Ethnicity, Institutions, and Sectoral Favoritism

Abstract

The existing theoretical and empirical literature examining the role of institutions in explaining differences in levels urban bias across developing countries assumes cities and/or economic sector are the basis of collective action. However, much of the conflict literature in political science and economics stresses the role of ethnic identity in mobilizing resistance against governments. This paper assesses the role of ethnicity in explaining differences in levels of government rent-seeking across regions. A simple two-sector, open economy model is developed where an autocratic, rent-maximizing government collects revenue from an ethnic enclave, which can choose to rebel. This model predicts that, within a particular enclave, the urban/rural income difference should be zero, while larger and more urban enclaves should have higher average income. To test the model, we then use geographic location and self-reported ethnicity of survey respondents from the 2005 Afrobarometer Survey and examine the relationship between these variables and reported number of assets owned by the respondent. Within autocratic countries, we find a statistically significant, positive relationship between the urbanization of the respondent's ethnic enclave and respondent wealth, but no relationship between respondent's urban residence and assets within enclaves, consistent with the model. By contrast, within democratic countries, we do not find a statistically significant relationship between the urbanization of the respondent's ethnic enclave and the asset-based wealth index, but find a statistically significant, positive relationship between respondent urban residence and the index within enclaves.