in Peru, several new electoral vehicles emerged that utilize indigenous symbols and names, including MARQA in the department of Puno, the Inka Pachacuteq Party in the department of Cusco, and the Popular Llapanchik Front in the department of Apurímac. However, such parties reflect regional and class-based demands rather than an explicit ethno-cultural agenda (Meléndez, 2003). In Chile, the indigenous struggle is also a localized one. There is no overarching organization to bring together the country’s geographically dispersed indigenous peoples. The most significant attempt to form an indigenous-based party, which occurred in 1989 during the transition to democracy, failed. The Party for Land and Identity (PTI) was an attempt by indigenous leaders across Chile to end the role of nonindigenous intermediaries in the representation of indigenous interests before the state. However, the party’s founders were unable to complete the requirements for party registration as key members opted instead to run on the lists of the established parties (Marimán, 1993). Consequently, Peru and Chile serve as important examples of cases where the expected parties did not emerge. Social scientists have been keen to explain the emergence of these new indigenous political vehicles, given the extremely adverse socioeconomic, institutional, and political conditions under which most have emerged. Most scholars have focused on the most electorally successful cases—Bolivia and
716
Comparative Political Studies
Ecuador—primarily through monographic, qualitative analyses of either country, or comparisons of the two.6 As the number of countries with indigenous parties increased in the late 1990s, qualitative, comparative analyses of multiple cases were undertaken (Van Cott, 2005), and the topic began to attract attention from political scientists employing cross-national statistical methods. Most of these recent quantitative efforts are still in the conference paper, in press, or unpublished stage (e.g., Birnir, 2004, in press; Birnir & Van Cott, in press; Madrid, 2005a, 2005b). Our article contributes to this emerging area of research.
Research Design The authors independently conducted field research in six contiguous South American countries that share a variety of institutional, social, and political features: Bolivia, Chile, Colombia, Ecuador, Peru, and Venezuela.7 While controlling for region, the six countries vary significantly with respect to the size and level of political mobilization of their indigenous populations and with respect to the emergence and performance of indigenous political parties. As noted above, Bolivia and Ecuador have proportionally large indigenous populations (62.05% and 24.85%, respectively); vibrant, diverse, indigenous social movements; and nationally competitive indigenous parties that are contenders for the presidency. Colombia and Venezuela have minuscule indigenous populations (2.7% and 1.48%, respectively), regionally competitive indigenous movements, and indigenous parties with robust local and regional support and some national representation. In Chile and Peru, which are 7.06% and 38.39% indigenous, respectively, indigenous organizations have been less geographically extensive and effective, and their attempts to form competitive parties unsuccessful.8 Taken together, the 6. On Ecuador, see, for example, Beck & Mijeski, 2001; Collins, 2001; and Pallares, 2002. On Bolivia, see, for example, Ticona, Rojas, & Albó, 1995; Secretaría Nacional de Participación Popular, 1997; Van Cott, 2000, 2003. Doctoral dissertations by Andolina (1999) and Lucero (2002) compared parties in both countries. 7. Rice conducted field research in Chile in April-May 2004, in Bolivia in January-March 2004, in Peru in October-December 2003, and in Ecuador in July-September 2003. Van Cott conducted field research in Bolivia in June-July 2002, in Venezuela in May 2000, in Ecuador in July 1999, and in Colombia in February-March 1997. 8. Indigenous population data is from Deruyttere (1997, p. 1), except Bolivia (Censo Nacional de Población y Vivienda, 2001, available at ine.gov.bo/esn) and Colombia (1993 census) http://www.dane.gov.co/.
Rice, Van Cott / Indigenous Peoples’ Political Parties
717
six countries contain more than one half the indigenous population of Latin America and represent a range of relationships between indigenous peoples and the political system. The six countries also present diverse histories of party system development. Colombia and Venezuela have a long history of continuous civilian, elected rule and two-party dominance. Peru and Venezuela saw their party systems collapse in the 1990s. Ecuador, Bolivia, and Peru returned to elected, civilian rule in the late 1970s or early 1980s and have suffered from high and increasing party system fragmentation. Ecuador and Peru only enfranchised illiterates—affecting much of the indigenous population— during the democratic transition. Chile was the last to emerge from authoritarian rule and has been dominated by a center-left electoral alliance since that time. Thus, our selection of cases enables us to hold a variety of variables constant, while observing the impact of variations in key political, institutional, and demographic characteristics. Whereas a larger collection of cases would have advantages for cross-national statistical research, the small number of country cases included in the current study enables the authors to obtain a deeper and more nuanced understanding of the interaction between indigenous peoples and their political systems, knowledge that is particularly useful in the preliminary stages of theory construction and testing. This method also reduces the likelihood of dramatic measurement error on key variables such as indigenous population size—a serious problem in Latin American countries—because there is presumably less error in measurement on this and on socioeconomic data across subnational levels within countries than among countries (Jones, 1997, pp. 350-351; Snyder, 2001, p. 94). Of course, this method does not eliminate the possibility of measurement error—likely to remain significant with respect to indigenous population size—but it minimizes such error relative to large-N cross-national studies. Using subnational data enables us to increase our N to 248, despite the fact that we have only included six countries in the current study. We rejected a time-series analysis in favor of what we have termed a twin snapshot approach because of data constraints at the subnational level and the notion that many of our key variables change quite slowly (e.g., size of indigenous population, poverty levels, population density). Time-series analysis would have significantly underestimated the impact of these variables relative to political variables, which vary more widely from year to year. Moreover, the time period during which indigenous parties have competed in multiple countries is brief (1995-2002). Therefore, we collected data
718
Comparative Political Studies
from two important time periods: (a) the last election in the 1980s or the beginning of the 1990s, before most indigenous political parties first emerged and (b) the most recent elections, when indigenous parties reached their greatest level of formation and success thus far. Because these two “snapshots” are taken roughly a decade apart, they capture the maximum variation on socioeconomic variables, which change slowly, and demographic variables, which typically are measured each decade.9 We ran two separate models with two different dependent variables: a dichotomous variable measuring indigenous party emergence and existence,10 and a continuous variable measuring the combined vote share for indigenous parties. Because of the skewed distribution of cases of indigenous party formation, with only 58 cases of party formation and 190 cases in which no indigenous party emerged, the party formation variable was collapsed into a dichotomous variable to predict party emergence and existence in general as opposed to an event count variable, in which the number of party formations would be predicted. None of the cases had more than two indigenous parties in existence in a given year. Because of the dichotomous nature of the party emergence and existence variable, a logistic regression analysis was performed. The indigenous party share of the vote variable ranged from a value of 0% to 39.25% (for descriptive statistics see Appendix).11 The vote share was calculated using the combined 9. The country-election years included in our model are as follows: Bolivia, 1989 and 2002; Chile, 1993 and 2001; Colombia, 1994 and 2002; Ecuador, 1990 and 2002; Peru, 1990 and 2001; and Venezuela, 1988 and 2000. Electoral data was derived from the following: Bolivia, Corte Nacional Electoral (www.cne.org.bo) and the Political Database of the Americas (www.georgetown.edu/pdba); Chile, Servicio Electoral (www.servel.cl); Colombia, Lijphart Elections Archive (http://dodgson.ucsd.edu/lij/data/Colombia.col.data.html); Ecuador, Tribunal Supremo Electoral (www.tse.gov.ec); Peru, Oficina Nacional de Procesos Electorales (www. onpe.gob.pe); and Tuesta Soldevilla (2001); Venezuela, Asamblea Nacional de la República Bolivariana de Venezuela (www.asambleanacional.gov.ve). 10. If one or more indigenous parties competed in an election in a department/state/ province within a given year it was assigned a value of 1; if no indigenous party competed in that year it was assigned a value of 0. The indigenous parties included in the data set are: Bolivia, 1989 = Movimiento Revolucionario Túpac Amaru (MRTKL) and Frente de Unidad de Liberación Katarista (FULKA), 2002 = Movimiento al Socialismo (MAS) and Movimiento Indígena Pachakuti (MIP); Chile 1993 and 2001 = none; Colombia 1994 = Alianza Social Indígena (ASI), Autoridades Indígenas de Colombia (AICO), 2002 = ASI, AICO, Partido Indígena Colombiano; Ecuador 1990 = none, 2002 = Partido Sociedad Patriótica (PSP)/Movimiento de Unidad Plurinacional Pachakutik-Nuevo País (MUPP-NP) and Movimiento Indígena Amauta Jatari (MIAJ); Peru 1990 and 2001 = none; Venezuela 1988 = none, 2000 = Pueblo Unido Multiétnico de Amazonas (PUAMA). 11. The cases of Peru and Chile were dropped from the indigenous party electoral performance model because no indigenous parties were present.
Rice, Van Cott / Indigenous Peoples’ Political Parties
719
percentages of the parties’ votes in national legislative elections at the level of the department/state/province in each country. A standard ordinary least squares (OLS) regression analysis was performed on the indigenous party vote share variable.12 The independent variables of the current study include institutional, structural, and demographic variables. We tested three institutional variables. The first is district magnitude (DM), which scholars of both ethnic parties and Latin American party systems found to be decisive in determining the number of parties competing effectively in a political system. According to Birnir and Van Cott (in press), Cox (1997), Ordeshook and Shvetsova (1994), and Taagepera and Shugart (1989), the greater the average number of seats available in electoral districts, the higher the number of parties in the political system. Most of these studies focused on cases of established democracies in developed countries. Using data mainly from new democracies, Rosenblum and Huelshoff (2004) found the opposite effect: Multipartism is more likely to occur in smaller districts because of the distinct way that DM interacts with social cleavages in countries where political institutions are new and, thus, their effects on political outcomes less pronounced. In a study using data only from Latin America, Mainwaring and Shugart (1997) found DM to have no significant impact. These studies use mean or median measures of DM because their unit of analysis is the political system, rather than subnational districts. Our model, in contrast, tests for the actual effects of the magnitude of each district.13 Moreover, variations in DM enable us to partially capture the variable of population size in each district because this is typically a main determinant of DM. We expected that greater DMs will be associated with the emergence and enhanced electoral performance of indigenous peoples’ parties. Our second institutional variable is the creation of new electoral districts,14 which occurred in many Latin American countries during the decentralization drives of the 1990s. Several new districts were carved out of large rural and Amazonian territories and, thus, coincided with significant indigenous populations. For example, in Venezuela, the two most-indigenous 12. We opted to use the ordinary least squares (OLS) regression method, rather than weighted least squares (WLS), because autocorrelation did not pose a serious problem since we have only two observations per case and a relatively large number of cross sections. 13. District magnitude (DM) is measured by the absolute number of seats per subnational unit in the lower house of deputies. 14. A dummy variable was used to test the impact of new electoral districts on indigenous party emergence and performance.
720
Comparative Political Studies
states—Amazonas (49.71% indigenous) and Delta Amacuro (19.80% indigenous)—were created in 1988 out of former federal territories. New districts present particularly favorable ground for new political vehicles because they do not have to compete with incumbents, who might otherwise have a tight grip on political resources and established ties with voters. Therefore, it is expected that new electoral districts will be associated with a greater propensity to form indigenous parties and the enhanced electoral performance of such parties. The final institutional variable in our model is party system fragmentation.15 We measured the level of party system fragmentation using Laakso and Taagepera’s (1979) formula for effective number of parties for seats (ENPS), which is calculated by taking the proportion of seats won by each party, squaring each proportion, adding them together, and dividing one by that number. We expect that higher degrees of party system fragmentation will lower barriers to entry for new parties because fewer votes are needed to gain seats. This indicator is particularly important for the emergence of new party system cleavages (Dalton, Flanagan, & Beck, 1984, p. 466). Therefore, we expected that a greater effective number of parties for seats would be associated with the emergence and improved performance of indigenous peoples’ parties. We also tested for the effects of two structural variables. First, given the long relationship between indigenous voters and organizations and leftist parties16 in Latin America (Bengoa, 1999; Chiriboga & Rivera, 1989, p. 195; 15. Because of data constraints at the subnational level, a lagged effective number of parties for seats (ENPS) using electoral data from the previous election was not calculated, thereby presenting a minor endogeneity problem given that if an indigenous party formed in the year under investigation, it automatically increased the overall number of parties. However, our measure is largely determined by the configuration of nonindigenous parties and, thus, can determine whether indigenous party formation and success is influenced by the degree of party system fragmentation. 16. Leftist parties were identified using the Latin American political party classifications developed by Coppedge (1997a) and the authors’ familiarity with the cases. Parties and election years included in our analysis are: Bolivia, 2002: Socialist Party (PS), 1989: United Left (IU) and Socialist Party-One (PS-1); Chile, 2001: Communist Party of Chile (PCCH), 1993: Communist Party of Chile (PCCH), Unitary Popular Action Movement (MAPU); Colombia, 2002: Democratic Alliance/M-19 (ADM-19) and Revolutionary Independent Workers’ Movement (MOIR), 1994: Patriotic Union (UP) and Democratic Alliance/M-19 (ADM-19); Ecuador, 2002: Popular Democratic Movement (MPD), Socialist Party-Broad Front (PS-FA), 1990: Popular Democratic Movement (MPD), Socialist Party of Ecuador (PSE); Peru, 2001: none, 1990: United Left (IU) and Socialist Left (IS); Venezuela, 2000: Movement for the Fifth Republic (MVR), Movement Toward Socialism (MAS), Communist Party of Venezuela (PCV), Fatherland for Everyone (PPT), Bandera Roja-Red Flag, and Radical Cause (LCR); 1988: Movement Toward Socialism (MAS), Communist Party of Venezuela (PCV), The People’s Electoral Movement (MEP), New Alternative, Socialist League (LS), Radical Cause (LCR).
Rice, Van Cott / Indigenous Peoples’ Political Parties
721
Degregori, 1998; Ruiz Hernández & Burguete Cal y Mayor, 2001, p. 25; Van Cott, 2005), we expected that a higher vote share for political parties of the Left, which reflects the salience of class-based identities in society, would decrease the likelihood of indigenous party emergence and success. The disarticulation of traditional class-based collective action and the crisis of the political Left as a result of the shift toward neoliberal economic policies in Latin America have created a void in popular sector interest representation out of which new, identity-based political parties are emerging. In countries where leftist, class-based organizing predominates, autonomous patterns of popular-sector mobilization create class identities that may impede the articulation and mobilization of ethnic identities to the extent that where indigenous peoples mobilize politically, they do so as workers or as peasants and tend to vote for leftist rather than ethnic parties.17 The second structural variable included in our model is the level of poverty in each subnational unit.18 There is a strong and positive correlation between indigenous population size and levels of poverty (Psacharopoulos & Patrinos, 1994).19 Indigenous peoples are much more likely to live in conditions of extreme poverty than nonindigenous peoples. Much of the programmatic appeal of these newly emerged indigenous parties is based on their rejection of the neoliberal economic model and their emphasis on poverty alleviation.20 Consequently, indigenous peoples’ parties have broad 17. Our expectations are complicated by the fact that in several countries in our sample it is common for indigenous parties to run in coalition with leftist parties. In fact, this occurred regularly in Ecuador and Venezuela. Where indigenous parties run in coalition with leftist parties a large vote share for the Left may actually encourage the formation and successful performance of indigenous parties who are able to form alliances with them, whereas in districts where indigenous parties run on their own, competing against leftist parties, the opposite effect can be expected. 18. Poverty is measured here as the percentage of the population living in extreme poverty. Poverty data was drawn from the following sources: Bolivia, 2001 and 1992 national censuses, available at www.ine.gov.bo; Chile, 1998 and 1992, Encuesta de Caracterización Socioeconómica Nacional, available at www.mideplan.cl; Colombia, 2000, Human Development Report for Colombia, available at www.pnud.org.co/Informed/IDH_COL-2000.pdf, and 1998, Human Development Report for Colombia, available at www.pnud.org.co/Informes/IDH_COL-1998.pdf; Ecuador, 2001, Human Development Report for Ecuador, available at pnud.org.ec/Idh2001/ informe.php; Peru, 2001, Encuesta Nacional de Hogares and 1993 national census, both available at www.inei.gob.pe; Venezuela, 2002 national census, available at www.ocei.gov.ve, and 1996 Human Development Report for Venezuela, available at www.pnud.org.ve/Idh96. 19. A correlation matrix of the indigenous population and poverty variables in our subnational data set confirmed a positive correlation of .4 between the two variables. 20. See, for example, Web sites of Ecuador’s Pachakutik party at www.pachakutik.org.ec and Bolivia’s Movement toward Socialism at www.masbolivia.org.
722
Comparative Political Studies
appeal to Latin America’s poor, indigenous and nonindigenous alike. However, one study of the 1996 and 1998 Ecuadorian elections using parroquia [parish] level data found no, or a slightly negative, relationship between levels of poverty and the likelihood of voting for an indigenous party (Beck & Mijeski, 2004). We believe this finding may reflect the use of different poverty measures, or perhaps differences between parroquiaand provincial-level data or politics, or factors peculiar to the Ecuadorian case. Therefore, we expected that indigenous peoples’ parties will emerge and perform better in subnational units that have higher rates of poverty. The demographic variables included in our model are the population density of each subnational unit and the percentage of the population in each unit that is indigenous. The population density21 variable enables us to control for the degree to which our district is rural or urban and to partially account for the population size of each district, with low density typically corresponding to a smaller population. In fact, we omitted district population size as a variable because it should correlate strongly with DM and population density. In Latin America, conservatives have used their influence to ensure that rural districts, traditionally dominated by conservative elites, are relatively overrepresented in national legislatures. Where well-organized indigenous movements are able to upset traditional relations of dominance, they can use malapportionment to their own advantage. For example, in Amazonas, Venezuela, in the 2000 national legislative elections the indigenous party PUAMA elected a nominal deputy with only .04% of the national vote (Van Cott, 2005). Therefore, it is hypothesized that lower population densities, indicating that a district is rural and may potentially possess more organized, indigenous supporters, will be associated with a greater possibility of indigenous party emergence and enhanced electoral performance. Finally, we included in our model a variable on indigenous population size.22 We expected that the more indigenous voters in the district, the greater the likelihood that indigenous peoples would form their own electoral vehicles. Following the work of Chandra (2004), it is also expected that 21. Measured as the number of inhabitants per km2. Recent data drawn from the national censuses of each country; earlier years from Europa World Year Book, (Europa Publications, 1994). 22. Indigenous population data were drawn from the following sources: Bolivia, 2001 and 1992 national census, available at www.ine.gov.bo; Chile, 2002 and 1992 national census, available at www.ine.cl; Colombia, 1993 national census, available at www.dane.gov.co; Ecuador, 2001 national census, available at www.inec.gov.ec; Peru, 1993 national census, available at www.inei.gob.pe; Venezuela, 1992 indigenous census, available at www.pueblosindigenas.org.ve.
Rice, Van Cott / Indigenous Peoples’ Political Parties
723
ethnic parties are more likely to succeed when they target votes from their own ethnic category. It is important to remember, however, that sheer numbers alone are not decisive. A variable that is perhaps more important, although difficult to measure and insert into a quantitative analysis, is the maturity and degree of unity, organization, and mobilization of indigenous peoples’ social movement organizations within the district. In Colombia, for example, indigenous political parties have performed better in Cauca, which is less than 15% indigenous, than in Vaupés and Guainía, with 90.5% and 98.7%, respectively. Logically, we can expect that the higher the degree of maturity, unity, organization, and mobilization of indigenous peoples’ social movement organizations in a particular electoral district, the more likely that an indigenous organization will (a) seek to participate in elections, encouraged by its ties of loyalty to a large number of voters in the district; (b) will have the organizational resources to surmount registration barriers (e.g., the ability to gather sufficient signatures); and (c) will be able to convert its social movement ties into votes on election day. This notion is supported by a six-country cross-national study undertaken by one of the authors, which found that all indigenous social movements forming political parties were founded on average 14 years before spawning a party (Van Cott, 2005). Given the lack of financial resources available to indigenous movements, they must rely on organizational resources to inform voters of their programs and candidates, provide logistical support for campaigns, and get their voters to the polls. Should a scholar develop in the future a reliable method of measuring this complex qualitative variable, it should not be correlated significantly with the variables we have already mentioned, and we can expect that its importance lies somewhere in the unexplained portion of our model and in the decade dummy described below. Our model included a key control variable:23 a decade dummy to control for the decade in which the phenomenon under investigation occurred. The decade control variable accounts for changes in the political, economic, and social conjuncture in South America during the 1990s, while also capturing the “demonstration effect” produced by the prior formation and the ability to win office of indigenous parties in individual districts and countries, as well as cross-border diffusion within and among countries.24 The decade control may also potentially capture important unmeasured factors, such as 23. Country dummy variables were also used to control for fixed unit effects. 24. A value of 1 was assigned to the data of the second snapshot period of the early 2000s, and a value of 0 was assigned to data of the previous snapshot period of the late 1980s or early 1990s.
724
Comparative Political Studies
Table 1 Logistic Regression on Indigenous Peoples’ Party Emergence and Existence by Institutional, Structural, and Demographic Variables Variables
Indigenous Party Emergence
District magnitude ENPS New electoral district Leftist share of vote Poverty Population density Indigenous population Decade dummy Constant Pseudo R2 Overall % predicted correctly N
.236* (.094) –.163 (.274) 1.226 (1.426) –.022 (.031) .047 (.028) .000 (.001) .067** (.020) 3.432*** (.848) –5.805 (1.159) .792 94.8 248
Note: ENPS = effective number of parties for seats. Entries are unstandardized logistic regression coefficients with standard errors in parentheses. *p < .05. **p ≤ .001. ***p ≤ .000. Two-tailed tests.
the maturity and growth of indigenous movement organizations. Anecdotal evidence suggests that indigenous political movements often look to their cohorts in neighboring countries for lessons learned and inspiration. Therefore, it is hypothesized that the decade control variable will be positively associated with the emergence and electoral strength of indigenous peoples’ parties in the most recent round of elections as a result of ideological diffusion and political learning within and across countries as well as the organizational maturity of the region’s indigenous movements.
Results The results of our analysis offer significant support for the hypothesis that political institutions, structural conditions, and demographic factors determine the emergence and electoral performance of indigenous peoples’ parties. Permissive electoral rules, large indigenous population size, and the decade control dummy were found to encourage the emergence of indigenous political parties (see Table 1). All three variables were significant in predicting indigenous party emergence and existence and in the hypothesized
Rice, Van Cott / Indigenous Peoples’ Political Parties
725
Table 2 Ordinary Least Squares Regression on Electoral Performance of Indigenous Peoples’ Parties by Institutional, Structural, and Demographic Variables Variables District magnitude ENPS New electoral district Leftist share of vote Poverty Population density Indigenous population Decade dummy Constant Adjusted R2 N
Indigenous Party Share of Vote .298* (.105 –.663 (.458) –.520 (1.888) –.111** (.030) .108* (.036) .000 (.001) .098** (.022) 6.594** (.882) –2.800 (1.875) .523 172
Note: ENPS = effective number of parties for seats. Entries are unstandardized regression coefficients with standard errors in parentheses. *p ≤ .01. **p ≤ .000. Two-tailed tests.
directions. Greater DM was found to be associated with the increased likelihood of indigenous party emergence. Indigenous peoples’ parties were also found to be more likely to emerge in subnational districts with higher indigenous populations. Last, the decade control variable was found to be positive and significant in predicting the emergence of indigenous parties. This finding suggests that independent of all the variables included in the model, indigenous peoples’ parties are just more likely to emerge in the current time period, presumably because of a demonstration effect produced by the prior formation and electoral competitiveness of indigenous parties in the region. The discovery that the emergence of indigenous peoples’ parties is the product of indigenous population size, favorable institutional conditions, and timing suggests that indigenous party formation in the future will occur mainly in districts where indigenous populations and DM are relatively large. Our second model, which tests for factors that determine the electoral strength of indigenous peoples’ parties, offers a different picture. In this model, structural conditions were found to be significant predictors of indigenous-party electoral performance, in addition to favorable underlying political conditions (see Table 2). The vote share of leftist parties and poverty rates, alongside DM, indigenous population, and the decade control
726
Comparative Political Studies
variable, were all significant in predicting the superior electoral performance of indigenous peoples’ parties and in the hypothesized directions. A lower vote share for political parties of the Left, representing less salient classbased identities, was found to successfully predict higher vote shares for indigenous parties. Higher rates of poverty also were found to be associated with the enhanced electoral performance of indigenous peoples’ parties. These findings indicate that although favorable institutional and demographic conditions may result in the emergence of indigenous parties, it is largely structural and political conditions that determine the electoral strength and, thus, the longevity of these new political parties. Consequently, if indigenous peoples’ parties are to continue to thrive in Latin America, they will need to take economic conditions and relations with leftist parties into account, particularly when winning elections depends on attracting nonindigenous voters. The variables that did not yield the predicted results are also of some theoretical interest.25 Of the institutional variables included in our models, neither the creation of new electoral districts nor the degree of party system fragmentation at the district level were significant in predicting indigenous party emergence or electoral performance. The failure of the new electoral district dummy to yield results may be explained by the sparsity of such districts relative to the number of established districts in our data set, with only 10 new electoral districts of a total of 248 districts. Perhaps most surprising, the effective number of parties for seats variable was not found to be a significant predictor of the emergence or performance of indigenous parties. It may be that the party systems in the current study are all fairly highly fragmented, such that the difference among them, during the elections when our “snapshots” were taken, was not sufficient to affect party emergence and performance. Of the demographic variables included in our models, only population density proved to be insignificant in the emergence and electoral performance of indigenous peoples’ parties. This finding may be explained by the conditionality of the hypothesis that lower population densities, indicating rural districts, may be associated with the emergence of and strength of the vote for indigenous peoples’ parties only if well-organized 25. Note: the country dummy variables were not found to be significant in the party emergence model but yielded mixed results in the electoral performance model, with only the dummies for Bolivia, Ecuador, and Peru significant, indicating that there is an unmodeled factor that makes an indigenous party more likely to have greater electoral success in the first two but not the last.
Rice, Van Cott / Indigenous Peoples’ Political Parties
727
indigenous movements can upset traditional relations of dominance to their own electoral advantage. Overall, the findings of our models shed considerable new light on the factors that influence, and those that fail to influence, new party formation and performance.
Conclusion The factors giving rise to ethnic parties in Latin America, and those helping to determine their electoral performance, are more varied and complex than the social cleavages and institutionalist literatures suggest. Although the size of the ethnic population in question is, indeed, a significant factor, in Latin America, given the overwhelming poverty of this population and its long-standing preference for leftist parties that appeal to its class interests, we must also take into account socioeconomic indicators and the performance of existing leftist parties in each district. Of the institutional variables that were available and appropriate for our research design, only DM appears to have the predicted effect on new party formation. Perhaps most important from a theoretical standpoint, the significance of the decade control variable included in our models indicates that something very different is going on in Latin America at the beginning of the new millennium. Perhaps political learning and diffusion, as well as the organizational maturation of indigenous movements, play a key role in the emergence and performance of ethnic parties independent of underlying structural or institutional conditions. In addition, it reflects significant changes in political conditions. The most recent round of elections occurred following more than two decades of intense indigenous political mobilization, the dealignment and increasing fragmentation of political party systems throughout the region, and the codification of ethnic rights in most of the region’s constitutions. Thus, the cohesion and strength of indigenous peoples as collective political actors experienced a significant increase in most countries during the past decade, whereas traditional political actors—particularly political parties and labor movements—suffered a notable decline. We hope that our findings will inspire and complement qualitative studies of particular indigenous parties, electoral districts, and country case studies to produce a clearer picture of the complex variables that help to shape political party systems so that we can better understand one of the most interesting political phenomena to emerge in Latin America in the past decade: the politicization and particization of indigenous ethnic identity.
728
Comparative Political Studies
Qualitative studies will help to flesh out important but difficult-to-measure variables such as the relative cohesion and strength of indigenous identity (and of subgroups within the indigenous category, e.g., Aymara vs. Quechua); the maturity and consolidation of indigenous social movements; the availability of talented leadership; and the complex and ever-changing electoral alliances that indigenous-movement-based parties enter into with leftist and populist parties. In their other work, the authors are wrestling with these questions. Future research should also incorporate more sophisticated measures of ethnic identity that capture ethnic diversity within the indigenous category, while also controlling for the geographic concentration and/or dispersion of indigenous voters within particular districts. Should indigenous parties continue to flourish and expand to more countries, the prospects for statistical research should improve as the data set on this nascent phenomenon expands.
Appendix Descriptive Statistics, N = 248 Variables Indigenous party Emergence Indigenous party Share of vote District magnitude ENPS New electoral district Leftist share of vote Poverty Population density Indigenous population Decade dummy
Minimum Value
Maximum Value
M
SD
.00
1.00
.2339
.4242
.00
39.25
2.1174
6.3545
1.00 1.00 .00 .00 3.00 .20 .00 .00
49.00 10.78 1.00 74.56 91.50 5264.00 98.70 1.00
6.5956 2.6279 .0403 12.4398 35.6685 123.3327 16.1361 .5040
6.4111 1.2258 .1971 16.9879 23.5745 505.2668 23.6335 .501
Note: ENPS = effective number of parties for seats.
References Albó, Xavier. (1996). Andean people in the twentieth century. In Frank Salomon & Stuart B. Schwartz (Eds.), The Cambridge history of the native peoples of the Americas. Vol. III, South America, part 2 (pp. 765-871). Cambridge, UK: Cambridge University Press. Albó, Xavier. (2002). Pueblos Indios en la Política [Indian peoples in politics]. La Paz, Bolivia: Centro de Investigación y Promoción del Campesinado.
Rice, Van Cott / Indigenous Peoples’ Political Parties
729
Andolina, Robert. (1999). Colonial legacies and plurinational imaginaries: Indigenous movement politics in Ecuador and Bolivia. Unpublished doctoral dissertation, University of Minnesota, Minneapolis. Beck, Scott H., & Mijeski, Kenneth J. (2001). Barricades and ballots: Ecuador’s Indians and the Pachakutik political movement. Ecuadorian Studies, 1(Sept.). Available at www. ecuatorianistas.org/journal/1/1.html. Beck, Scott H., & Mijeski, Kenneth J. (2004). Ecuador’s Indians in the 1996 and 1998 elections: Assessing Pachakutik’s performance. Latin Americanist, 67(3/4), 46-74. Bengoa, José. (1999). Historia de Un Conflicto: El Estado y los Mapuches en el Siglo XX [History of a conflict: The state and the Mapuches in the 20th century]. Santiago, Chile: Editorial Planeta. Bengoa, José. (2000). La Emergencia Indígena en América Latina [The indigenous emergence in Latin America]. Santiago, Chile: Fondo de Cultura Económica. Birnir, Jóhanna Kristín. (in press). Ethnicity and electoral politics. New York: Cambridge University Press. Birnir, Jóhanna Kristín. (2004). In press. Disunity and diversity: Party system fragmentation and the dynamic effect of ethnic heterogeneity in Latin America. Latin American Research Review. Birnir, Jóhanna Kristin, & Van Cott, Donna Lee. (2004, October). Disunity in diversity: Party system fragmentation and ethnic heterogeneity in Latin America. Paper presented at the international meeting of the Latin American Studies Association, Las Vegas, NV. Brysk, Alison. (2000). From tribal village to global village: Indian rights and international relations in Latin America. Stanford, CA: Stanford University Press. Chandra, Kanchan. (2004). Why ethnic parties succeed: Patronage and ethnic head counts in India. New York: Cambridge University Press. Chhibber, Pradeep. (1999). Democracy without associations. Ann Arbor: University of Michigan Press. Chiriboga, Manuel, & Rivera, Fredy. (1989). Elecciones de enero 1988 y participación indígena [January 1988 elections and indigenous participation]. Ecuador Debate, 17, 181-221. Collins, Jennifer N. (2001, September). Opening up electoral politics: Political crisis and the rise of Pachakutik. Paper presented at international meeting of the Latin American Studies Association, Washington, DC. Coppedge, Michael. (1997a). A classification of Latin American political parties (Kellogg Institute Working Paper No. 244). Notre Dame, IN: Notre Dame University, Kellogg Institute. Coppedge, Michael. (1997b). District magnitude, economic performance, and party-system fragmentation in five Latin American countries. Comparative Political Studies, 30(2), 156-185. Cox, Gary W. (1997). Making votes count: Strategic coordination in the world’s electoral systems. Cambridge, UK: Cambridge University Press. Dalton, Russell L., Flanagan, Scott C., & Beck, Paul Allen. (Eds.). (1984). Electoral change in advanced industrial democracies: Realignment or dealignment? Princeton, NJ: Princeton University Press. Degregori, Carlos Iván. (Ed.). (1998). Comunidades: Tierra, Instituciones, Identidad [Communities: Land, institutions, identity]. Lima, Peru: Diakonía-CEPES-Arariwa. Deruyttere, Anne. (1997). Indigenous peoples and sustainable development. The role of the Inter-American Development Bank. Washington, DC: Inter-American Development Bank. Europa Publications Limited. (1994). Europa world year book (Vols. 1 & 2). London: Europa Publications. Harmel, Robert, & Robertson, John D. (1985). Formation and success of new parties: A crossnational analysis. International Political Science Review, 6(4), 501-523.
730
Comparative Political Studies
Horowitz, Donald. (1985). Ethnic groups in conflict. Berkeley: University of California Press. Htun, Mala. (2004). From “racial democracy” to affirmative action: Changing state policy on race in Brazil. Latin American Research Review, 39(1), 60-89. Hug, Simon. (2001). Altering party systems: Strategic behavior and the emergence of new political parties in Western democracies. Ann Arbor: University of Michigan Press. Jones, Mark P. (1997). Racial heterogeneity and the effective number of candidates in majority runoff elections: Evidence from Louisiana. Electoral Studies, 16(3), 349-358. Joseph, Adrian. (2001a). Indigenous party meets opposition. Latinamerica Press, 33(5), 4-5. Joseph, Adrian. (2001b). Racial divisions mar elections. Latinamerica Press, 33(11), 3. Kitschelt, Herbert. (1989). The logics of party formation: Ecological politics in Belgium and West Germany. Ithaca, NY: Cornell University Press. Laakso, Murkuu, & Taagepera, Rein. (1979). Effective number of parties. A measure with application to western Europe. Comparative Political Studies, 12, 3-27. Levitsky, Steven. (1999). Fujimori and post-party politics in Peru. Journal of Democracy, 10(3), 78-92. Levitsky, Steven, & Cameron, Maxwell A. (2001, September). Democracy without parties? Political parties and regime collapse in Fujimori’s Peru. Paper presented at the international meeting of the Latin American Studies Association, Washington, DC. Lipset, Seymour Martin, & Rokkan, Stein. (Eds.). (1967). International yearbook of political behavior research (vol.7). New York: Free Press. Lucero, José Antonio. (2002). Arts of unification: Political representation and indigenous movements in Bolivia and Ecuador. Unpublished doctoral dissertation, Princeton University, Department of Politics, Princeton, NJ. Madrid, Raúl. (2005a). Ethnic cleavages and electoral volatility in Latin America. Comparative Politics, 38(1), 1-20. Madrid, Raúl. (2005b). Indigenous voters and party system fragmentation in Latin America. Electoral Studies, 24(4), 689-707. Mainwaring, Scott. (1999). Rethinking party systems in the third wave of democratization: The case of Brazil. Stanford, CA: Stanford University Press Mainwaring, Scott, & Scully, Timothy. (Eds.). (1995). Building democratic institutions: Party systems in Latin America. Stanford, CA: Stanford University Press. Mainwaring, Scott, & Shugart, Matthew. (1997). Presidentialism and democracy in Latin America. New York: Cambridge University Press. Marimán, José. (1993). Movimiento Mapuche: Transición Democrática en Chile (1989-1993): ¿Nuevo Ciclo Reivindicativo Mapuche? [Mapuche movement: Democratic transition in Chile (1989-1993): New Mapuche claim cycle?] Unpublished doctoral dissertation, Universidad de la Frontera, Temuco, Chile. Meléndez Guerrero, Carlos. (2003). Ultimo Mapa Político: Análisis de los Resultados de Las Elecciones Regionales de Noviembre del 2002 [Latest political map: Analysis of the results of the regional elections of November 2002]. Lima, Peru: IEP. Mozaffar, Shaheen, Scarritt, James R., & Galaich, Glen. (2003). Electoral institutions, ethnopolitical cleavages and party systems in Africa’s emerging democracies. American Political Science Review, 97(3), 379-390. Ordeshook, Peter C., & Shvetsova, Olga. (1994). Ethnic heterogeneity, district magnitude, and the number of parties. American Journal of Political Science, 38(1), 100-123. Pallares, Amalia. (2002). From peasant struggles to Indian resistance. The Ecuadorian Andes in the late twentieth century. Norman: University of Oklahoma Press.
Rice, Van Cott / Indigenous Peoples’ Political Parties
731
Psacharopoulos, George, & Patrinos, Harry Anthony. (1994). Indigenous people and poverty in Latin America: An empirical analysis. Washington, DC: World Bank. Roberts, Kenneth M. (1998). Deepening democracy? The modern left and social movements in Chile and Peru. Stanford, CA: Stanford University Press. Roberts, Kenneth M. (2001, September). Political cleavages, party-society linkages, and the transformation of political representation in Latin America. Paper presented at the international meeting of the Latin American Studies Association, Washington, DC. Roberts, Kenneth M. (2002a). Party-society linkages and democratic representation in Latin America. Canadian Journal of Latin American and Caribbean Studies, 27(53), 9-34. Roberts, Kenneth M. (2002b). Social inequalities without class cleavages in Latin America’s neoliberal era. Studies in Comparative International Development, 36(4), 3-33. Roberts, Kenneth M., & Wibbels, Erik. (1999). Party systems and electoral volatility in Latin America: A test of economic, institutional, and structural explanations. American Political Science Review, 93(3), 575-590. Rosenblum, Marc R., & Huelshoff, Michael. (2004). Not whether but when: Institutions, structure, and political parties. Unpublished manuscript, University of New Orleans. Ruiz, H., Margarito, & Burguete Cal y Mayor, Aracely. (2001). Indigenous peoples without political parties: The dilemmas of indigenous representation in Latin America. In Kathrin Wessendorf (Ed.), Challenging politics: Indigenous peoples’ experiences with political parties and elections (pp. 20-63). Copenhagen, Denmark: International Work Group for Indigenous Affairs. Schattschneider, E. E. (1960). The semisovereign people. New York: Holt, Rinehart & Winston. Secretaría Nacional de Participación Popular. (1997). Indígenas en el poder local [Indians in local power]. La Paz, Bolivia: Ministerio de Desarrollo Humano. Secretaría Técnica del Frente Social. (2001). Sistema Integrado de Indicadores Sociales del Ecuador (SIISE) [CD-ROM]. Quito, Ecuador: SIISE. Snyder, Richard. (2001). Scaling down: The subnational comparative method. Studies in Comparative International Development, 36(1), 93-110. Stavenhagen, Rodolfo. (1992). Challenging the nation-state in Latin America. Journal of International Affairs, 34(2), 421-440. Taagepera, Rein, & Shugart, Matthew Soberg. (1989). Seats and votes: The effects and determinants of electoral systems. New Haven, CT: Yale University Press. Tanaka, Martin. (1998). Los espejismos de la democracia: El colapso del sistema de partidos en el Perú, 1980-1995. Lima, Peru: Instituto de Estudios Peruanos. Ticona Alejo, Esteban, Rojas, Gonzalo, & Albó, Xavier. (1995). Votos y Wiphalas: Campesinos y Pueblos Originarios en Democracia [Votes and Wiphalas: Peasants and original peoples in democracy]. La Paz, Bolivia: CIPCA. Tuesta Soldevilla, Fernando. (2001). Perú Político en Cifras: 1821-2001 [Political Peru in figures: 1821-2001]. Lima, Peru: Fundación Friedrich Ebert. Van Cott, Donna Lee. (2000). Party system development and indigenous populations in Latin America: The Bolivian case. Party Politics, 6(2), 155-174. Van Cott, Donna Lee. (2003). From exclusion to inclusion: Bolivia’s 2002 elections. Journal of Latin American Studies, 35(4), 751-776. Van Cott, Donna Lee. (2005). From movements to parties in Latin America: The evolution of ethnic politics. New York: Cambridge University Press. Yashar, Deborah J. (1998). Contesting citizenship: Indigenous movements and democracy in Latin America. Comparative Politics, 31(1), 23-42.
732
Comparative Political Studies
Roberta Rice is a doctoral candidate in political science at the University of New Mexico. She is completing a dissertation titled From Peasants to Politicians: The Politicization of Ethnic Cleavages in Latin America, which examines how institutional constraints and incentives in conjunction with different patterns of incorporation of the peasantry shaped the prospects for the emergence of indigenous political movements in Latin America. Donna Lee Van Cott is an assistant professor of political science at Tulane University. She is author of From Movements to Parties in Latin America: The Evolution of Ethnic Politics (2005) and The Friendly Liquidation of the Past: The Politics of Diversity in Latin America (2000). Her new book project is titled, “Radical Democracy in the Andes: Ethnic Parties and the Quality of Democracy.”
