Abstract
Despite the well-documented large returns to good teachers, many developing countries face challenges in attracting and retaining qualified teachers in public schools. While various interventions have focused on improving teacher quality, little is known about the effect of job protection on teachers and students. In this paper, I use a regression discontinuity design and unique administrative data on teacher applications and students to analyze the impact of a tenure policy in Peru on the selection of teachers and student learning. The policy ties access to permanent jobs in Peruvian public schools to passing a centralized entrance exam and a local evaluation. I show that teachers who barely pass the exam are 3.8 percentage points less likely to exit from public schools. Using value added estimates of teacher quality, I show that this retention effect is driven by higher value-added teachers. Furthermore, using a within-student, across-subject strategy, I document positive returns to both having a teacher who just passed the exam; and a teacher who got a permanent job. However, these returns are largely driven by high ability students and students in urban schools. Finally, I examine the selection effects of the policy through a difference-in-differences analysis and show that it successfully attracted higher quality teachers to public schools.