![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
ABSTRACT
The 2012 Bolivian census produced a surprising result: 41% of the adult population identified themselves as belonging to an indigenous group. This represents a dramatic decline in the indigenous population compared with the 62% that was registered in the previous 2001 census. This turnaround is especially surprising considering the election of Bolivia’s first indigenous president in 2006 and the implementation of a series of policies granting indigenous groups particular rights and benefits, which would have tended to increase indigenous identity. Using survey data from between 2004 and 2014 from LAPOP’s AmericasBarometer, this paper shows that the primary reason for this change is a difference in the instrument used for identifying the indigenous population: the self-identification question employed in the Bolivian censuses. The paper also explores socioeconomic factors associated with identification as indigenous and with other relevant ethnic categories in Bolivia, finding evidence suggesting that part of this identification is based on political grounds, and providing empirical elements that could contribute to the debate on how to measure ethnic identity in census and surveys.
Acknowledgements
An earlier version of this paper was presented at the Latin American Studies Association Congress held in Puerto Rico in 2015. The paper benefitted greatly from generous comments by Edward Telles. At the review process at LACES, the paper benefitted from generous comments by three anonymous reviewers and the editor. The work also received valuable inputs from Gonzalo Vargas and Daniela Osorio at Ciudadanía, with whom the author teamed at an early stage for a research project funded by PIEB, Programa de Investigación Estratégica en Bolivia, through the call for research proposals on the Bolivian Nation, 2013–2014. I also thank the Latin American Public Opinion Project (LAPOP) and its major supporters for making the data used in this paper available.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. The inequality argument, which demonstrates that the socioeconomic and political conditions of indigenous were substantially inferior to the rest of the national population (Hall and Patrinos Citation2006; Pascharopoulos and Patrinos Citation1994), was, of course, also taken into account in the debate.
2. For a wide description of the Constituent process in Bolivia see, among others, Shavelson (Citation2012). For a description of the Bolivian political process and the forces that led to the Constitutional Assembly see Mayorga (Citation2014, Citation2011).
3. Former President Carlos Mesa (see his opinion piece on Pagina Siete, 10 August 2013) is a visible member of the group of personalities claiming this, which also includes Jorge Lazarte (Página Siete, 8 August 2013) and Carlos Toranzo. For a wider discussion, see the debate on national newspapers in August including opinion articles from Fernando Mayorga (La Razón, 18 August 2013) and Jorge Komadina (La Razón, 29 September 2013).
4. The criticism over technical difficulties in the conduction of the 2012 census even led to the resignation of INE’s national director in August 2013.
5. The constructivist approach in the research on ethnic identities conducted by different authors has shown convincingly that identities do change within some limits across time as a result of social and political processes. For instance, see Chandra (Citation2006); Chandra and Laitin (Citation2002); Hoddie (Citation2006). For the Bolivian case, see Moreno Morales (Citation2008); Moreno Morales et al. (Citation2008); Moreno Morales, Vargas, and Osorio (Citation2014); also Burman (Citation2014).
6. Several works illustrate the process of de-indigenization of different countries in the Americas. For instance, see the work of de la Cadena in Peru or Adams and others in Guatemala (Adams and Bastos Citation2003; Cojti Citation1998; De la Cadena Citation2000; Hale Citation2005).
7. Leonardo Tamburini (CEJIS, 15 October 2013) and Pablo Stefanoni (Página Siete, 5 August 2013) are among the few that point to the question employed as a possible cause in the decrease of indigenous population between the two censuses.
8. Some of these findings were previously published locally in Bolivia, see Moreno Morales (Citation2013); Moreno Morales, Vargas, and Osorio (Citation2014).
9. Sampling design in the LAPOP surveys emphasizes unbiasedness as a result of a probability sample that randomly selects the homes where the interviews are conducted using a multi-stage sampling strategy. LAPOP samples in Bolivia until 2014 were not only designed to represent national averages, but also regional means, using each of the country’s nine departments as sampling strata, and considering also the urban and rural distribution of the population within strata. The fact that interviews are conducted all over the country using a relatively large and disperse sample guarantees that no particular group of the national population, particularly one with a relatively large share of the population as those belonging to an indigenous group, is excluded. For more information on the LAPOP surveys visitThe AmericasBarometer by the Latin American Public Opinion Project (LAPOP): http://www.lapopsurveys.org.
10. The biannual results are presented by Ciudadanía and LAPOP in their periodic report on the political culture of democracy in Bolivia; these reports discuss indigenous identity after the publication of the initial 2001 census. Reports can be downloaded from: http://www.ciudadaniabolivia.org.
11. Responses include geographical locations (such as la cumbre); names of towns (pueblos in Spanish) such as Punata or Vallegrande; non-indigenous identity categories (camba); regional belonging categories (potosino), and many other possibilities that could be hardly classified as ‘indigenous‘ but are offered by respondents when asked this specific question. For more information on the type of responses offered by interviewees and on the way responses to this question are coded and processed see Molina and Albó. (Citation2006); Moreno Morales (Citation2014).
12. The inclusion of this question in the 2001 census had the political intention to officially register, for the first time ever, identification with an indigenous group; this is need is consistent with the official recognition of the multi-cultural nature of the country in the 1994 Constitutional amendment. The item was not intended to be an exhaustive ethnic identification question (and that is why it does not offer any non-indigenous options) and it is not a question about racial self-identification either; this question was designed to identify the indigenous population and not to ethnically or racially classify the Bolivian population in a set of relevant categories.
13. The author of this paper served as a consultant for the Bolivian government on how to identify the indigenous population in the 2012 census and directly witnessed some of the tensions and negotiation processes that finally led to the use of this particular wording.
14. We know that changes in the wording of questions and in the way in which they are administered, particularly whether they are filtered or not, have large effects on survey results (Krosnick Citation2013; Tourangeau, Rips, and Rasinski Citation2000); this methodological difference is at the center of the differing results between the two censuses in Bolivia.
15. The Quechua category is the one that shows the largest absolute reduction between 2001 and 2012, more than a quarter million individuals. More than an indigenous nation, the Quechua group can be understood as a linguistic community whose members’ ancestors belonged to different peoples and voluntarily adopted the Quechua language or were forced to do so by the Inca empire or the Spanish colonial rule. Conversely, the Aymara group shows very little variation between 2001 and 2012. This has to do with the fact that the Aymara identity is more strictly based on language and origin, and that members of this group tend to marry non-Aymara less frequently than members of other indigenous groups.
16. Needless to say, using census data at the individual level would be ideal for this purpose. Unfortunately, census data at the individual level are not made available by the INE, so this research strategy cannot be put in practice.
17. There is one caveat that should be mentioned here. This methodological decision implies the assumption that there is some level of homogeneity in the factors related to the self-identification of individuals with the different categories that define the larger ‘indigenous‘ group. However, we know that there are relevant differences in the factors associated with these categories, particularly the Aymara category, for which racial and linguistic affiliation seem to be a stronger component (Moreno Morales, Vargas, and Osorio Citation2014). All categories are grouped and treated together in this paper for simplicity reasons, assuming that differences in the relationship between the categories and individual variables are mitigated in the pooled sample, showing only results that are particularly robust.
18. Recording the respondent’s skin color is a measure introduced in the LAPOP surveys since 2010 by Edward Telles and the PERLA team at Princeton. This measure of a person’s skin tone has proven to be very useful explaining phenomena such as discrimination and social stratification based on racial differences, and also for understanding how racial characteristics are associated with identities, see Telles and Paschel (Citation2014); Telles (Citation2014).
19. CONAMAQ is the Consejo Nacional de Ayllus y Markas del Qollasuyo, the organization representing most indigenous nations in the highlands. CIDOB, the Confederación Indígena del Oriente Boliviano, unites most lowland indigenous peoples.
20. In 2010 the national government announced plans to build a road crossing the heart of the Territorio Indígena y Parque Nacional Isiboro Sécure, an amazonic jungle area granted protected status by being both a national park and an indigenous territory. This decision was resisted by indigenous organizations and environmental groups, who finally won the dispute and at least temporarily stopped the construction.
21. Virtually all surveys that ask the respondents to choose between these categories have resulted in a majority of the population identifying as indigenous. See, for instance, the surveys conducted by Fundación UNIR (Citation2008) and UNDP (Calderón and Toranzo Citation1996).
22. This finding becomes clear when the meaning of the questions is discussed with respondents. Qualitative research shows that individuals tend to associate ‘mestizo‘ with race, while they are willing to grant belonging to an indigenous group a broader, more cultural meaning (Moreno Morales, Vargas, and Osorio Citation2014).
Additional information
Notes on contributors
Daniel E. Moreno Morales
Daniel E. Moreno Morales is a Bolivian sociologist with a PhD in Political Science from Vanderbilt University (2008). He is currently Executive Director at Ciudadanía, Comunidad de Estudios Sociales y Acción Pública, a local research NGO in Bolivia (www.ciudadaniabolivia.org). Has a methodological expertise in public opinion research and has worked with the main comparative surveys in Latin America. Among his recent publications is ‘Nación, diversidad e identidad en el marco del estado plurinacional’ (with G. Vargas and D. Osorio), PIEB 2014.