5/13/2007

Racial discrimination by NBA referees ?

A recent paper, after studying 13 NBA seasons from 1991 through 2004, proved that white referees called fouls against black players at a higher rate, about 4.5% higher, than they did against white players.

This research imply "that if LeBron James faced only white refs the whole season relative to having only black refs the whole season, he would be expected to run up an extra 11 or 12 fouls over the course of the season and score about .3 fewer points per game".

Although these are not very big numbers, results did show NBA referees having implicit, unconscious biases.

I don't believe NBA referees consciously uses race to referee a game. They should be the group people that least likely to exhibit racial biases. They receive a great deal training. Their decisions are consistently reviewed. They make decisions in the front of large audiences.

But in a fast-paced contest, where decisions must be made immediately, it is not surprising to see such bias sneak out. I hope that we will learn more about race and society.

3 comments:

Jay Livingston said...

The difference was tiny: if an all-white team played an all-black team, the difference would be one foul per game. An all-white NBA team is, of course, purely hypothetical. The Spurs sometimes have as many as three white guys on the floor (only one an American), so that's 0.6 fouls per game. With a 90% free-shooter at the line for two shots, that's about one point.

The research finding was "statistically significant" because the researchers used a huge sample. In statistics, "significant" is not the same as "meaningful."

(I too blogged this story.)

FBT said...

did they also study any discrepancies between black refs calling fouls against white players vs black players? Or are there not enough black referees for that to be a meaningful/ possible study?

Anonymous said...

Is it possible the Black Referees have a bias that causes them to notice less fouls by Black players?

The data is what is it, but the conclusions drawn from it aren't automatically correct.

Causation and coorelation are different things.