On the Link Between Ethnic Politics and Identification: Lessons from Bolivia Paper prepared for the Political Studies Association International Conference 2013 Draft, please do not cite without permission. Comments welcome. Anaïd Flesken Institute of Latin American Studies, German Institute of Global and Area Studies [email protected] Abstract The politicization of ethnic diversity has long been regarded as perilous to ethnic peace and national unity, its detrimental impact memorably illustrated in Northern Ireland or Rwanda. The process of indigenous mobilization in Bolivia over the past decade has hence been seen with some concern by observers in policy and academia alike. This paper presents an analysis of the link between ethnic politics and identification in Bolivia by, first, qualitatively examining the development of political identity discourses and, second, quantitatively examining two dimensions of indigenous identification through time-series survey data. The analyses show that ethnic identifications do indeed change with changing political discourses, that they do so more quickly than expected, and not necessarily in the manner as expected: the type of reaction in identification depends on the type of dominant discourse, demonstrating that it is necessary to distinguish different elements of ethnic identification. 1 Introduction The politicization of ethnicity has long been regarded as perilous to ethnic peace and national unity, its detrimental impact memorably illustrated in Northern Ireland, former Yugoslavia, or Rwanda. Ethnic identities, so the common assertion, may be the basis of political mobilization of their group members, resulting in more salient identification with the in-group, in an ever more entrenched hatred for members of the out-group and thus, potentially, in conflict (see also Davis and Brown 2002; Gibson 2006). However, the causal processes as such have rarely been studied directly (see also Fearon and Laitin 2000; Green and Seher 2003). Instead, inferences about individuals – those who purportedly identify with their in-group and harbor ethnic hatred – are often made on the basis of observations at the aggregate level, such as the occurrence of violence within a country (e.g. Cederman, Wimmer, and Min 2010; Gurr 1993; Rabushka and Shepsle 1972; see also Joras and Schetter 2004). Such analyses may not only lead to ecological fallacies, overestimating the relationship at the individual level, but also to biases in theories on the causal mechanisms linking ethnicity and conflict. This paper addresses this gap, analyzing the link between the politicization of ethnicity at the macro- level and ethnic identification at the micro-level for the case of Bolivia. Bolivia provides the setting for a particularly interesting development of a process of ‘reindianization’ which has recently swept the Latin American continent; the emergence and evolution of new cultural and political indigenous discourses (Jackson and Warren 2005, 549). Long marginalized in both the social and political spheres, indigenous Bolivians began to mobilize in social movements in the 1970s. The mobilizations have hence given rise to indigenous-based parties in the 1990s, which successfully contested in local, regional, and national elections. Between 2000 and 2005, they contributed to the ousting of two presidents and today support the first indigenous president of the country, Evo Morales Ayma. His presidency saw the introduction of policies designed to end indigenous marginalization, including the nationalization of the country’s gas resources, the passing of an anti-discrimination law, and, perhaps most importantly, the passing of a new constitution in 2009. This constitution formally established Bolivia as Plurinational State of Bolivia, in special recognition of the country’s thirty-six indigenous nations as well as its Afro-descendant population. This paper presents a systematic analysis of the development of ethnic identifications in Bolivia over the past decade, examining the extent to which identifications have been affected by ethnic politics. To this aim, both the political discourses as well as ethnic identifications are examined. First, drawing on mainly anthropological accounts as well as expert interviews, I map out the development of political identity discourses and identify three strands: i) a rising salience of indigeneity and its inclusion into the notion of Bolivian national identity or ‘Bolivianhood’, particularly during the so-called protest cycle from 2000 to 2005; ii) a rising assertiveness of indigeneity, specifically surrounding the election and inauguration of Morales in 2005 and 2006; and iii) rising tensions between different 2 interpretations of indigeneity during the constitutional assembly, recently played out in the still- ongoing TIPNIS conflict. 1 The second part quantitatively examines indigenous identification in Bolivia, using representative survey data collected by the Latin American Public Opinion Project (LAPOP) over the past decade. 2 The analyses show changes in ethnic identifications that can be linked to the political identity discourse. They show that there has been indeed an increase in the number of people who identify as indigenous but that this increase may be due to the inclusive discourse during the protest cycle, rather than due to the rising assertiveness during Morales’ first years in presidency. Instead, the latter has given rise to an increased sense of cohesion among members of the diverse indigenous peoples as one community. Quarrels during the writing and implementation of the new constitution reversed this process, however, if they did not even further lessen this sense of community. Besides shedding light on the situation in Bolivia, this analysis thus also contributes to the wider debate on the effects of ethnic politics. It shows that ethnic identifications do indeed change with changing political discourses, that they do so quicker than expected, and not necessarily in the manner as expected: the type of reaction in identification depends on the type of dominant discourse, demonstrating that it is necessary to carefully distinguish different elements of ethnic identification in the analysis of ethnopolitics. Ethnic Groupness Today, scholars of ethnicity agree virtually unanimously that ethnic identities are not primordial and fixed but constructed, multiple, and situational (e.g. Chandra 2012; Eriksen 1993; Gil-White 1999; Hale 2004).3 Ethnic identities are constructed in that they are not determined by a structure essential to them but come into being in social interaction; they are multiple in that one individual may be, for example, Bavarian, German, and Jewish all at the same time; and they are situational in that one of these identities may be more salient to the individual in one situation than in another. One important impetus for the changing salience of ethnicity is said to be its political mobilization through ethnic entrepreneurs. In the struggle for popular support, politicians define or activate ethnic cleavages which, as a result, begin to structure not only political but also social relations (e.g. Chandra 1 TIPNIS is the Spanish acronym for Indigenous Territory and National Park Isiboro–Securé. Government plans for the construction of a road through the territory was protested in autumn 2011 with a 700km-long march from the city of Trinidad to the seat of government in La Paz. The protest march was met with incomprehension by government officials and violent action by the police. This reaction increased public support for the protesters and the lowland indigenous marchers were joined by other organizations (see Bolivia Information Forum 2011). The TIPNIS crisis was resolved in favour of the protesters but ongoing tensions gave rise to a second march to La Paz at the time of writing. 2 I thank the Latin American Public Opinion Project and its major supporters (the United States Agency for International Development, the United Nations Development Program, the Inter-American Development Bank, and Vanderbilt University) for making the data available. 3 They do not, however, agree on the potential for change of ethnic identities once constructed (see also Hale 2004). 3 2004; Horowitz 2000; Lake and Rothchild 1997; Rabushka and Shepsle 1972). Through politicization, categories do not only become more salient but also a source of identification and sense of community, if not of collective action and inter-group rivalry (see also Lieberman and Singh 2012). While this effect has sometimes been explained with reference to social psychological processes (e.g. Birnir 2007; Horowitz 2000), it has rarely been studied directly. Indeed, in spite of the widespread acceptance of the view that ethnicity is socially constructed, studies analyzing the effect of political processes on ethnicity – rather than vice versa – are scarce (see also Cooney 2009; Green and Seher 2003). It is the aim of this paper to analyze the effects of ethnic politicization on in-group processes. In doing so, the paper is based on the distinction – common in the theoretical, much less the empirical literature – between ethnic category and ethnic group. An ethnic category is defined by its attributes, such as a common language, culture, history, locality, and/or physical appearance. On the ground, these attributes are perceived to arise from a common descent of the members of each ethnic category. An ethnic group, on the other hand, is marked by cohesion and solidarity among