posted on 2020-08-18, 20:38authored byKarthik Nagarajan
Controlled school choice policies aim to provide families a choice as to which school their child will attend while maintaining diversity in schools. These policies are of- ten implemented in the form of minimum and maximum quotas on student types (that can be based on race, ethnicity or socioeconomic factors). When these quotas are interpreted as hard bounds, standard fairness and non-wastefulness properties are compromised, imposing a cost on student welfare. By interpreting the minimum quota as a flexible guideline (a ”soft” bound), I employ a mixed interpretation of these quotas to formulate a modified version of the student-optimal deferred acceptance algorithm that guarantees fairness. Furthermore, I show that both this and a purely soft inter- pretation of the diversity bounds can be used to formulate effective mechanisms that ensure fairness, even in settings with notionally hard bounds.