Elite Secondary Schools and Student Achievement: Regression Discontinuity Evidence from Kenya

Abstract

This paper estimates the causal effect of attending an elite (or high performing) secondary school on student achievement using data from Kenyan secondary schools. The admission of students into government secondary schools in Kenya is centralized and is based solely on student scores on the national primary school exit exam, district quotas, and students' stated preferences. The assignment rules cause students of similar ability and preferences near the implied assignment score cutoffs to be assigned to secondary schools of different quality. We use a regression discontinuity design to obtain causal estimates of the effects of elite school attendance on student progression through secondary school and achievement in the secondary school exit examination. As the assignment rule generates different score thresholds for each district, we can examine whether the effects of attending an elite secondary school differ by the student's initial (primary school) test score. For students whose primary school test scores are near the threshold of an elite school, we find that elite school attendance does not affect timely progression through secondary school or achievement.