Ethnic favouritism in Kenyan education reconsidered:when a picture is worth more than a thousand regressions

Does a leader's ethnicity affect the regional distribution of basic services such as education in Africa? Several influential studies have argued in the affirmative, by using educational attainment levels to show that children who share the ethnicity of the president during their school-aged years have higher attainment than their peers. In this paper we revisit this empirical evidence and show that it rests on problematic assumptions. Some models commonly used to test for favouritism do not take adequate account of educational convergence and once this is properly accounted for the results are found to be unstable. Using Kenya as a test case, we argue that there is no conclusive evidence of ethnic favouritism in primary or secondary education, but rather a process of educational convergence among the country's larger ethnic groups. This evidence matters, as it shapes how we understand the ethnic calculus of politicians.