Social Exclusion in Peru: An Invisible Wall A Research Program on the Relationship between Ethnicity and Economic and Social Outcomes

Peru is a country were social exclusion is very profound but at the same time very subtle. Sociological and anthropological evidence suggest that there are several mechanisms through which ethnic and racial discrimination affect the lives of a large part of the population. However, precise measurements of the extent to which the potentially excluded groups are affected, are almost nonexistent. In Peru the majority of the urban population has a mixed background, and most of them will be mestizos. However, this mixed population is and is perceived as highly heterogeneous. Our initial research in this field has shown that even among mestizos, different racial backgrounds have important implications over socioeconomic outcomes, in particular over earnings. We know much less about the specific exclusion mechanisms. This proposal is based on the idea that there are social exclusion mechanisms related to ethnic and racial differences that have effects on the access and accumulation to different public and private assets, and that exclusion also affect the returns to some of those assets in the labor market, with crucial implications over poverty and well being. In this proposal we set the plan to continue the analysis of a data set constructed to approximate the ethnic heterogeneity in Peruvian urban areas. Our data includes self-reported discrimination events, race, mother tongue of the parents, language spoken at home and at school; and education and origin of the previous generation among others. We will analyze the relation between ethnic and racial heterogeneity and occupational segregation, access to education and to social networks. Additionally, we propose the collection and analysis of a new data set that will allow us to explore the extent to which there are exclusion mechanisms operating in the hiring process in the urban labor market. Social Exclusion: An Invisible Wall A Research Program on the Relationship between Ethnicity and Economic and Social Outcomes I. Background and Motivation Social exclusion in the form of racial or ethnic discrimination is pervasive in Peru, a highly multiracial and multilingual country. Poverty and economic and social wellbeing are clearly related to ethnicity, race and cultural traits. Indigenous are more likely to be poor than any other ethnic group. While overall poverty rate in Peru is 49%, the poverty rate of the indigenous population (those whose mother's tongue is Quechua, Aymara or other native language reach 70%. Moreover, more than 75% of the indigenous people can be found in the three bottom deciles of the income distribution. Although the government recognizes the problem, little has been done to alleviate it. In 1998 the Peruvian Congress passed the Law No. 26772 prohibiting discrimination practices in labor hiring and in the education admission processes. However, the lack of understanding of the channels through which effective discrimination prevails make this kind of rules little more than lip service. Social exclusion goes beyond poverty. Many groups in the society are excluded from markets for certain goods and services in a process that may affect economic outcomes through different channels. First, it may affect the access to public and private assets (human, physical, financial, or organizational capital). For example, more than 80% of nonindigenous people have access to public water supply or access to electricity while less than 45% of indigenous people have access to it. On average, non-indigenous individuals attain 8.1 years of schooling while the same figure for indigenous people is 5.5 years, just to mention two of the most striking differences (see MacIsaac, 1993). Second, it may affect the rate of return of those assets: for instance, similarly educated people may show differences in the economic returns they get from that education and certain groups do not have access to better paid jobs in a process that generates the high levels of market segmentation and occupational segregation observed in Peru. Many studies account for the various forms of social exclusion that take place in Peru. Gender discrimination has been studied in reference to the access of women to political leadership (Alfaro, 1996); in educational enrollment and attainment (Oliart, 1989; Rossetti, 1989; Guillén, Soto and Yáñez, 1996; and Mendoza, 1995) and in labor market performance: participation and wage differentials (Guzmán, 1987 and Saavedra, 1997). Ethnic discrimination has also been studied, usually through case studies and ethnographic studies. Callirgos (1993) gives a global overview of the origins and particular characteristics on the Peruvian racism. Oliart (1989), Pozzi (1989) and Mendoza (1993) tried to tackle ethnic and cultural discrimination. Finally, using case studies Sulmont (1995) has documented some of the elements of social exclusion that may be present in Peruvian labor markets. Despite the fact that racial and ethnic discrimination has been the focus of attention of numerous sociological and anthropological studies and debates, many unanswered questions remain, in particular regarding the economic effects of social exclusion. In an economy that is inevitable more and more integrated to the world, equality of opportunities will be crucial to assure a fair distribution of the benefits of economic and cultural integration. As we will see in this proposal and is suggested by a few studies, in Peru, racial and ethnic differences explain differences in access to public services, to education, health and to social networks. This in turn affect human capital accumulation and hampers social mobility prospects of individuals with specific backgrounds Despite the obvious importance of the topic for a country like Peru, there are very few data sources that can capture ethnic discrimination, and empirical work that tries to tackle exclusion and discrimination issues from a quantitative perspective is very scarce. Most of the work that has been done tries to approximate racial and ethnic discrimination with supposedly easily observable characteristics, in most of the cases mother tongue as in several of the World Bank studies. An example of this line of research is World Bank (1999) which shows that, after controlling for other covariates, the marginal effect of ethnic background over learning outcomes is significant. For instance, Quechua speaking students tend to do worse in school than Spanish speakers. Over-aged student rates -the percentage of students that are in a grade below from what is expected for their ageare considerably larger among indigenous population than among the native Spanish speaking population (Saavedra and Cárdenas, 2002). Gleewwe (1988) shows that, other things equal, the budget share of food and education is higher if the head of household has an indigenous background However, the approximation of ethnicity using mother tongue usually used in empirical work is clearly incomplete, as there are ethnic and racial differences within the Spanish and Quechua speaking populations. In this research program we take a step forward by approximating ethnicity using variables related to several dimensions of the concept, such as mother tongue, parental background, race, and religion. In Peru the long process of race mixing started centuries ago. The process of mestizaje started since the Spaniards arrived in the XVI century and continued also with the increase in black population and Asian immigration. As a consequence the majority of the urban population has a mixed background. This mixed population is and is perceived by it as highly heterogeneous. Even if most of the urban population is mestiza, some are predominantly Indian and some 1 In the anthropology literature, the concept of ethnicity is defined in a general way as the community of individuals that share cultural elements and that organize their daily life around these. Generally it is associated with the idea of native communities that are isolated or that have a reduced contact with other communities. In urban settings, ethnic characteristics are associated, in a complex and hotly debated relationship, with culture, religion, language, traditions and race, among other dimensions. 2 Other attempts to approximate ethnic characteristics are found in some household surveys. As an example, the last Living Standards Measurement Survey (LSMS 2000) inquires about racial characteristics. More than 98% of the rural population is self-reported as "Mestiza" (mixed race), and only a tiny percentage of the sample self-report themselves as Indigenous, Asian, Black or White. predominantly White. As our initial research in this field has shown, different racial backgrounds have implications over socioeconomic outcomes, for instance, in the labor market. Several disciplines debate the complex interplay that exists between race and ethnicity. Here it is recognized that there is a social construction called “race” which is one dimension of ethnicity. Race, together with other ethnic characteristics, may be linked to differences among people that may have measurable consequences over economic opportunities. The use of Race as a measurable instrument in empirical work is not a trivial task. We approximate it using indicators based on self-report data and pollsters’ data using a score-based procedure, that try to approximate the racial heterogeneity observed in Peru. Using different dimensions of ethnicity, this research project tries to understand the channels through which this subtle although effective mechanism of discrimination prevails in Peru. The purpose of this research program is to analyze the extent and consequences of exclusion in different markets, learning how the different forms of exclusion have impacts on the access to opportunities for socioeconomic advancement. In particular, we use a data set that collects self-reported discrimination events, better empirical approximations of ethnicity, like race, mother tongue of the parents, language spoken at home and at school; and education of the previous generation. We plan to analyze the relation between ethnic and racial heterogeneity and occupational segregation, access to education and to social networks. Additionally, we propose the collection and analysis of a new dataset that will allow us to explore the extent to which there are exclusion mechanisms operating in the hiring process in the urban labor market. The understanding of exclusion mechanisms will help us in exploring potential elements to be taken into account by policy makers in order to reduce ethnic and racial related inequality of opportunities. II. Analysis and explanation of the project design and research strategy People may be excluded based on their possession of different traits, some of them considered as endowments –like raceand some of them acquired through life. Among the characteristics through which people maybe excluded or discriminated we find gender, race, ethnicity, age, weight or height. All these characteristics may be more or less important to explain why somebody is temporally or permanently excluded from certain markets as can be shown in Figure 1. Figure 1 How Does Exclusion Operate? Markets: Credit Labor Services: Education Health Civil and Political Rights: Right to vote, citizenship. Access?