Setting Priorities in School Choice Enrollment Systems: Who Benefits from Placement Algorithm Preferences? Technical Report

Many cities with school choice programs employ algorithms to make school placements. These algorithms use student priorities to determine which applicants get seats in oversubscribed schools. This study explores whether the New Orleans placement algorithm tends to favor students of certain races or socioeconomic classes. Specifically, we examine cases where families of Black and White or poor and non-poor children request the same elementary school as their top choice. We find that when Black and White applicants submit the same first-choice request for kindergarten, Black applicants are 9 percentage points less likely to receive it. Meanwhile, students in poverty are 6 percentage points less likely to receive a first-choice placement than other applicants for the same kindergarten program. However, these biases are not inevitable. In non-entry grades, where placement policies favor students whose schools are closing, Black and low-income applicants are more likely to obtain first-choice placements than their peers. We examine these priorities and simulate placements under alternate specifications of a deferred-acceptance algorithm to assess the potential of algorithm reform as a policymaking tool.