King Kaufman's Sports Daily

If a study about racial bias in umps is to be believed, the news is that such bias is easy to fix. Are you listening, NBA? Plus: Phil Rizzuto.

Read more: Sports, Baseball, Race, NBA, Basketball, Major League Baseball, King Kaufman, Sports Daily, MLB

Aug. 15, 2007 | Hot on the heels of a study that found a slight racial bias in the work of NBA referees comes one that finds the same thing in the work of major league baseball umpires.

Next they'll be telling us there's racial bias in law enforcement. I mean, where does it end?

The baseball study, by economics researchers at Auburn and McGill universities and the University of Texas at Austin, asserts that "controlling for umpire, pitcher, and batter fixed effects and other factors, strikes are more likely to be called if the umpire and pitcher match race/ethnicity."

Like the NBA study, which was reported in May, this one has not yet been subjected to peer review. There has been some criticism that the study is flawed because of sample-size problems, since there are so few minority umpires. Eighty-seven percent of major league umpires are white, as are 71 percent of pitchers, according to the study.

I'm pleased to be able to address those concerns in simple terms: I have no idea if those concerns are valid.

But there does at least seem to be something there with the group that is large enough, whites. The researchers, who looked at the 2.1 million pitches that were thrown from 2004 to 2006, write that the likelihood a pitch will be called a strike goes up by about 1 percent when the pitcher and umpire have the same race or ethnicity, which usually means they're both white. Since about 53 percent of the 275 pitches in an average game are called balls or strikes, the study says, something like one and a half pitches per game are affected.

That could be huge, of course. We've all seen a game turn on a single ball-strike call. But that's kind of beside the point. These researchers are economists, after all, and when economists study sports, they're usually not interested so much in the games themselves as in the fact that they are such handy laboratories. Here's what they're getting at:

"The results suggest that attempts to measure salary discrimination generally may be flawed, since the productivity measures can themselves be contaminated by the effects of racial preferences."

And that makes the rest of their findings even more interesting: The researchers say that the bias effect disappears in games where QuesTec, a computerized system that measures the accuracy and consistency of umpires' ball-strike calls, is used. It also goes away in well-attended games, and when there are three balls or two strikes on the hitter, in which case the next called pitch will result in the end of the at-bat.

Next page: The study suggests racial bias can be reduced, and fairly easily. Plus: Phil Rizzuto