The Rise of Ethnopopulism in Latin America: Raúl L. Madrid

The Rise of Ethnopopulism in Latin America
Raúl L. Madrid
World Politics, Volume 60, Number 3, April 2008, pp. 475-508 (Article)
Published by The Johns Hopkins University Press
For additional information about this article

http://muse.jhu.edu/journals/wp/summary/v060/60.3.madrid.html
Access Provided by University of Notre Dame at 04/06/11 1:44PM GMT

The Rise of Ethnopopulism in
Latin America
By Raúl L. Madrid*
L ATIN America had long been the one region in the world without
major ethnic parties. In recent years, however, important parties
that are based to varying degrees in the indigenous population have
emerged in the region. The most successful of these movements have
been ethnopopulist parties, inclusive ethnically based parties that adopt
classical populist electoral strategies.1 Whereas exclusionary ethnic par-
ties have registered little electoral success, ethnopopulist parties have
won significant legislative or presidential victories in the Andean na-
tions. In Bolivia, Evo Morales and the Movimiento al Socialismo (mas)
won a resounding victory in the 2005 presidential elections, after com-
ing in second in the 2002 elections. In Ecuador another ethnopopu-
list party, the Movimiento Unidad Plurinacional Pachakutik (mupp),
has maintained a significant presence in the legislature since 1996 and
helped elect Lucio Gutiérrez president in 2002 and Rafael Correa in
2006.
What accounts for the success of these new ethnically based parties
in Latin America?2 Why have they been more successful than tradi-
tional ethnic parties? And why have they combined populist and ethnic
appeals?
Understanding the causes of the success of these new ethnically
based parties is important from a practical standpoint because these
movements are already having important effects. The election of Evo
* The author would like to thank Daniel Brinks, Jason Brownlee, Henry Dietz, Jorge Domínguez,
Ken Greene, Merilee Grindle, Austin Hart, Juliet Hooker, Wendy Hunter, Steve Levitsky, Tse-Min
Lin, Scott Mainwaring, Robert Moser, Kurt Weyland, the students in his graduate seminar on Latin
American politics, and the three anonymous referees for very helpful comments on earlier versions of
this article. The Teresa Lozano-Long Institute for Latin American Studies at the University of Texas
at Austin provided funding for the field research that made this article possible.
1
I define an inclusive party as one that recruits members of various ethnic groups for the top
leadership positions of the party, forms alliances with organizations that represent a diversity of ethnic
groups, eschews exclusionary rhetoric, and emphasizes that it seeks to represent all members of the
nation.
2
In social science parlance, the dependent variable of this study is the performance of ethnically
based parties in Latin America.
World Politics 60 (April 2008), 475–508

476 w o r l d p o li t i c s
Morales in Bolivia, for example, has led to state intervention in the

country’s natural gas industry, the passage of land reform, the expansion
of education in indigenous languages, and the convening of a constitu-
ent assembly to redesign the country’s political institutions. Similarly
dramatic changes may well occur if ethnopopulist leaders take power
elsewhere in the region.
The rise of the new ethnically based parties is puzzling from a theo-
retical perspective because it contradicts some of the expectations of
the scholarly literatures on populism and ethnic parties. Whereas the
literature on ethnic parties would not expect such parties to make eth-
nically inclusive appeals, the literature on populism would not expect
populist parties to make ethnic appeals at all. Thus, neither the schol-
arly literature on ethnic parties nor the literature on populism can eas-
ily account for the emergence and success of ethnopopulist parties.
Another scholarly literature, the indigenous politics literature, has
generated important insights into why powerful indigenous movements
have arisen in some Latin American countries in recent years, but this
literature has not typically sought to explain why some of these move-
ments have transformed themselves into parties or achieved electoral
success.3 One exception is Van Cott, who argues that institutional re-
forms, among other factors, helped foster the rise of indigenous-based
parties in Latin America.4 As we shall see, however, institutional re-
forms did not play an important role in the rise of the most important
ethnopopulist party to date, although some of the other variables that
Van Cott and others have discussed, especially the strength of the in-
3
Deborah J. Yashar, Contesting Citizenship in Latin America: Indigenous Movements and the Post-
liberal Challenge (New York: Cambridge University Press, 2005); José Antonio Lucero, “Arts of Uni-
fication: Political Representation and Indigenous Movements in Bolivia and Ecuador” (Ph.D. diss.,
Princeton University, 2002); Alison Brysk, From Tribal Village to Global Village: Indian Rights and
International Relations in Latin America (Stanford, Calif.: Stanford University Press, 2000); Claudia
Dary, ed., La Construcción de la Nación y la Representación Ciudadana en México, Guatemala, Perú, Ec-
uador y Bolivia [The Construction of the Nation and Citizen Representation in Mexico, Guatemala,
Peru, Ecuador, and Bolivia] (Guatemala City: flacso, 1998).
4
See Donna Lee Van Cott, “Institutional Change and Ethnic Parties in South America.” Latin
American Politics and Society 45 (Summer 2003); and Donna Lee Van Cott, From Movements to Parties
in Latin America: The Evolution of Ethnic Politics. New York: Cambridge University Press, 2005); for
other institutional explanations for the rise of indigenous-based parties in Latin America, see Jóhanna
Kristín Birnir, “Party System Stabilization in New Democracies: The Effect of Ethnic Heterogene-
ity on the Volatility of Electoral Preferences” (Ph.D. diss., University of California, Los Angeles,
2001); Jóhanna Kristín Birnir “Stabilizing Party Systems and Excluding Segments of Society,” Stud-
ies in Comparative International Development 39 (September 2004); Jennifer N. Collins, “Democra-
tizing Formal Politics: Indigenous and Social Movement Political Parties in Ecuador and Bolivia,
1978–2000” (Ph.D. diss., University of California at San Diego, 2006); and Roberta Lynne Rice,
“From Peasant to Politicians: The Politicization of Ethnic Cleavages in Latin America” (Ph.D. diss.,
University of New Mexico, 2006).
r i s e o f e t h n o p o p u li s m 477
digenous movement, have played a role in the success of the mas and
other ethnopopulist parties.5
This study argues that ethnopopulist parties have succeeded in
Latin America (and traditional ethnic parties have failed) in large part
because of the nature of ethnicity and ethnic relations in the region.
Specifically, the low levels of ethnic polarization and the ambigu-
ity and fluidity of ethnic identification in the region have meant that
indigenous-based parties can win votes not only from self-identified
indigenous people but also from people from other ethnic categories
who share some identification with indigenous cultures or who sup-
port the parties based on their positions on other issues. To win the
support of people from other ethnic categories, ethnopopulist parties
have avoided exclusionary rhetoric, reached out to members of differ-
ent ethnic groups, and employed traditional populist appeals.
This study differs from most previous studies of indigenous politics
in Latin America not only in the arguments it develops but also in the
data and methods it employs. Previous studies of indigenous politics
in Latin America have typically used qualitative methods and have re-
lied mostly on elite interviews, secondary accounts, and, in some cases,
highly aggregated electoral data to support their arguments.6 This study
makes use of all of those sources of data, but also employs individual-
level survey data and provincial-level census and electoral data to test
the claims being made.
The article is divided into six main sections. The first section sets
forth an explanation for the appeal of ethnopopulism in the region and
discusses why existing theories of ethnic parties and populism cannot
explain the rise of ethnopopulist parties. The second section examines
existing explanations for the rise of Bolivia’s mas. The third section
discusses how the mas’s inclusive ethnic appeal made it possible for the
party to win votes across a range of different ethnic groups in Bolivia.
The fourth section explores how the mas also used populist strategies
to earn votes. The fifth section tests some of the previous arguments
5
See Van Cott (fn. 4, 2005); Robert Andolina, “Colonial Legacies and Plurinational Imaginaries:
Indigenous Movement Politics in Ecuador and Bolivia” (Ph.D. diss., University of Minnesota, 1998);
and Patricia Marenghi and Manuel Alcántara, “Los Partidos Étnicos de América del Sur: Algunos
Factores que Explican Su Rendimiento,” in Salvador Martí i Puig, ed., Pueblos Indígenas y Política en
América Latina (Barcelona: Bellaterra-cidob, 2007).
6
One important exception is Mijeski and Beck’s work on indigenous voting in Ecuador. See, for
example, Kenneth J. Mijeski and Scott H. Beck, “Ecuador’s Indians in the 1996 and 1998 Elections:
Assessing Pachakutik’s Performance,” Latin Americanist 3 (Spring 2003); and Scott H. Beck and Ken-
neth J. Mijeski, “Did Ecuador’s Indians Elect the President in 2002?” (Paper presented at the annual
meeting of the South Eastern Council of Latin American Studies, Santo Domingo, D.R., March 4–6,
2004).
with a multinomial logit analysis of public opinion data on the 2005

Bolivian elections. The sixth section accounts for why ethnopopulist
parties succeed in some countries and elections but not in others, and it
distinguishes the ethnopopulist parties from other populist and leftist
parties that have achieved important electoral victories in recent years.
The conclusion discusses the implications of these arguments for theo-
ries of populism and for the literatures on ethnic parties and voting.
Explaining the Rise of Ethnopopulist Parties

The literature on populism would not predict the rise of ethnopopulist
parties. Populist movements in Latin America have not traditionally
had an important ethnic component, and Weyland goes so far as to
suggest that ethnic appeals might not marry well with populism be-
cause populists tend to appeal to “undifferentiated ‘people.’ ”7 Studies
of populism have typically stressed how populist movements emerge
in the cities and rely on the urban working class as their main base of
support, although they usually incorporate other classes and sectors as
well.8 Some of the literature on populism has also located it within a
particular historical epoch in Latin America, namely, the 1930s through
the 1960s, when socioeconomic modernization, import-substitution
industrialization, and political liberalization made feasible the con-
struction of populist coalitions.9
The literature on ethnic parties cannot easily account for the rise of
ethnopopulist parties either. This literature would not expect leaders of
ethnic parties to be inclusive—indeed, much of the literature on ethnic
parties defines them as parties that draw their support from a single
ethnic group.10 Nor would this literature expect such inclusive appeals
to be successful if they were tried. Much of this literature suggests that
the leaders of ethnic parties will concentrate on mobilizing members of
7
Kurt Weyland, “Neoliberal Populism in Latin America and Eastern Europe,” Comparative Politics
31 ( July 1999), 383.
8
Michael Conniff, “Introduction: Toward a Comparative Definition of Populism,” in Michael
Conniff, ed., Latin American Populism in Comparative Perspective Populism in Latin America (Albu-
querque: University of New Mexico Press, 1982); Paul Drake, “Conclusion: Requiem for Populism?”
also in Conniff; and Kurt Weyland, “Clarifying a Contested Concept: Populism in the Study of Latin
American Politics,” Comparative Politics 34 (October 2001).
9
Fernando Henrique Cardoso and Enzo Faletto, Dependency and Development in Latin America
(Berkeley: University of California Press, 1979); Guillermo O’Donnell, Modernization and Bureau-
cratic-Authoritarianism (Berkeley: Institute of International Studies, University of California, 1979);
Drake (fn. 8); and Weyland (fn. 8).
10
Donald L. Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985),
291; and Richard Gunther and Larry Diamond, “Species of Political Parties: A New Typology,” Party
Politics 9 (2003), 183.
their own ethnic group on the assumption that reaching out to mem-
bers of other ethnic groups would be futile.11 Horowitz, for example,
writes that an ethnic party, “recognizing that it cannot count on de-
fections from members of the other ethnic group, has the incentive
to solidify the support of its own group.”12 Leaders of ethnic parties
mobilize members of their own group by exaggerating the threat posed
by members of other ethnic groups and adopting exclusionary rhetoric
and platforms. This leads to elections that are increasingly polarized
along ethnic lines in a process known as “outbidding.”13
The literature on ethnic parties focuses on societies that are eth-
nically polarized and where politics, in the words of Horowitz, are
“unidimensional—along an ethnic axis.”14 In societies that are not
ethnically polarized, however, we would expect party competition to
be multidimensional and ethnic issues to be of lesser salience. Thus,
a party based in one ethnic group would presumably be able to attract
support from members of other ethnic groups. As a result, ethnically
based parties in nonpolarized societies would have greater incentives to
eschew exclusionary appeals and instead reach out to members of other
ethnic groups, as ethnopopulist parties have done.15
The arguments of Horowitz and others also assume that individu-
als have a single ethnic identity and that the boundaries dividing eth-
nic groups are clear and relatively stable.16 Constructivists, however,
have shown that ethnic identification is often fluid and that individuals
frequently have multiple ethnic identities.17 Where ethnic identifica-
tion is multiple and fluid, an exclusionary strategy would be counter-
productive because it would alienate those people whose ethnic identi-
ties comprise the excluded as well as the included group. By contrast,
an ethnically based party that adopts an inclusive strategy might appeal
11
Alvin Rabushka and Kenneth Shepsle, Politics in Plural Societies (Columbus: Merril, 1972); Tim-
othy D. Sisk, Power Sharing and International Mediation in Ethnic Conflicts (Washington,D.C.: U.S.
Institute of Peace Press, 1996); Benjamin Reilly, “Electoral Systems for Divided Societies,” Journal of
Democracy 13 (April 2002); Horowitz (fn. 10); and Gunther and Diamond (fn. 10).
12
Horowitz (fn. 10), 318.
13
Ibid., 526–30.
14
Ibid., 304.
15
I define ethnic polarization as the existence of widespread hostilities between members of differ-
ent ethnic groups, resulting in relatively frequent incidents of ethnically related violence.
16
See Kanchan Chandra, “Cumulative Findings in the Study of Ethnic Politics,” APSA-CP 12
(Winter 2001); and Kanchan Chandra, “Ethnic Parties and Democratic Stability,” Perspectives on Poli-
tics 3 ( June 2005), 235–52. The assumption that ethnic identities are clear, singular, and fixed may be
realistic in ethnically polarized societies since ethnic conflict can harden ethnic identities and elevate
certain identities to the exclusion of others, but it is less realistic in societies where ethnic polarization
is low. See Stephen Van Evera, “Primordialism Lives!” APSA-CP 12 (Winter 2001).
17
Fredrik Barth, ed., Ethnic Groups and Boundaries: The Social Organization of Cultural Difference
(Boston: Little, Brown, 1969); David Laitin, Hegemony and Culture: Politics and Religious Change among
the Yoruba (Chicago: University of Chicago Press 1986); and Chandra (fn. 16, 2001).
to all people who share a given ethnic identity without alienating those
who also have other ethnic identities.
In Latin America ethnic polarization has tended to be relatively low.
Indeed, Latin America has seen many fewer incidents of ethnic con-
flict than have most other regions of the world.18 Moreover, ethnicity
in the region is characterized by a great deal of fluidity, and people of-
ten identify, at least partially, with multiple, intersecting ethnic groups.
Some of this fluidity and multiplicity is the result of widespread mes-
tizaje, or miscegenation, which has blurred the lines between different
ethnic or racial groups and ensured that most Latin Americans have
mixed lineage. Prejudice and discrimination have also led many indi-
viduals to identify, at least some of the time, with ethnic groups that
are accorded higher social status, regardless of their own ethnic lineage.
Many Latin Americans who are mostly or entirely of indigenous de-
scent, for example, do not typically identify as indigenous, preferring
to identify themselves as mestizos. Many of these people nevertheless
have indigenous features, speak indigenous languages, respect certain
indigenous traditions, and sympathize with some of the demands of
the indigenous movement, leading some scholars to refer to them as
“indigenous mestizos.”19
Indigenous-based parties that adopt exclusionary rhetoric are likely
to alienate nonindigenous people as well as those indigenous people
who also identify as mestizo. Even some people who identify exclu-
sively as indigenous may find these parties’ exclusionary rhetoric and
platforms unpalatable, given the traditionally low levels of ethnic po-
larization in Latin American society. Moreover, in many Latin Ameri-
can countries, the indigenous population is itself divided along regional
or ethnolinguistic lines, and exclusionary indigenous parties may end
up alienating members of some of these indigenous communities. In-
clusive indigenous-based parties, by contrast, have a much broader po-
tential base of support. They have the potential to win support not only
from those people who self-identify exclusively as indigenous but also
from those people who have divided ethnic loyalties. Moreover, given
the low levels of ethnic polarization prevailing in Latin America, inclu-
sive indigenous-based parties may also attract some votes from people
18
Ted Robert Gurr, Minorities at Risk: A Global View of Ethnopolitical Conflicts (Washington, D.C.:
U.S. Institute of Peace, 1993); Matthew R. Cleary, “Democracy and Indigenous Rebellion in Latin
America,” Comparative Political Studies 33 (November 2000).
19
Marisol de la Cadena, Indigenous Mestizos: The Politics of Race and Culture in Cuzco, Peru,
1919–1991 (Durham, N.C.: Duke University Press, 2000); Programa de Naciones Unidas para el
Desarrollo (pnud), Interculturalismo y Globalización: Informe Nacional de Desarrollo Humano 2004 [In-
terculturalism and Globalization: National Report on Human Development] (La Paz: pnud, 2004).
who do not self-identify as indigenous but who nevertheless sympa-

thize with some aspects of their platforms. In Latin America, then, it is
not just feasible for an indigenous party or movement to be inclusive, it
is also the most rational electoral strategy to pursue.
In order to win votes from members of a variety of different eth-
nic groups, inclusive indigenous-based parties have had to develop
an appeal that extends beyond indigenous issues. The most successful
indigenous-based parties, the ethnopopulist parties, have used classi-
cal populist strategies to attract votes. The ethnopopulist parties, like
classical populist parties, have attracted politically disenchanted voters
by denouncing the traditional parties and the existing elites. They have
also won support by adopting a highly redistributive, nationalist, and
state interventionist agenda just as the classical populist parties once
did. They have, for example, opposed neoliberal reforms and other pol-
icies promoted by the U.S. government, advocating the nationalization
of natural resources and the redistribution of wealth. Finally, ethno-
populist leaders, like traditional populist caudillos, have attracted voters
by dint of their charismatic personalities. These strategies have enabled
the ethnopopulist parties to fuse traditional populist constituencies—
politically disenchanted urban mestizos with nationalist and statist
views—to their rural, largely indigenous base.
As Table 1 indicates, ethnopopulist parties combine elements of tra-
ditional ethnic and populist parties, but they are distinct from neolib-
eral populist parties in most ways. Ethnopopulist parties are like ethnic
parties in that they typically originate in a single ethnic group and de-
rive many of their leaders and supporters from that group. Moreover,
ethnopopulist parties, like traditional ethnic parties, make ethnic ap-
peals. They often portray themselves as the legitimate representative
of a particular ethnic group, they frequently prioritize the demands of
that group, and they typically use cultural and political symbols associ-
ated with it. However, unlike ethnic parties, ethnopopulist parties are
inclusive. Whereas ethnic parties use exclusionary rhetoric and plat-
forms to mobilize members of a single ethnic group, ethnopopulist
parties have sought to appeal to a variety of different ethnic groups.
Thus, they have eschewed exclusionary rhetoric, developed broad-
based platforms, and recruited leaders and candidates from a range of
different ethnic groups. Just as traditional populist movements reached
out to the middle classes, the peasantry, and the industrial bourgeoi-
sie to supplement their main base of support among the urban work-
ing class, ethnopopulist movements have sought to recruit supporters
among urban whites and mestizos, while still relying to a large extent
Table 1
Key Distinguishing Characteristics of Populist and Ethnic Parties
Based on Their Electoral Appeals
Classical Neoliberal
Ethnopopulist Ethnic Populist Populist
Makes ethnic appeals? yes yes no no
Has an ethnically inclusive
platform and leadership? yes no yes yes
Adopts nationalist rhetoric
and ideas? yes sometimes yes no
Advocates state intervention
and redistribution? yes sometimes yes no
Makes antisystem & anti-
establishment appeals? yes sometimes yes yes
Employs personalistic appeals? yes sometimes yes yes
on their rural indigenous core. As noted above, ethnopopulist parties

are also similar to populist parties in that they use classical populist ap-
peals to win votes.
Some studies have sought to define populism according to strictly
political criteria, specifically an uninstitutionalized, personalistic, and
plebiscitarian leadership style.20 Other studies have defined it accord-
ing to economic criteria, in particular, the implementation of redis-
tributive programs, deficit spending, and widespread state interven-
tion in the economy.21 Both of these types of studies have focused on
populism as a style of governance—that is, as a means of governing the
economy (economic populism) or the polity more generally (political
populism). By contrast, this article classifies populist parties according
to their electoral appeals because the focus of this study is explaining
why ethnopopulism has been a successful electoral strategy. I use both
economic and political criteria to define populism, since ethnopopulist
leaders have used populist economic strategies (such as calls for redis-
tribution of income and nationalization of natural resource companies)
as well as populist political strategies (for example, antiestablishment
and antisystem appeals) to win votes.
20
Weyland (fn. 8).
21
Rudiger Dornbusch and Sebastian Edwards, eds., The Macroeconomics of Populism in Latin Amer-
ica (Chicago: University of Chicago Press, 1991).
Institutional Explanations for the Rise of the MAS

The remainder of this article examines the causes of the rise of the most
successful ethnopopulist party to date: the Movimiento al Socialismo
in Bolivia. Some studies have attributed the rapid growth of the mas in
part to the institutional reforms that Bolivia carried out in the 1990s.
Van Cott, for example, maintains that Bolivia’s pr system, which used
large multimember districts, “made it difficult for geographically con-
centrated indigenous movements to win enough votes nationwide to
gain national office.”22 She and others have argued that Bolivia’s con-
stitutional reform of 1994 facilitated the rise of the mas by creating
smaller single-member districts for elections to one tier of the lower
chamber of the legislature.23 This explanation is unconvincing, how-
ever, because the mas or its predecessor would have won approximately
the same number of seats in 1997 and 2002 if Bolivia had retained the
previous proportional representation system. For example, the vote that
the mas earned in the 1997 elections would have gained it three leg-
islative seats under the old system, as opposed to the four seats it won
under the new system. Indeed, the German-style mixed system that
Bolivia adopted was designed to achieve proportionality by using the
pr tier to offset any disproportionality created by the outcomes of the
races in the single-member districts.
Another institutional explanation for the rise of the mas has focused
on the decentralization process Bolivia underwent in the mid-1990s.
The 1994 Law of Popular Participation created 311 municipalities
nationwide and called for elections to be held for mayoral and council
member positions throughout the country. The mas was formed in the
wake of the passage of this law, and some scholars have argued that
the municipal elections enabled the mas to gain a foothold at the local
level, which its members used as a stepping-stone to national office.24
This explanation for the rise of the mas is also problematic, however.
To begin with, the municipal electoral victories of the mas and its pre-
decessor in 1995 and 1999 were confined largely to the department of
22
Donna Lee Van Cott, “From Exclusion to Inclusion: Bolivia’s 2002 Elections,” Journal of Latin
American Studies 35 (November 2003), 756.
23
Ibid.; Van Cott (fn. 4, 2005); Pablo Stefanoni, “Algunas reflexiones sobre el mas-ipsp,” Temas
Sociales 25 (2004); and author interview with Jorge Lazarte, La Paz, August 2, 2004.
24
Andolina (fn. 5); Van Cott (fn. 4, 2003); Miguel Urioste, “Ninguno de los Indígenas que Está en
el Parlamento Hoy en Día Hubiera Llegado a ese Nivel si no Era a Través del Proceso de la Partici-
pación Popular,” in Diego Ayo, ed., Voces Críticas de la Descentralización [Critical Voices on Decentral-
ization] (La Paz: Friedrich Ebert Stiftung, 2004); author interview with Silvia Rivera Cusicanqui, La
Paz, July 16, 2004; and author interview with Gustavo Torrico, La Paz, July 22, 2004.
Cochabamba.25 It therefore seems unlikely that the mas’s strong elec-

toral performance outside of Cochabamba in the 2002 elections could
be explained by the resources, experience, or reputations won by the
party’s mayors. Moreover, there is no evidence that the mas has per-
formed better in those municipalities where it has elected mayors or
council members. To the contrary, between 1997 and 2002 the mas
increased its share of the total vote by a much smaller margin in those
municipalities where it had elected mayors in 1995 than in those mu-
nicipalities where it did not control the mayoralty—the mas boosted its
vote by 13 percentage points in the former municipalities and 19 points
in the latter. Thus, it does not appear that either the decentralization
law or the shift to a mixed electoral system can explain the rapid rise
of the mas.
The Inclusive Ethnic Appeal of the MAS

This article, by contrast, argues that one of the main factors behind
the success of the mas was the party’s inclusive ethnopopulist appeal.
To be sure, other factors, such as the strengthening of indigenous con-
sciousness and organization and growing disenchantment with the tra-
ditional parties and their record of governance, played a role in the rise
of the mas, but as the following discussion shows, the mas’s ethno-
populist rhetoric and platform helped the party take advantage of these
developments.
Several features of the ethnic landscape in Bolivia helped make this
appeal successful. To begin with, a large proportion of the Bolivian
population is of indigenous ancestry, and, as a result, parties that seek to
appeal to the indigenous population in Bolivia have an important po-
tential pool of voters to attract. According to the 2001 census, roughly
half of the Bolivian population speaks an indigenous language and
almost two-thirds of the population identifies with some indigenous
ethnolinguistic category. The indigenous population in Bolivia is frag-
mented, however, most notably between the Quechua-speaking and
Aymara-speaking populations. According to the census, 27.6 percent
of the total population speaks Quechua, 18.4 percent speaks Aymara,
and 1.2 percent of the population speaks a variety of different lowlands
indigenous dialects. To win support from these disparate groups, indig-
enous leaders and parties have needed to be inclusive.
25
Gonzalo Rojas, “La Elección de Alcaldes en los Municipios del País en 1999–2000: Persistencia
de la Coalición Nacional,” Opiniones y Análisis 49 (March 2000).
The low level of ethnic polarization and the fluidity of ethnic identi-
fication in Bolivia also favor an inclusive approach. Although discrimi-
nation against indigenous people is commonplace in Bolivia, ethnic vi-
olence is rare, and relations between members of different ethnic groups
are relatively harmonious. In Bolivia, as elsewhere in Latin America,
the state has actively promoted mestizaje and suppressed indigenous
identities.26 Partly as a result, most Bolivians self-identify as mestizo
rather than as indigenous. Surveys by the Ministry of Human Devel-
opment, the United Nations Development Program, and the Latin
American Public Opinion Project (lapop) have found that between
60 and 70 percent of the Bolivian population self-identifies as mestizo,
whereas less than 20 percent of the population self-identifies as in-
digenous.27 Nevertheless, many of these self-identified mestizos speak
indigenous languages and identify to some degree with indigenous cul-
ture. Indeed, in a recent survey by lapop, 55 percent of the people who
self-identified as mestizo spoke an indigenous language and 70 percent
of the people who so self-identified stated that they belonged to an
indigenous ethnolinguistic category, mostly Quechua or Aymara.28 As
we shall see, the mas’s inclusive indigenous profile appealed not only
to self-identified indigenous people but also to the numerically much
larger group of indigenous mestizos. It even won the support of some
whites and mestizos who did not identify as indigenous at all.
26
Félix Patzi Paco, Insurgencia y Sumisión: Movimientos Indígeno-Campesinos (1983–1998) (La Paz:
Muela del Diablo, 1999), 27–34; Javier Sanjinés C., Mestizaje Upside Down: Aesthetic Politics in Modern
Bolivia (Pittsburgh, Pa.: University of Pittsburgh Press, 2004); Rachel M. Gisselquist, “Ethnicity, Class
and Party Competition: The Bolivian Case” (Paper presented at the annual meeting of the American
Political Science Association, Philadelphia, August 31–September 3, 2006); and Yashar (fn. 3).
27
The surveys have posed some variation of the following question: “Do you consider yourself
white, mestizo, or indigenous?” In some cases, additional categories such as cholo, black, or other are
included. See Gonzalo Rojas and Luis Verdesoto, La Participación Popular como Reforma de la Política:
Evidencias de una Cultura Democrática Boliviana [Popular Participation as a Political Reform: Evidence
of a Bolivian Democratic Culture] (La Paz: Ministerio de Desarrollo Humano, 1997); Programa de
Naciones Unidas para el Desarrollo (fn. 19); Mitchell A. Seligson, La Cultura Política de la Democracia
Boliviana [The Political Culture of Bolivian Democracy] (La Paz: Encuestas y Estudios, 1999); idem,
La Cultura Política de la Democracia en Bolivia: 2000 [The Political Culture of Democracy in Bolivia:
2000] (La Paz: Universidad Católica Boliviana, 2000); idem, Auditoria de la Democracia: Bolivia, 2002
[Audit of Democracy: Bolivia, 2002] (La Paz: Universidad Católica Boliviana, 2003). Mitchell A.
Seligson, Daniel Moreno Morales, and Vivian Schwarz Blum, Democracy Audit: Bolivia 2004 Report
(Nashville: lapop, 2004); Mitchell A. Seligson, Abby B. Cordova, Juan Carlos Donoso, Daniel More-
no Morales, Diana Orcés, and Vivian Schwarz Blum, Democracy Audit: Bolivia 2006 Report (Nashville:
lapop, 2006).
28
The 2006 lapop survey included a question about indigenous identity that was modeled on a
question from the 2001 census. It asked: “Do you consider yourself to belong to one of the following
native or indigenous peoples? Quechua; Aymara; Guaraní; Chiquitano; Mojeño; other native; none
of the above.” In the 2001 census 62 percent of the population chose one of these indigenous ethno-
linguistic categories, and in the 2006 lapop survey 71 percent of the population selected one of the
indigenous categories. This question was criticized widely, however, in part because it did not include
the option of self-identifying as mestizo.
The mas’s appeal to people with indigenous backgrounds had a

number of different components. First, in contrast to the parties that
traditionally dominated Bolivian politics, most of the mas’s leadership
and candidates were indigenous, at least initially, and these candidates
attracted indigenous voters. According to a 2006 lapop survey, 29 per-
cent of people who self-identified as indigenous felt better represented
in the government and legislature by leaders of their same ethnic back-
ground. Second, the mas had strong ties to indigenous organizations
throughout the country. Indeed, the predecessor of the mas, the iu/
asp, had been founded by indigenous organizations. These organiza-
tional ties provided the mas with legitimacy in many indigenous areas
as well as a network of activists and supporters. Third, the mas styled
itself as the representative of the indigenous population and made nu-
merous rhetorical and symbolic appeals to it. The leaders of the mas,
for example, invoked indigenous sayings, used indigenous clothing and
banners, and trumpeted the achievements of indigenous civilizations.
Fourth and finally, the mas developed a platform that had broad ap-
peal within the indigenous population. The mas embraced many tra-
ditional indigenous demands, from agrarian reform to bilingual edu-
cation. Moreover, many of its positions on nonethnic issues, such as
the nationalization of natural resource industries, also appealed to most
indigenous people.
Previous indigenous-based parties, commonly known as the Katar-
ista parties, had also made ethnic appeals, but these parties failed to
reach out beyond their narrow base in the Aymara population.29 With
their incendiary rhetoric, the more radical Katarista parties, such as the
Movimiento Indio Tupak Katari (mitka) and the Movimiento Indí-
gena Pachacuti (mip), alienated voters, especially whites and mestizos
but also many indigenous people.30 For example, the leader of the mip,
Felipe Quispe, has become well known for his ethnonationalist rhet-
oric.31 In an interview with Canessa, Quispe denounced whites and
spoke of creating an Aymara homeland:
Those lying q’aras [a pejorative term for whites]. When the Pachamama walks
again in Qullasuyu, when her laws reign, then we will be able to judge them.
29
Conciencia de Patria (condepa) formulated a somewhat successful ethnopopulist appeal in the
1990s, but condepa, like the Katarista parties, never developed a following or an organizational base
outside of Aymara areas. Moreover, the party, like the traditional parties, was led by mestizos, and that
ultimately undermined its appeal in indigenous areas. It fell apart in the wake of leadership disputes
caused by the death of its charismatic founder, Carlos Palenque.
30
These parties have frequently been referred to as Indianista parties.
31
Quispe has frequently denounced whites, saying, for example, that “they want to bathe them-
selves in indigenous blood”; author interview with Felipe Quispe, La Paz, July 29, 2004.
Those who want to leave can go; but those who stay will eat what we eat; they
will work the way we work, dripping with sweat; they will have blisters on their
hands; they will suffer like we do. Then truly the Aymara nation, what people
call the indigenous [nation], what we call Qullasuyu, will come forth.32
Other Katarista parties, such as the Movimiento Revolucionario Tu-

pak Katari de Liberación (mrtkl) and the Frente Único de Liberación
Katarista (fulka), adopted more moderate and inclusive stances, and
their leaders typically avoided exclusionary rhetoric. Even these par-
ties, however, failed to take many steps to win non-Aymara voters. The
Katarista parties did not recruit many white, mestizo, and Quechua
leaders to their organizations.33 Nor did they establish ties to many
organizations outside of the Aymara heartland.34 As a result, these par-
ties fared quite poorly outside of Aymara areas, never winning more
than 3.3 percent of the vote in majority Quechua-speaking provinces
or more than 1.8 percent of the vote in majority Spanish-speaking
provinces. The most successful Katarista party, Quispe’s mip, won 26
percent of the vote in 2002 in provinces that had an Aymara-speaking
majority, but it won less than 2 percent of the vote in other provinces,
which added up to a mere 6 percent of the national vote.
The leaders of the mas, by contrast, have avoided exclusionary rhet-
oric and have emphasized the welcoming nature of the party. Indeed,
in his inaugural speech, Morales stated that “we all have the right to
live in this land” and emphasized that “the indigenous movement is
not exclusionary; it’s inclusive.”35 Initially, however, Quechua speak-
ers dominated the mas. The party sprang from the largely Quechua-
speaking coca growers’ unions based in rural Cochabamba, and at the
outset these unions provided most of the party’s supporters as well as its
candidates. As a result, the party fared poorly outside of these unions’
base. In 1997, for example, the predecessor of the mas, the iu/asp, won
30 percent of the vote in rural Cochabamba and less than 1.4 percent
32
Andrew Canessa, “Todos Somos Indígenas: Towards a New Language of National Political
Identity,” Bulletin of Latin American Research 25 (April 2006), 251.
33
There were some exceptions. The mrtkl, for example, recruited Filemón Escobar, a mestizo
union leader, as its vice presidential candidate in 1985, and it elected Walter Reinaga, a Quechua
leader, as a deputy from Potosí that same year.
34
In a few cases, the Katarista parties did establish alliances with traditional parties, such as the
udp and the mnr, but these were unequal alliances, which typically resulted in the subordination and
co-optation of the indigenous parties and leaders. See Javier Hurtado, El Katarismo (La Paz: Hisbol,
1986), 112–18; Esteban Ticona, Gonzalo Rojas, and Xavier Albó, Votos y Wiphalas: Campesinos y Pueb-
los Originarios en Democracia [Votes and Wiphalas: Peasants and Native Peoples under Democracy] (La
Paz: cipca, 1995), 121–56.
35
Evo Morales, “La reserve moral de la humanidad,” in Pablo Stefanoni and Hervé Do Alto, Evo
Morales de la Coca al Palacio: Una Oportunidad para la Izquierda Indígena [Evo Morales from Coca to
the Presidential Palace: An Opportunity for the Indigenous Left] (La Paz: Malatesta, 2006), 133.
elsewhere. In the early 2000s, however, the mas sought to diversify

its base. The party forged ties to Quechua groups outside of Cocha-
bamba, and that helped it significantly increase its share of the vote
in Quechua-speaking areas throughout the country. It struck an alli-
ance, for example, with Felix Vázquez, the powerful head of a peasant
federation in northern Potosí, who helped deliver substantial votes to
the party in this department. Even more importantly, the mas forged
ties to many Aymara groups and recruited numerous Aymara leaders
as candidates, including Evo Morales, who was the party’s presidential
candidate in 2002 and 2005.36 As a result, the party’s share of the vote
in majority Aymara-speaking provinces rose from only 3.3 percent in
1997 to 32.4 percent in 2002 and 75.2 percent in 2005.37
The mas also reached out to whites and mestizos. In a 2004 inter-
view with the author, Dionisio Núñez, an Aymara legislator from the
mas, explained that initially the party was dominated by indigenous
people, but
in the end we came to understand that we didn’t want to go from being ex-
cluded to excluding others, that we had to include more people, business people,
the middle classes. . . . Originally, there were three peasant organizations that
founded the mas. Two years ago, the reformulation of the mas began. . . . The
mas ceased to be solely indigenous and peasant.38
Similarly, in a 2007 interview with the author, Ricardo Díaz, a senator

from the mas, acknowledged that the mas “has moderated in the sense
of greater inclusion. Before I saw it as very biased toward the indig-
enous. Now we are taking into account professionals, urbanites.”39 In
an effort to win white and mestizo votes, the party nominated a promi-
nent white/mestizo leftist intellectual as its vice presidential candidate
in 2002 (Antonio Peredo) and again in 2005 (Alvaro García Linera). It
also recruited numerous whites and mestizos as candidates for the leg-
islature. Whereas in 1997 the mas’s legislative contingent was entirely
indigenous, in 2002 more than one-third of the party’s legislators were
white or mestizo and by 2005 whites and mestizos represented at least
half of the contingent.
36
Although Evo Morales is Aymara, he migrated to a Quechua-speaking area as a young man,
learned Quechua, and became a leader of the Quechua-dominated coca grower unions. He thus has a
certain panindigenous appeal. See Canessa (fn. 32), 250.
37
The mas also forged alliances with indigenous groups in the Amazon. For example, it struck an
alliance with the Coordinadora de Pueblos Étnicos de Santa Cruz (cpesc) and allowed it to help select
candidates in the department of Santa Cruz in 2002. See Van Cott (fn. 4, 2005), 91.
38
Author interview with Dionisio Nuñez, La Paz, July 21, 2004.
39
Author interview with Ricardo Díaz, La Paz, August 17, 2007.
The mas’s candidate strategies and ethnic appeals varied somewhat

from constituency to constituency and election to election. In national
or departmental elections, the mas’s discourse tended to be very in-
clusive and in these elections the party recruited numerous white and
mestizo candidates.40 In municipal races and in the legislative elections
for the uninominal districts, the mas would recruit candidates that
reflected the social composition of the municipality or district. Thus,
in rural indigenous districts and municipalities the candidates typi-
cally came from indigenous backgrounds and oriented much of their
discourse toward indigenous people. By and large, however, the mas
sought to maintain a balance between rural indigenous and urban mes-
tizo candidates, and it sought to include as many groups as possible.41
The mas also tried to boost its standing among whites and mestizos
by establishing ties to a variety of unions and other organizations com-
posed mostly of mestizos and centered for the most part in urban areas.
These included organizations of teachers, pensioners, small business-
people, adjudicators, artisans, truck drivers, and the self-employed. In
addition, the mas forged ties to some smaller, mestizo-dominated left-
of-center political parties. For example, in the 2005 elections, the mas
allied with the Movimiento Sin Miedo (msm) of Juan del Granado,
the mayor of La Paz.42 These allied organizations provided the mas
with human and material resources from activists to transportation and
food, and they gave the party an organizational base outside of the in-
digenous movement, outside of rural areas, and outside of the Depart-
ment of Cochabamba and the highlands more generally.
The mas’s efforts to reach out to white and mestizo voters have
been largely successful and over time the mas made important inroads
among both groups. As Table 2 indicates, the mas won 32 percent of
the vote of people who self-identify as white in the 2005 elections, up
from 6 percent in 2002, according to the 2006 lapop survey.43 It was
the large mestizo vote, especially the indigenous mestizo vote, that was
crucial to the mas’s victory in 2005, however. The mas won 51 percent
of the vote of people who self-identified as mestizo in 2005, up from
20 percent in 2002. Moreover, mestizos represented 62 percent of the
40
In Bolivia, senators, prefects, and 60 of the 130 deputies are elected at the departmental level.
The remaining 70 deputies are elected from uninominal districts.
41
Author interview with Leonilda Zurita, La Paz, August 20, 2007; author interview with Ricardo
Díaz (fn. 39).
42
The mas has eschewed alliances with the traditional parties, however, on the grounds that those
sorts of alliances might compromise its autonomy or political project.
43
According to the 2006 lapop survey, the mas also substantially increased its share of the urban
vote, winning the support of almost half of urban voters in 2005, as opposed to less than 20 percent
in 2002.
Table 2
The MAS’s Share of the Self-Reported Vote of People from
Various Ethnic Categories in 2002 and 2005
2002 Elections 2005 Elections
(%) (%)
Self-identified indigenous 37.8 71.1
Self-identified mestizos who speak indigenous languages 27.6 63.6
Self-identified mestizos who speak only Spanish 11.5 34.2
Self-identified whites 5.8 31.6
All self-reported voters 21.6 53.3
Source: lapop 2006 Bolivia survey.
party’s total vote in 2005.44 More than two-thirds of the mestizos who
reported voting for the mas had grown up speaking an indigenous lan-
guage. Indeed, as Figure 1 indicates, these so-called indigenous mes-
tizos accounted for 43 percent of the mas’s total vote, more than any
other ethnic group. Had the mas adopted a more exclusionary plat-
form, it likely would have alienated many of the indigenous mestizos.
A more exclusionary approach presumably would also have alienated
the mas’s white supporters as well as the nonindigenous mestizos, who
together accounted for more than a quarter of the mas’s total vote.
Many of these nonindigenous people (as well as many of the mas’s
indigenous supporters) were drawn to the party by its populist rhetoric
and platform, but the mas’s inclusive approach helped make them feel
comfortable with the party.
The mas’s inclusive approach has caused tensions within the party,
however, particularly with respect to candidates for the legislature and
bureaucratic and ministerial posts. Some indigenous leaders have com-
plained that middle-class whites and mestizos have seized many of the
key positions within the government and the party. In a 2007 interview
with the author, Lino Villca, an indigenous senator and longtime mas
leader, complained that “the indigenous movement is isolated. We have
the president and the Ministry of Foreign Relations, but the middle
class has the rest of the ministers. . . . Now the middle class defines
the strategy of Evo Morales. The indigenous class is only for mobili-
zations.”45 Nonetheless, as Villca acknowledged, the inclusive strategy
44
According to the 2006 lapop survey, whites represented 7 percent of the mas’s total vote in
2005, although this represented an increase from only 3 percent in 2002. By contrast, self-identified
indigenous people represented 28 percent of the mas’s total vote in 2005, down slightly from 33 per-
cent in 2002.
45
Author interview with Lino Villca, La Paz, August 15, 2007.
white other
mestizo (Spanish 7% 3% indigenous or native
only speaker) 28%
19%
43%
mestizo (indigenous
language speaker)
Figure 1
The Ethnic Composition of the MAS’s 2005 Vote
Source: lapop 2006 Bolivia survey.
of the mas has yielded results, and the party is unlikely to abandon it
anytime soon in spite of any tensions it might cause.
The MAS’s Populist Appeal

Equally crucial to the success of the mas was its embrace of traditional
populist electoral strategies. The mas used three principal populist ap-
proaches to attract supporters. First, it adopted an antiestablishment
message, taking advantage of widespread disenchantment with the tra-
ditional parties and elites. Second, the mas espoused redistributive, na-
tionalist, and state interventionist policies, feeding on growing unrest
with neoliberal policies and U.S. intervention. Third and finally, the
mas relied heavily on the charismatic appeal of Evo Morales. These
populist strategies helped the party win the support of people of all dif-
ferent ethnic backgrounds.
Like many populist parties, the mas has relied in part on personal-
istic linkages with voters. Evo Morales is a polarizing figure, but his
down-to-earth grassroots style has appealed to many voters, particularly
poor and indigenous people. Morales has proved to be a tireless cam-
paigner who carried his campaign to areas of Bolivia long neglected by
national-level politicians. During the 2002 campaign Antonio Peredo,
the mas’s vice presidential candidate in 2002, observed that many peo-
ple would come to the party’s rallies simply to see Morales.46 Morales’s
46
Author interview with Antonio Peredo, La Paz, July 22, 2004.
personal popularity has grown so much in recent years that some ana-
lysts now speak of the cult of Evismo, similar to that surrounding some
other populist leaders like Juan Perón.47 Not surprisingly, Morales has
consistently fared better than the party as a whole in general elections.
In the 1997 elections, when he ran as a candidate for the legislature, he
earned more votes than any other candidate in the country, and in the
2005 elections, he performed better in every department than did the
mas’s prefectural candidates. Unlike some populist parties, however,
the mas also has a strong organizational base because of its links to
social movements. Indeed, according to the lapop surveys, many of the
mas’s supporters in the 2002 and 2005 elections participated in unions
or trade associations.
Populist parties have traditionally taken advantage of disenchant-
ment with existing parties, and political leaders and the mas is no ex-
ception. Between 1985 and 2003 Bolivia maintained what has become
known as a system of pacted democracy in which the three traditional
parties—the Movimiento Nacionalista Revolucionario (mnr), the Ac-
ción Democrática Nacionalista (adn), and the Movimiento de Izqui-
erda Revolucionaria (mir)—took turns governing the country, usually
in alliance with each other and younger parties. These parties carried
out many important reforms and attracted considerable support for a
time, but a stalling economy, repeated corruption scandals, and a grow-
ing number of protests gradually undermined their support. By late
2001 almost half of the population said that parties were not necessary
for democracy, as opposed to only 17 percent holding that opinion in
1993.48 In the 2002 elections the mnr managed to finish first, but with
a mere 22.4 percent of the vote, which put it less than two percentage
points ahead of the mas. The popularity of the mnr government, more-
over, steadily deteriorated as it failed to get the economy back on track
or resolve the widening protests. Support for the government of Presi-
dent Gonzalo Sánchez de Lozada hit rock bottom in October 2003,
when it violently repressed protests in the city of El Alto, leading to the
deaths of more than fifty people. Abandoned by most of his former al-
lies, Sánchez de Lozada resigned and fled the country, leaving the rep-
utation of his party and the parties that had supported him in tatters.49
47
“El Evismo Ensalza a Evo en el Poder,” La Razón, August 5, 2006, www.la-razon.com/
versiones/20060805%5F005624/nota_244_316940.htm (accessed August 10, 2006); “En la Eman-
cipación de los Pueblos, Evo es Sustituible,” La Razón, August 5, 2006, www.la-razon.com/
versiones/20060805%5F005624/nota_244_316937.htm (accessed August 10, 2006).
48
Fernando Calderón and Eduardo Gamarra, Crisis y Reforma de los Partidos en Bolivia (La Paz:
pnud, 2004), 17.
49
A survey carried out in 2004 found that parties were the least trusted institution in Bolivia that
year. See Seligson, Moreno, and Blum (fn. 27), 102.
So poor was the reputation of the traditional parties that only the mnr
opted to compete in the 2005 presidential elections, and it won a mere
6.5 percent of the vote.
The mas was in a good position to take advantage of disenchant-
ment with the traditional parties because of its outsider status.50 Un-
like the main parties, the mas never participated in the various coali-
tion governments that ruled Bolivia between 1985 and 2003. To the
contrary, the mas consistently criticized the ruling parties and their
policies, and it participated in numerous social protests against them,
ranging from marches and demonstrations to roadblocks. The mas also
differed from the traditional parties in that it had no party bureaucracy
to speak of, and its candidates were typically social movement leaders
rather than career politicians. Indeed, the mas was hardly a political
party at all, but rather was a collection of numerous social organiza-
tions. Thus, the mas had strong outsider credentials, which helped it
to capture the support of those voters who were fed up with the tradi-
tional parties and political elites.
In 2002 the mas faced strong competition for politically disen-
chanted voters from a new party, the Nueva Fuerza Republicana (nfr),
which ran an antiestablishment campaign. In the 2005 elections, how-
ever, the mas was the only one of the main contenders that had a clear
antiestablishment profile. podemos, the mas’s main rival in 2005, was
made up almost entirely of politicians who had recently left the tradi-
tional parties, especially the adn. Indeed, Jorge Quiroga, the presiden-
tial candidate of podemos in 2005, was a former leader of the adn who
had served as president of Bolivia in 2001and 2002.51 Thus, podemos
did not hold much appeal for antiestablishment voters. According to
the 2006 lapop survey, in the 2005 elections the mas won 54.7 per-
cent of the votes of people who expressed no trust in parties, whereas
podemos won only 19.5 percent. The mas’s appeal to politically disen-
chanted voters was even more apparent from the high levels of support
it won from voters who participated in protests against previous gov-
ernments. The mas won 77 percent of the vote of people who partici-
pated in protests against the administration of Carlos Mesa, whereas
podemos earned the support of less than 10 percent of these voters.
50
Carlos Böhrt Irahola, “Voto Presidencial y Voto Uninominal en las Elecciones de 2002,” Opin-
iones y Análisis 58 (September 2002); Salvador Romero Ballivián, Geografía Electoral de Bolivia [Elec-
toral Geography of Bolivia], 3rd ed. (La Paz: fundemos, 2003); René Antonio Mayorga, “Bolivia’s De-
mocracy at the Crossroads,” in Scott Mainwaring and Frances Hagopian, eds., The Third Wave of
Democratization in Latin America (Cambridge: Cambridge University Press, 2005).
51
Unidad Nacional (un), which finished third in the 2005 elections, was also composed principally
of former members of the traditional parties, especially the mir.
The mas has attracted voters not only because of its antiestablish-
ment rhetoric, but also because of its populist economic policy positions.
The market-oriented policies Bolivian governments implemented be-
ginning in the mid-1980s stabilized the economy and generated some
initial growth, but by the late 1990s the Bolivian economy had be-
gun to stagnate. In 2005 gross domestic product per capita in Bolivia
was actually lower than it had been in 1998, leading to widespread
disenchantment with the neoliberal economic model. The mas capi-
talized on this disenchantment by denouncing neoliberal policies and
proposing state interventionist measures, including the recuperation of
privatized companies, in order to redistribute income and generate an
economic recovery. In opposing neoliberal policies, the mas often ap-
pealed to nationalist sentiments. For example, in its 2002 governing
program, the mas declared:
The neoliberal parties such as the mnr, adn, mir, mbl, ucs, nfr, condepa and
other small groupings of their corrupt circle, are characterized by the submis-
sion and betrayal of the country, by the handing over of the national patrimony
almost without charge to the voraciousness of international capital and its di-
rectors, who impose conditions of poverty on the legitimate owners of natural
resources.52
The mas made control of Bolivia’s considerable natural gas depos-

its a centerpiece of its platform, particularly in the 2005 campaign. It
helped block Sánchez de Lozada’s plan to export gas through Chile
and then pressed the ensuing government of Carlos Mesa to renegoti-
ate its contracts with the foreign firms that exported the gas. After
being elected president, Morales went so far as to seize control of the
natural gas fields and demand that the foreign firms pay a higher share
of their profits to the state. These moves proved quite popular. Indeed,
Morales’s public approval rating soared by 13 points to 81 percent in
the wake of his takeover of the gas fields.53
The mas also successfully appealed to nationalist sentiments in op-
posing the coca eradication program that the Bolivian government ex-
panded in the late 1990s under pressure from the U.S. government.
The mas’s opposition to coca eradication programs provoked the inter-
vention of the U.S. ambassador to Bolivia, Manuel Rocha, who gave a
speech shortly before the 2002 election warning Bolivians not to vote
for Morales. In the wake of this speech, which many Bolivians viewed
52
Movimiento al Socialismo, Programa de Gobierno: Territorio, Soberanía, Vida [Program of Gov-
ernment: Territory, Sovereignty, and Life] (La Paz: mas, 2002), 9.
53
“Bolivians Love Evo: Want Closer Ties with the U.S.,” Latin American Weekly Report, May 23,
2006, 16.
as inappropriate interference in their internal affairs, Morales’s support

went up by five points in surveys taken in the principal cities of the
country.54
Left-of-center voters have been particularly attracted by the mas’s
statist and nationalist agenda.55 According to the 2006 lapop survey,
the mas won 75.3 percent of the vote of people who identified with
the left in the 2005 election, whereas podemos won only 12.0 percent
of these voters.56 Nearly half of the people who reported voting for
the mas in 2005, as in 2002, identified with the left, with most of the
remainder self-identifying as centrists. Many of these left-wing voters
felt abandoned by the traditional parties’ embrace of the United States
and neoliberal policies. The mas, moreover, aggressively courted this
left-of-center constituency by recruiting well-known leftists to serve as
candidates for the vice presidency and the legislature and by develop-
ing a traditional left-wing platform in many areas.57
The mas’s nationalist and state-interventionist agenda appealed to
some centrist and right-wing voters as well, however. Bolivians who
self-identified as being on the right or in the center were almost as
likely to support the nationalization of the gas industry as people on the
left, and many of these nationalistic centrists and rightists voted for the
mas.58 According to the 2006 lapop survey, 53.8 percent of centrists who
strongly supported the nationalization of the gas industry reported voting
for the mas, as opposed to only 41 percent of those who mildly sup-
ported it and 29.6 percent of those who disapproved of it. Overall,
the mas won 48.2 percent of the vote of people who self-identified
as centrists and 32.3 percent of the vote of people who self-identified
as being on the right—an impressive performance for a left-of-center
party.59
54
Romero Ballivián (fn. 50), 251.
55
Salvador Romero Ballivián, “La Elección Presidencial 2002: Una Visión de Conjunto,” Opiniones y
Análisis 57 (September 2002), 191; Seligson et al. (fn. 27, 2006), 89–90.
56
The survey asked people to place themselves on a left-right scale of 1–10. Here, I classify people
who place themselves at 1–4 as leftists, 5–6 as centrists, and 7–10 as rightists. In the 2006 lapop
survey, 31 percent of the people who answered this question identified themselves as being on the left,
46.5 percent on the center, and 22.5 percent on the right. However, a significant percentage of the
interviewees (24.5 percent) failed to respond.
57
María Teresa Zegada Claure, “Sorpresas de la Elección: mnr, mas, nfr y adn,” Opiniones y Análi-
sis 57 (September 2002), 51; Félix Patzi Paco, “De Movimiento Indígena al Fracaso en la Escena del
Parlamento,” Temas Sociales 25 (2004); Van Cott (fn. 4, 2005).
58
According to the 2006 lapop survey, 62 percent of people who self-identified as being on the
left strongly approved (8–10 on a 10 point scale) of the nationalization of the gas industry, as did 60
percent of centrists and/or rightists.
59
Ethnic ties also help explain why many people on the center and the right supported the mas. Ac-
cording to the 2006 lapop survey, 61.4 percent of centrists and 56.8 percent of rightists who grew up
speaking an indigenous language supported the mas in 2005, as opposed to only 30.0 percent of those
centrists and 14.3 percent of those rightists who did not grow up speaking an indigenous language.
An Analysis of the MAS’s 2005 Vote

Table 3 presents the results of a multinomial logit analysis of the char-
acteristics of voters for the mas in the 2005 elections.60 For simplicity,
I report only the results of a comparison of the likelihood of voting for
one of the two main parties—the mas and podemos—which together
won 82 percent of the valid vote.61 Each parameter estimate in the first
column represents the predicted marginal effect of the variable on the
log-odds ratio of voting for the mas versus podemos. The data for this
analysis come from the 2006 lapop survey.62 As with most postelec-
tion surveys, this survey slightly overestimated the percentage of the
population that reported voting for the winner of the election, the mas
in this case, and slightly underestimated the percentage of the popula-
tion that reported voting for the other parties.63 In addition, the survey
overestimated voter turnout in the election: 90.9 percent of registered
voters reported voting in the survey, whereas only 84.5 percent of reg-
istered voters actually voted in the election.64
The analysis generated a number of noteworthy findings. First, the
mas’s populist appeal—that is, its antiestablishment, nationalist, state
interventionist, and redistributive message—was clearly successful in
winning support. Left-wing voters, people who favored the national-
ization of the gas industry, and people who had participated in protests
against the administration of Carlos Mesa (2003–5) were all signifi-
cantly more likely to report having voted for the mas, according to
the analysis. In order to estimate precisely what effect these variables
had on the probability of voting for the mas, I carried out a series of
60
Dow and Endersby argue that multinomial logit is superior in some aspects to multinomial
probit, particularly for applications such as this, where “a voter casts a ballot for a candidate or party
selected from a fixed, stable pool of alternatives”; see Jay K. Dow and James W. Endersby, “Multi-
nomial Probit and Multinomial Logit: A Comparison of Choice Models for Voting Research,” Elec-
toral Studies 23 (March 2004), 108. For a contrasting view, see Michael R. Alvarez and Jonathan Na-
gler, “When Politics and Models Collide: Estimating Models of Multiparty Competition,” American
Journal of Political Science 42 ( January 1998).
61
None of the six other parties that competed in this election earned more than 7 percent of the
vote.
62
For more information on the survey and the wording of the questions, see Seligson et al. (fn.
27, 2006).
63
In the 2006 lapop survey, 53 percent of voters reported casting their ballots for the mas and 25
percent reported voting for podemos. According to the official returns, the mas received 50 percent of
the total vote and podemos earned 26 percent.
64
There is little evidence to suggest that the increase in votes for the mas in 2005 is due to in-
creased voter turnout. Voter turnout as a percentage of the estimated voting-age population actually
declined between 2002 and 2005, both nationwide and in majority indigenous provinces. This decline
in turnout was largely a result of the purging of the voter rolls as required by a change in Bolivian elec-
toral laws. Data on turnout are from Corte Nacional Electoral (cne), Resultados Elecciones Generales y
de Prefectos 2005 (La Paz: cne, 2006).
Table 3
Predictors of Voting for the MAS (over PODEMOS)
in the 2005 Bolivian Election
(multinomial logit model)
Coefficient Standard Error P > |z|
Constant 3.722 .949 .000
Self-identifies as indigenous .561 .273 .040
Self-identifies as white –.676 .247 .006
Aymara maternal language 1.434 .285 .000
Quechua maternal language .637 .199 .001
Other indigenous maternal language .413 .548 .452
Prefers ethnic representation .299 .207 .149
Supports indigenous language education
(1–7 scale) .129 .056 .023
Trust in political parties (1–7 scale) –.042 .055 .448
Protested against Mesa administration .835 .283 .003
Left-right ideological self-placement
(left-right 1–10 scale) –.298 .041 .000
Support for nationalization of gas industry
(1–10 scale) .106 .037 .004
Participation in trade associations (1–4 scale) .268 .100 .007
Resides in media luna (southern and
eastern departments) –1.030 .193 .000
Urbanization level (1–4 scale) –.013 .078 .864
Monthly income (0–8 scale) –.142 .069 .038
Female –.016 .166 .921
Age (in years) .015 .006 .013
Pseudo R2 .184
N 1162
simulations using Clarify.65 The sloping lines in the figures represent

the predicted probabilities of voting for the mas (over podemos) and
the horizontal lines represent the 95 percent confidence intervals of
these estimates. The simulation in Figure 2 shows that someone who
strongly favored the nationalization of the gas industry had a 63 per-
cent likelihood of voting for the mas when other variables are held at
their means, whereas someone who strongly opposed it had only a 36
percent probability of doing so. The effects of ideological self-place-
ment were even stronger, as Figure 3 indicates. Someone who identi-
fied strongly with the left had a 78 percent likelihood of voting for the
65
Gary King, Michael Tomz, and Jason Wittenberg, “Making the Most of Statistical Analyses:
Improving Interpretation and Presentation,” American Journal of Political Science 44 (April 2000).
0.8
0.7
0.6
of Voting for the MAS

Predicted Probability 0.5
0.4
0.3
0.2
0.1
0
1 2 3 4 5 6 7 8 9 10
Degree of Support for Nationalization of Gas Industry
Figure 2
The Effect of Support for Nationalization of the Gas Industry on
Probability of Voting for the MAS
0.9
0.8
0.7
Predicted Probability
0.6
0.5
0.4
0.3
0.2
0.1
0
1 2 3 4 5 6 7 8 9 10
Left-Right Ideological Placement
Figure 3
The Effect of Ideology on the Predicted Probability of Voting
for the MAS
mas, whereas someone who identified strongly with the right had only
a 27 percent probability of doing so. Meanwhile, voters who partici-
pated in protests against the administration of Carlos Mesa (2003–5)
had a 70 percent probability of voting for the mas when all other vari-
ables are held at their means, as opposed to only a 53 percent prob-
ability for those voters who did not participate in any protests against
the Mesa administration. Somewhat surprisingly, voters who had little
trust in political parties were not more likely to vote for the mas, but
this may be because by 2006 the mas had established itself as the most
important political party in the country, causing the level of trust in
political parties among supporters of the mas to increase.66
Self-identifying as indigenous also increases the likelihood of vot-
ing for the mas, even after controlling for ideology, dissatisfaction with
parties, participation in protests, support for nationalization of the gas
industry, and a host of other variables.67 This suggests that people did
not vote for the mas just because of the party’s populist platform, but
did so also because the party’s ethnic demands and profile presumably
appealed to them as indigenous people. The dual ethnic and popu-
list appeal of the mas helps explain why it fared significantly better in
the 2005 elections than traditional populist parties have fared in recent
elections. Indeed, while populist parties such as the nfr and ucs man-
aged to win approximately 20 percent of the vote in some elections
during the 1990s and early 2000s, none of them approached the 53
percent of the vote that mas captured in 2005.
The analysis also found that people who self-identify as white were
less likely to vote for the mas, even controlling for other variables.
However, the probability that a self-identified white person would vote
for the mas was still relatively high, other things being equal. A simu-
lation, depicted in Figure 4, found that when all other variables are
held at their means people who self-identify as white had a 42 percent
probability of voting for the mas, people who self-identify as mestizo
had a 57 percent probability of doing so, and people who self-identify
as indigenous had a 63 percent probability of voting for the mas.68 This
66
According to the 2002 lapop survey, mas supporters in 2002 were more likely to express low
levels of confidence in parties, presumably because in 2002 mas had not yet established itself as one of
the country’s main parties. In 2002, 38.8 percent of mas supporters reported having no trust in parties,
as opposed to 29.2 percent of all voters.
67
The modest level of statistical significance of the indigenous identification variable is presum-
ably the result of the fact that indigenous self-identification is correlated with other variables in the
analysis, such as the Aymara and Quechua linguistic variables. The vast majority (84 percent) of people
who self-identify as indigenous grew up speaking an indigenous language.
68
As the figure shows, the 95 percent confidence intervals of the estimates for self-identified indig-
enous people overlap with those of mestizos, so we do not have a high level of certainty that someone
0.8
0.7

0.6
Predicted Probability 0.5
0.4
0.3
0.2
0.1
0
Indigenous Mestizo White
Ethnic Self-Identification
Figure 4
The Effect of Ethnic Self-Identification on the Probability of
Voting for the MAS
suggests that the mas’s ethnopopulist appeals resonated particularly

strongly among self-identified indigenous people but that they were
inclusive enough to attract many mestizos and whites.
Speaking an indigenous language also increased the likelihood of
voting for the mas, even controlling for indigenous self-identification
and other variables. The variables measuring whether the respondent
grew up speaking Aymara or Quechua were both positive and highly
statistically significant, although the variable for a lowlands indigenous
language was not.69 As Figure 5 shows, with all other variables held
at their means, people who grew up speaking Aymara had a 69 per-
cent probability of voting for the mas, people who grew up speaking
Quechua had a 60 percent probability of doing so, and people who did
not grow up speaking an indigenous language had only a 45 percent
probability of supporting the mas. Why does speaking an indigenous
language have such a strong effect on the likelihood of voting for the
who self-identifies as indigenous is more likely to vote for the mas than is someone who self-identifies
as mestizo.
69
Speakers of lowlands indigenous languages may not have been significantly more likely to vote
for the mas because it may have been perceived as a party that principally represented highlands indig-
enous populations, reflecting ongoing Amazonian-highlands indigenous divides.
0.8
0.7

Predicted Probability
0.6
0.5
0.4
0.3
0.2
0.1
0
Aymara Quechua None
Indigenous Language Spoken in Home as a Child
Figure 5
The Effect of Maternal Language on the Probability of Voting
for the MAS
mas even after controlling for indigenous self-identification?70 The

main reason is that the mas’s inclusive ethnic appeal managed to win
the support not only of people who self-identified as indigenous but
also of many people who grew up speaking an indigenous language but
do not self-identify as indigenous. These indigenous mestizos may not
self-identify as indigenous, but their indigenous roots, their cultural at-
tachments, and their experiences with socioeconomic disadvantage and
discrimination made them receptive to the inclusive ethnic profile and
agenda of the mas.
The analysis also found that views on ethnic issues had only a mod-
est effect on the likelihood of voting for the mas, presumably because
the mas’s inclusive appeal attracted many people who did not have
a strong position on these issues. The ethnic representation variable
had the expected sign but it was not statistically significant, suggest-
ing that people who felt better represented by members of their own
ethnic group were not significantly more likely to vote for the mas. The
support for teaching indigenous languages variable, meanwhile, was
70
I view growing up speaking an indigenous language as a reasonable proxy for having indigenous
roots and the cultural attachments and life experiences that go with them. It is these cultural attach-
ments and life experiences, I assume, that draw indigenous language speakers to the mas.
positive and statistically significant, but a simulation revealed that we

cannot be sure (at the 95 percent confidence level) that someone who
strongly advocated that schools teach indigenous languages was more
likely to support the mas than someone who did not support education
in indigenous languages at all.
Taken as a whole, these findings provide support for the argument
that the mas’s inclusive ethnopopulist appeal played a crucial role in
its electoral victory, attracting the support of a broad variety of vot-
ers with indigenous backgrounds as well as of politically disenchanted
voters and of voters with leftist and nationalist views. The statistical
analysis, however, also showed that a number of other variables, not
directly related to the mas’s ethnopopulist appeal, influenced the likeli-
hood that a voter would support the mas. Other things being equal,
voters from the eastern and southern departments known as the media
luna were much less likely to support the mas. By contrast, older voters,
people with lower incomes, and people who participated in trade asso-
ciations (that is, organizations of professionals, merchants, peasants, or
producers) were significantly more likely to support the mas. Surpris-
ingly, however, neither urbanization level nor gender had a statistically
significant relationship with voting for the mas, once other variables
were controlled for.
Ethnopopulism in Comparative Perspective

The mas is only one of the many left-leaning governments that have
come to power in Latin America in recent years. In nearly a dozen
countries, from Central America to the Southern Cone, leftist lead-
ers have won important electoral victories. The causes of this resur-
gence are many of the same ones that propelled the mas to victory
in Bolivia: disenchantment with the existing parties and their policies,
especially their market-oriented economic policies.71 Market-oriented
policies have succeeded in reducing inflation and restoring foreign fi-
nancial flows, but they have typically failed to generate strong and sta-
ble economic growth and they have not resolved pressing problems of
inequality, crime, and, unemployment. Moreover, recurring problems
with corruption, legislative gridlock, and poor leadership have led to
growing disaffection with the existing political elites in many countries,
which leftist candidates have used to their advantage.
71
Matthew R. Cleary, “Explaining the Left’s Resurgence,” Journal of Democracy 17 (October
2006).
As many scholars have observed, however, the leftist leaders that

have come to power in recent years are not all alike.72 Some of them,
such as Hugo Chávez in Venezuela and Rafael Correa in Ecuador, have
taken on a populist bent, developing strong antiestablishment agen-
das and seeking a radical overhaul of the existing economic and politi-
cal order. Others, like Ricardo Lagos and Michelle Bachelet in Chile
and Luiz Inácio Lula da Silva in Brazil, have embraced the existing
economic and political institutions. These latter leaders have sought
increased income redistribution but they have advocated only modest
changes in the market-oriented economic model. The ethnically based
left-leaning parties, such as the mas in Bolivia and Pachakutik in Ecua-
dor, have fallen clearly on the populist side. These parties represent the
most marginalized and disadvantaged sectors of the population and,
as a result, they have the lowest stake in the existing system. The lead-
ers of these parties, moreover, are typically political outsiders with few
links to the political and economic establishment and thus they have
been more disposed to make classically populist antisystem appeals.
Nevertheless, the ethnopopulist parties differ from classical populist
and leftist parties in that their ethnic linkages to voters provide them
with certain advantages. First, these linkages have enabled them to win
overwhelming support from a constituency that traditional leftist and
populist movements have had a difficult time penetrating: the rural,
mostly indigenous peasantry.73 This has enabled the ethnopopulist
parties to make dramatic gains in short periods of time. In Ecuador,
Pachakutik won 20.6 percent of the valid vote in 1996, the first time it
competed in presidential elections. The rise of the mas in Bolivia was
even more spectacular. The precursor of the mas, Izquierda Unida, won
only 3.5 percent of the valid vote in 1997, the first time it competed
in presidential elections, but the mas won 20.9 percent of the valid
vote in 2002 and 50.3 percent in 2005. By contrast, most other leftist
parties that have registered important victories in recent years, such as
the Workers’ Party in Brazil, the Broad Front in Uruguay, the Socialist
Party in Chile, and the Sandinistas in Nicaragua, have built themselves
up only gradually.
72
Teodoro Petkoff, “Las Dos Izquierdas,” Nueva Sociedad 197 (2005); Jorge Castañeda, “Latin
America’s Left Turn,” Foreign Affairs 85 (May–June 2006).
73
It is true that some populist and leftist parties, such as Izquierda Democrática in Ecuador, Izqui-
erda Unida in Peru, and the Unión Democrática y Popular in Bolivia, have won significant levels of
support among rural, indigenous voters in the past, but they have not captured this constituency to the
same degree as the ethnopopulist parties. In Bolivia, only the mnr of the 1950s and 1960s rivaled the
mas in terms of its share of the vote among the rural, indigenous population.
Second, the ethnopopulist parties may also prove to be more stable

than classical populist parties. Traditional populist leaders often failed
to build organizations, relying instead on unmediated, personalistic
linkages to voters that have proved fragile, particularly in the event of
the death of the leaders.74 Support for populist leaders has often de-
pended in large part on their ability to deliver economic and social
goods, and when they have failed to do so, volatile urban constituencies
have been quick to abandon them.75 The identity-based linkages that
ethnopopulist parties enjoy with some portions of the electorate should
be more enduring than these personalistic or performance-based link-
ages.76 Indeed, numerous studies have found that ethnic diversity tends
to dampen electoral volatility in part because of the resilience of ties
between ethnic parties and the people whom they were established to
represent.77 Ethnopopulist parties may also benefit from their close ties
to the indigenous organizations, which have deep roots in many rural
communities and should provide the parties with some organizational
stability. For the ethnopopulist parties to endure, however, they will
need to avoid ethnonationalist appeals and reconcile the interests of
their rural indigenous core and the urban mestizo sectors that have
also supported the parties. To date, the mas has demonstrated a good
deal of electoral stability, successfully maintaining its rural indigenous
base while making important inroads in other sectors of the popula-
tion. During its first decade, Pachakutik also demonstrated a remark-
able degree of stability for an Ecuadorian party, but, as we shall see, it
has recently begun to disintegrate.
Ethnopopulist parties have fared well in recent years in part because
the sociopolitical climate has been conducive to ethnic as well as popu-
list appeals. Rising levels of indigenous mobilization and heightened
ethnic consciousness have politicized many indigenous voters in Latin
74
Weyland (fn. 8), 13.
75
Ibid.; Conniff (fn. 8), 20; Drake (fn. 8); Robert R. Kaufman and Barbara Stallings, “The Politi-
cal Economy of Latin American Populism,” in Rudiger Dornbusch and Sebastian Edwards, eds., The
Macroeconomics of Populism in Latin America (Chicago: University of Chicago Press, 1991).
76
The stability of ethnic ties should not be exaggerated, however, particularly in regions such as
Latin America, where ethnic identification is fluid. As the Ecuadorian case illustrates, indigenous vot-
ers may abandon an indigenous-based party for other parties if its appeals become exclusionary or if
they believe that other populist or leftist parties have better electoral possibilities.
77
Stefano Bartolini and Peter Mair, Identity, Competition, and Electoral Availability: The Stabilisa-
tion of European Electorates 1885–1985 (New York: Cambridge University Press, 1990); and Jóhanna
Kristín Birnir (fn. 4, 2001). In an earlier study, I found that indigenous areas in Latin America have
traditionally been more electorally volatile than nonindigenous areas, but I argued that this volatil-
ity stemmed from the failure of existing parties to represent the indigenous population adequately.
Indeed, the emergence of indigenous-based parties has reduced electoral volatility to date. See Raúl
Madrid, “Ethnic Cleavages and Electoral Volatility in Latin America,” Comparative Politics 38 (Oc-
tober 2005).
America at the same time that economic stagnation and inequality

have led many voters of all ethnic backgrounds to embrace populist
policies and leaders. Indeed, it is quite likely that ethnopopulist parties
would not have been nearly as successful as the mas or Pachakutik had
they emerged in Bolivia or Ecuador during earlier periods.
Nevertheless, even in the last decade, the level of success enjoyed by
ethnopopulist parties in Latin America has varied considerably from
country to country. Not surprisingly, the ethnopopulist parties have
tended to fare better in those countries with substantial indigenous
populations in large part because indigenous people tend to vote for
these parties at higher rates.78 Bolivia, which by most accounts has the
largest proportion of indigenous people in the Americas, has had by
far the most successful ethnopopulist party to date. An ethnopopulist
party also fared well for a time in Ecuador, which has a reasonably large
indigenous population, although the exact size of this population is
the subject of great debate.79 By contrast, ethnopopulist parties have
not fared particularly well in Colombia and Venezuela, in large part
because these nations have only small indigenous populations.
The strength of a country’s indigenous movement has also affected
the performance of the ethnopopulist parties.80 Strong indigenous
movements have provided useful resources to ethnopopulist parties in
Bolivia and Ecuador, including candidates, activists, and organizational
legitimacy. Where the indigenous movement is relatively weak or
highly fragmented, ethnopopulist parties have had a harder time tak-
ing root.81 According to the Guatemalan anthropologist Ricardo Falla,
the most important reason that the 2007 presidential candidacy of in-
digenous activist Rigoberta Menchú failed to catch on in Guatemala
was that “unlike Evo Morales in Bolivia, Rigoberta is not backed by a
mass organization or social movement. Nor can she count on a party
structure that covers the country’s many indigenous regions, much less
78
Mijeski and Beck (fn. 6); Raúl L. Madrid, “Indigenous Voters and Party System Fragmentation
in Latin America,” Electoral Studies 24 (December 2005); Roberta Rice and Donna Lee Van Cott,
“The Emergence and Performance of Indigenous Peoples’ Parties in South America: A Subnational
Statistical Analysis,” Comparative Political Studies 39 (August 2006).
79
Some scholars have claimed the indigenous population in Ecuador represents 30 percent or more
of the population, but the 2001 census found that only 6.1 percent of the population self-identified
as indigenous and only 4.6 percent reported speaking an indigenous language. The 2000 emedinho
survey found that 14.3 percent of the population either self-identified as indigenous, spoke an in-
digenous language, or had parents who spoke an indigenous language. See Mauricio León Guzmán,
“Etnicidad y exclusión en el Ecuador: Una mirada a partir del Censo de Población de 2001,” Iconos
(February 2003).
80
Van Cott (fn. 4, 2005); and Marenghi and Alcántara (fn. 5).
81
Van Cott (fn. 4, 2005).
the country as a whole.”82 Menchú also did not articulate a clear leftist or
populist agenda and chose as her running mate a wealthy businessman and
former president of the country’s most important business organization.
The absence of a strong indigenous movement in Peru has also ham-
strung ethnopopulist parties in that country.83 A number of candidates
and parties seeking to represent the indigenous population have emerged
in Peru, but these candidates have fared poorly in part because they have
not had the legitimacy or organizational resources that a strong indig-
enous movement might provide them. Ollanta Humala of the Partido
Nacionalista Peruano (pnp) did fare quite well in the 2006 Peruvian
presidential elections, particularly in indigenous areas, but Peru’s weak
and divided indigenous movement played no significant role in his
campaign. Moreover, although Humala may have benefited from his
Quechua name and the ethnic organizing of some of his family mem-
bers, he made only limited ethnic appeals, so it is more accurate to
describe his movement as populist, rather than ethnopopulist.
Finally, as we have seen, those ethnically based parties that have de-
veloped inclusive appeals have been much more successful than those
parties that have been more exclusionary. Indeed, the mas’s electoral
performance improved considerably over time as it became more inclu-
sive. By contrast, the main ethnically based party in Ecuador, Pachaku-
tik, became less inclusive over time, with negative effects on the party’s
fortunes. The Ecuadoran indigenous movement played the lead role
in the founding of Pachakutik, but at the outset Pachakutik had nu-
merous white and mestizo as well as indigenous leaders.84 Pachakutik
also initially established close ties to various mestizo-dominated orga-
nizations. In the elections of 1996 and 1998, for example, Pachaku-
tik formed an alliance with the urban-based Movimiento Ciudadano
Nuevo País and supported this movement’s mestizo leader, Freddy
Ehlers, for president. In 2002 Pachakutik allied with Lucio Gutiérrez,
a mestizo army colonel, and his party, the Partido Sociedad Patriótica.
This inclusive approach attracted many white and mestizo voters, as
well as indigenous people.85 As a result, the candidates Pachakutik sup-
ported won more than 20 percent of the vote on average in the first
round of presidential elections between 1996 and 2002, and the party
helped deliver the presidency to Gutiérrez in 2002.
82
Ricardo Falla, “Rigoberta Menchú: A Shooting Star in the Electoral Sky,” Revista Envio 312
( July 2007), http://www.envio.org.ni/articulo/3606 (accessed November 27, 2007).
83
See Yashar (fn. 3) for an illuminating analysis of why Peru has failed to develop a strong indig-
enous movement.
84
Andolina (fn. 5); and Collins (fn. 4).
85
Mijeski and Beck (fn. 6); Beck and Mijeski (fn. 6); and Madrid (fn. 78).
In 2006, however, the party shifted in a more ethnonationalist di-

rection in large part because many of its indigenous leaders grew tired
of ceding important candidatures to whites and mestizos. Pachakutik
declined to form alliances with other parties and opted to run its own
candidate for president, the indigenous leader Luis Macas. The party
also put forth fewer mestizo candidates for other positions than it had
in the past, and some prominent mestizo leaders, such as Augusto Bar-
rera and Virgilio Hernández, left the party because of their concerns
about its growing ethnocentrism.86 Many of the mestizo-dominated
unions and organizations that had supported Pachakutik in past elec-
tions—such as the Urban Forum, and the unions of teachers, oil work-
ers, and electrical workers—also abandoned it in 2006 in part because
of concerns about the growing dominance of the indigenista faction
within Pachakutik.87 Thus, Pachakutik increasingly resembled a tradi-
tional ethnic party rather than an ethnopopulist party. Partly as a result,
it fared poorly in the 2006 presidential elections, winning a mere 2
percent of the vote.88 Not only did white and mestizo voters abandon it
en masse, but so did many indigenous voters.89 To be sure, Pachakutik’s
ethnonationalist turn was not the only factor in its decline—stiff com-
petition from other populist parties also played an important role—but
its increasingly exclusionary policies hurt the party considerably.
Conclusion
The existing literature on ethnic parties has maintained that such par-
ties succeed by mobilizing their base through exclusionary appeals.
This article has shown that such exclusionary appeals are unlikely to be
successful in areas where ethnic polarization is low and ethnic identities
are fluid and multiple. In these areas, ethnically based parties may win
support from members of a variety of different ethnic groups by develop-
86
“Pachakutik Pierde su Fuerza Urbana.” El Comercio, December 14, 2005, www.elcomercio.com/
solo_texto_search.asp?id_noticia=9962&anio=2005&mes=12&dia=14 (accessed August 5, 2007);
“Pachakutik se Requesbraja por el Indigenismo,” El Comercio, December 15, 2005, www.elcomer-
cio.com/solo_texto_search.asp?id_noticia=10072&anio=2005&mes=12&dia=15 (accessed August 5,
2007).
87
Scott H. Beck and Kenneth J. Mijeski, “How to Lose by Winning: The Ecuadorian Indigenous
Movement after the 2002 Elections” (Paper presented at the international congress of the Latin Amer-
ican Studies Association, San Juan, Puerto Rico, March 15–18), 17; “La desintegración de Pachakutik
continua.” El Comercio, December 23, 2005, www.elcomercio.com/solo_texto_search.asp?id_noticia=1
0889&anio=2005&mes=12&dia=23 (accessed August 5, 2007).
88
Pachakutik fared even worse in the September 2007 elections for the constituent assembly, win-
ning less than 1 percent of the national vote.
89
Sara Báez Rivera and Víctor Bretón Solo de Zaldívar, “El Enigma del Voto Étnico o las Tribula-
ciones del Movimiento Indígena,” Ecuador Debate 69 (December 2006).
ing inclusive appeals. Indeed, this is exactly the approach that the mas
and, initially, Pachakutik used successfully in Bolivia and Ecuador.
These findings suggest that the literature on ethnic parties needs to
consider how ethnic identities and relations are constructed in theoriz-
ing about the electoral behavior of such parties. For example, the litera-
ture needs to recognize not only that ethnic identification is often fluid
and multiple but also that the degree of fluidity and multiplicity often
varies considerably over time and space and that this has important
electoral implications. It may well be that ethnic parties have incentives
to adopt exclusionary appeals where ethnic identities are singular and
relatively fixed, but ethnically based parties have greater incentives to
be inclusive in societies where ethnic identification is fluid and mul-
tiple. This study thus aims to bring the complexity of ethnic identity
back into the study of ethnic parties.
This article also aims to bring ethnicity into the study of populism.
The literature on populism has largely failed to recognize the poten-
tial compatibility between populist and ethnic appeals. In ethnically
homogeneous countries, populist parties may not need to make ethnic
appeals to win votes among the subaltern sectors, but in countries with
large indigenous populations or other socioeconomically disadvantaged
ethnic groups, populist parties may more effectively win lower-class
votes through a combination of populist and inclusive ethnic appeals.
Ethnic appeals, moreover, may create more enduring ties to voters than
personalistic or performance-based appeals, thus enabling ethnopop-
ulist parties to achieve a degree of electoral stability that traditional
populist parties have found elusive.

The Rise of Ethnopopulism in Latin America: Raúl L. Madrid

Uploaded by

Copyright:

Available Formats

The Rise of Ethnopopulism in Latin America: Raúl L. Madrid

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Rise of Ethnopopulism in Latin America: Raúl L. Madrid

Uploaded by

Copyright:

Available Formats

The Rise of Ethnopopulism in Latin America

Published by The Johns Hopkins University Press

For additional information about this article

Access Provided by University of Notre Dame at 04/06/11 1:44PM GMT

World Politics 60 (April 2008), 475–508

Morales in Bolivia, for example, has led to state intervention in the

with a multinomial logit analysis of public opinion data on the 2005

Explaining the Rise of Ethnopopulist Parties

who do not self-identify as indigenous but who nevertheless sympa-

on their rural indigenous core. As noted above, ethnopopulist parties

Institutional Explanations for the Rise of the MAS

Cochabamba.25 It therefore seems unlikely that the mas’s strong elec-

The Inclusive Ethnic Appeal of the MAS

The mas’s appeal to people with indigenous backgrounds had a

Other Katarista parties, such as the Movimiento Revolucionario Tu-

elsewhere. In the early 2000s, however, the mas sought to diversify

Similarly, in a 2007 interview with the author, Ricardo Díaz, a senator

The mas’s candidate strategies and ethnic appeals varied somewhat

Source: lapop 2006 Bolivia survey.

The MAS’s Populist Appeal

The mas made control of Bolivia’s considerable natural gas depos-

as inappropriate interference in their internal affairs, Morales’s support

An Analysis of the MAS’s 2005 Vote

simulations using Clarify.65 The sloping lines in the figures represent

of Voting for the MAS

of Voting for the MAS

suggests that the mas’s ethnopopulist appeals resonated particularly

of Voting for the MAS

mas even after controlling for indigenous self-identification?70 The

positive and statistically significant, but a simulation revealed that we

Ethnopopulism in Comparative Perspective

As many scholars have observed, however, the leftist leaders that

Second, the ethnopopulist parties may also prove to be more stable

America at the same time that economic stagnation and inequality

In 2006, however, the party shifted in a more ethnonationalist di-

You might also like