Why Discrimination Based on Stereotypes Often Reduces Firms’ Expected Profit

Discrimination based on stereotypes is often assumed to be rational from the perspective of individual firms (or individual decision makers) under conditions of imperfect information. This perspective is particularly common in the economics literature, where such discrimination is usually referred to as “statistical discrimination”. In this article, I challenge this perspective and investigate whether there are cases where discrimination based on stereotypes—even statistically accurate stereotypes—reduces firms’ expected profit. I show that, in hiring decisions, there are cases where collecting more information about individual applicants leads to a higher expected profit than discrimination based on (even statistically accurate) stereotypes does. I then show that, in callback decisions, there are different types of cases where inviting applicants from different groups to a job interview leads to a higher expected profit than discrimination based on (even statistically accurate) stereotypes does. Finally, I show that the results are robust to considering risk aversion in employers’ utility function. The results improve the understanding of stereotyping and discrimination and possibly open up new ways to combat discrimination in practice, because they show that even for employers who are only interested in profit maximization, it is often rational to avoid discrimination based on stereotypes