The Diversity Discrepancy, Or Why Your Classmates Are More Like You

Abstract

Summary Diversity has many potential benefits, and increasing diversity in the classroom and the workplace is a high-priority goal for managers and policy-makers. Although there is nearly uniform support for increasing diversity, there is less agreement about how to quantify diversity. This note observes that there is a systematic bias between how diversity is perceived by an individual and the natural measures of population diversity. In particular, most individuals will find themselves among people who are “more like them” than the aggregate statistics indicate. For example, considering racial diversity in the classroom, do you measure diversity as the proportion of minority students in the average class, or the average proportion as experienced by a member of the minority? Using examples from the US Census and the Philadelphia Public Schools we show that these two metrics can be quite different, and we give a general mathematical treatment of this discrepancy.