Racial Bias and Networking in Employment Outcomes: Evidence from NBA Coaches

Abstract

Increasing diversity in upper management positions is often cited as a way to improve a firm's performance and generate more diverse hires. We study the impact of management diversity on employment outcomes using rich data on NBA head coaches that includes information on the race of each coach and his supervisor. The data allows us to analyze how race affects multiple facets of the labor market, including hiring decisions, how quickly coaches are rehired, and the quality of jobs received. We find evidence that a supervisor is more likely to hire a coach of his own race. We also find that black coaches are less likely to be rehired to a second job within three years than their white counterparts, and that a head coach's race affects the quality of the opportunities he is offered. These findings are generally stronger for first-time coaches than experienced coaches, suggesting that observable information on performance mitigates bias. Finally, we argue that the results are consistent with taste-based discrimination.