Testing for Bias in Graduate School Admissions

Abstract

This paper provides an empirical examination of the factors that influence graduate admissions decisions. It exploits a unique, large data set on applications and admissions to 48 leading graduate schools in five disciplines, including economics. The analysis shows that these graduate schools in the aggregate gave substantial preference in four out of five fields to U.S. citizens over foreign applicants, modest preference in three fields to women over men, and substantial preference in all fields to underrepresented minorities over other U.S. citizens. The findings suggest that higher standards are applied to overrepresented groups to achieve more diverse enrollments.