Abstract

If the costs and benefits of discriminating at the hiring stage differ from those at the retention and promotion stages, as recent evidence suggests, the effect of an individual’s race on a firm’s hiring decision should differ from its effect on the firm’s retention and promotion decisions. This paper presents the first direct empirical test of this proposition. Using data of players drafted into the National Football League (NFL), we show that after controlling for draft selection, position, team, draft year, collegiate division, and team wins the prior season, white players have a 0.13 lower probability of having an active contract and start 1.56 games less than nonwhite players. When compared with the average probability of a drafted player having an active contract, 0.48, and the average number of starts, 2.6 games, these results provide strong evidence that nonwhite players face hiring discrimination in the NFL but are treated more equitably in retention and promotion decisions. 1.

Citations