Own-Race Bias Among NBA Coaches

Abstract

This article finds that National Basketball Association (NBA) coaches gave greater minutes per game to players of their own race during the 1996-2004 seasons after controlling for player quality using performance statistics and player fixed effects. The authors estimate that having the same race as the coach increased playing time by between 45 and 55 seconds per game on average. One possible explanation for this result is an own-race bias on the part of NBA coaches, in which they subconsciously exhibit a preference for players whose race matches their own. The estimates reveal that the impact of race on playing time was relatively strong in the late 1990s but that it has declined over time since then. The authors hypothesize that racial bias may emerge more strongly in decisions made under pressure, such as substitution patterns during the flow of the game, than in decisions made after careful consideration, such as the choice of which player to start. Empirical tests provide little evidence, however, that deliberation completely removes the own-race bias since coaches both gave more playing time and were more likely to start players of their own race.