Admissions quotas in university education: Targeting and mismatch under Sri Lanka’s affirmative action policy

The state university system in Sri Lanka is characterised by free tuition, central administration and an affirmative action policy based on geographical region aimed at reducing inequality in access to higher education. Using administrative data on applicants for university admission, we study the effectiveness of the policy in targeting under-privileged regions and students and test for academic mismatch among medical students. We find that the policy ensures representation of disadvantaged regions while effectively targeting disadvantaged students. However, beneficiaries of the policy lag slightly behind the rest and this mismatch cannot be explained away by differences in university quality, though the estimated mismatch is likely to be an upper bound of mismatches occurring across the university system. These results suggest that region-based targeting is a viable alternative to the more common race/caste based affirmative action policies.