Student Outcomes and Teacher Identity: Evidence from Assam's Teacher Randomization

Beginning in 2012, the state government of Assam, in India, implemented a policy in which new primary school teachers are posted to government schools based on random assignment to increase transparency. In 2016 alone, 7,234 teachers were posted across 10 districts using a computerized lottery system. We plan to exploit this state-wide policy to understand how teachers’ characteristics (including teacher test scores, distance to school, and linguistic match, etc.) influence children’s schooling outcomes and school choice. We find supportive evidence for the role of teacher characteristics when matching the 2016 teacher assignment with school-level outcomes data (DISE): primary schools randomly assigned to receive at least one teacher from a marginalized group register a significant increase in student enrollment from these groups. Since randomized teacher assignment is at the school (rather than classroom) level, we will be equipped to also study the broader general equilibrium effects of this policy, including effects on private school enrollment, teacher retention, and community-level outcomes. We further plan to use the random assignment to inform a structural model of school choice and student outcomes to examine the optimal allocation of teachers to schools. A primary goal of our travel/project development grant work is to understand key details of the implementation of the teacher lottery, as well as explore potential research partnerships with the Government of Assam.