Racial Diversity and Measuring Merit: Evidence from Boston's Exam School Admissions

Abstract

Abstract The impact of admissions process design on the racial diversity of schools and colleges has sparked heated debates. We study the pipeline into Boston's three public exam schools to understand racial gaps in enrollment. Admission to these schools has historically been based on a combination of grade point average (GPA) and a score on an optional test from a private developer. We document racial gaps in test-taking rates, test scores, GPAs, preferences for the most selective school, and ultimate admission rates to all three schools. These gaps persist even among students with similarly high baseline achievement as measured by the state's mandatory standardized test. Substantial numbers of high-achieving Black and Hispanic students do not apply to the exam schools and to the most selective school in particular. The choice of standardized test used to measure academic merit strongly affects who is admitted.