In 2007 a study by Joseph Price and Justin Wolfers drew widespread attention to implicit racial bias in NBA referee calls – namely, that white referees called fouls against black players more often than against white players (and vice-versa). The authors recently revisited the issue and found something surprising: since the publication of their paper that bias had completely disappeared.

The graphic below looks at foul-calling against white and black players based on the racial make-up of the three-man officiating referee crews. The first chart, of the time period immediately prior to the publication of the 2007 study, shows that the more white referees are involved in a game, the fewer fouls are called against white players, with a much smaller decline for black players. Study author Justin Wolfers calls this difference “small, subtle and significant.” The second chart looks at the period after the study came out and shows two things: the first is that there’s no longer a meaningful difference in foul calls based on referee make-up. Second, the overall number of foul calls decreased slightly.


What changed between the two time periods? According to the NBA, “absolutely nothing.” The league told the study’s authors that it neither spoke with referees about the initial study, or changed referee incentives in any way.

Now, a little back-story is in order. The 2007 study led to a flurry of media attention and a fierce backlash from NBA officials all the way up to Commissioner David Stern, who called it “a bum rap.” The NBA even did their own study of referee bias in response and claimed it undercut Price and Wolfers’ conclusions.

But upon closer scrutiny that NBA report actually confirmed Price and Wolfers’ findings. “I don’t think the NBA fully understood their own study,” Wolfers said in an interview. He surmises that the NBA was so quick to denounce the original study – and to deny any subsequent changes in policy – due to a fear of litigation. A systemic pattern of racial bias – even at the implicit, subconscious level – could theoretically open the door to a discrimination lawsuit.

For his part, Wolfers calls it “unlikely” that the NBA did absolutely nothing in response to the initial paper, as they claim. This opens up two interesting possibilities. If NBA did indeed make an effort to correct referee bias, the latest study shows that they succeeded spectacularly. This could be a huge win for the organization, and there could be lessons here for others looking to reduce their own exposure to bias.

But it’s more interesting to take the NBA at their word. Even if there was no explicit policy change in response to the original study, the referees would have known about it – the front-page newspaper coverage and widespread discussion on major news networks and the internet would have seen to that. In this scenario, simply becoming aware of their implicit biases caused the refs to alter their decision-making process.

“Racial bias is a malleable trait,” says study co-author Joseph Price. “Large-scale public focus on a specific type of racial bias in a specific group can make it go away.” Price says the same could hold true for any situation in which implicit racial bias plays a role – from police officers deciding who to pull over to teachers deciding how to grade an essay.

“Large-scale public focus” here is key. This suggests that good social science research, and good media discussion of that research, has the ability to move the needle on thorny, intractable problems like implicit bias.