Does racial diversity improve academic outcomes? A natural experiment in higher education classrooms

This paper estimates the causal effect of racial diversity on academic outcomes in a higher education classroom setting. I exploit a natural experiment where first-year college students in a mandatory writing course are assigned to discussion conferences with varying racial compositions. Within-classroom diversity is effectively random conditional on scheduling availability. I find that a higher degree of classroom diversity increases GPA at graduation, improves first year grades of female students, and affects the major choice of white students. I develop a trimming procedure to bound estimates and account for selection into the sample of completers. My results highlight the potential value of racial diversity in higher education and contribute to the debate over race-based admissions policies.