Pro Basketball Coaches Display Racial Bias When Selecting Lineups

Research finds that NBA coaches give slightly more playing time to players of their own race, but the gap disappears at playoff time. Research by Letian Zhang.
by Michael Blanding

As the National Basketball Association kicks off its regular season this week, a recent study makes a surprising discovery: Coaches favor players of their own race, giving them slightly more playing time than might be expected.

And in follow-on research soon to be published, results show that coaches demonstrate less racial preference when their team is on a losing streak or in playoff games.

More than any other American sport, basketball is dominated by African American players. Three-quarters of athletes running up and down the court are black, compared to 23 percent white (with 2 percent Latino and 0.6 percent Asian). When it comes to NBA coaches, however, the exact opposite is true: there are six black head coaches among 30 teams, or just 20 percent of the league.

Research by Harvard Business School Assistant Professor Letian Zhang, who studies organizational theory and strategy with a focus on social inequalities and status hierarchies, reveals that whether consciously or not, white coaches and black coaches subtly favor players of their own race. That, in turn, can hurt the league overall.

“When you favor somebody because of their race, you are not playing your best players—or using your best workers”

“Racial bias is costly,” says Zhang, who speculates the findings could apply equally in a number of business contexts. “When you favor somebody because of their race, you are not playing your best players—or using your best workers.”

Zhang, who is both a basketball fan and a sociologist, began looking at the NBA several years ago as an ideal field in which to study the insidious effects of discrimination.

“Racial bias is really hard to measure,” Zhang says. In most business scenarios interpersonal differences makes it difficult to isolate the effects of race on people’s behavior. Performance, too, is tricky to gauge objectively. “Most workplaces have a rough measure of productivity, but that is usually biased itself.”

The NBA, however, obsessively tracks every action of every player, giving a moment-by-moment snapshot of performance, including shots, points, assists, fouls, and minutes played. And because players frequently change teams, it’s easy to compare how a player fares when playing for different coaches.

For a paper Zhang published on the topic last year in Administrative Science Quarterly, Zhang downloaded data for each player by season from the website, and then compared how many minutes each player played under black coaches versus white. He found that players averaged 40 second less per game when playing for a coach of a different race, regardless of his actual performance on the court.

The effect was particularly strong in the first year a player played for that coach, when he averaged a minute less each game. “That is pretty substantial, considering that most players only get 15 minutes per game,” Zhang says. “It’s a significant enough factor that it can influence a career.”

Zhang doesn’t think that coaches are racist per se. As fans know, however, coaches tend to favor certain players over others, and sociology research shows that in general, people have inherent preferences for people who look like them.

“Every coach has their favorite players,” Zhang says. “It just so happens that those players tend to be the same race.”

The data does show that those preferences generally go away after three years, as a coach has repeated contact with a player. That effect, however, only applies to that particular player, and not to other players of that race in the future. (Zhang calls that the “Driving Miss Daisy” effect, in which one member of a particular race, but not others, is exempted from bias.)

In a new paper to be published in Organizational Science, Zhang found that the bias a coach shows isn’t set, but varies from game to game. In some games, coaches favored players of the same race, while in others, race seemed to have no effect on the players they put on the court. In order to figure out why, Zhang scraped game-by-game data on players.

Overall, he found that coaches tended to show more bias when they were winning games rather than when they were losing. When a coach won nine out of 10 of his last games, for example, players of the same race got an extra minute of playing time. When a coach lost the same number, no bias was shown. Zhang attributes that phenomenon to the “slack” that coaches get when they are on a hot streak, when fans and owners put less pressure on them to perform, and they can exercise more of their own individual player preferences.

By contrast, when performance is weak, coaches feel more pressure to improve, and so they put aside their internal biases in favor of more impartial measures of performance such as game stats. That can also be the case when the stakes are high, as in the NBA playoffs, where Zhang found little evidence of bias no matter how many games a coach won.

“In the playoffs, the pressure is always high,” Zhang says. “Even when you have won several games in a row, you have to keep winning to stay alive. There really isn’t any slack.”

Bias in business

In a business context, Zhang surmises that racial bias might similarly be more prevalent when performance is high. It’s been well documented that company managers sometimes show hubris when a firm is ahead on earnings projections. “When CEOs perform well, they think they are smarter than anyone else, and then start to trust their own judgment and ignore objective information,” Zhang says. “When they don’t feel that pressure to perform they might exercise more biases.”

“Every coach has their favorite players. It just so happens that those players tend to be the same race”

Extrapolating from his research, Zhang says it’s not enough to put together diverse teams in order to reduce bias. Regularly working with people of other races does not guarantee that managers will treat those who look differently without bias. That, in turn, could cause them to pass over better workers in favor of those of their own race or background.

The danger is particularly high when things are going well. At those times, it may be easy to become complacent, Zhang says, and not realize the drag that subtle racial bias may be putting on future performance—on the court or in the cubicle.

“When you are really high-performing, it makes you sometimes forget about the more objective metrics and you could tend to become more racially biased and exercise more personal preferences,” Zhang says. “Being aware of that tendency is important for better decision making, as well as for reducing inequality within a firm.”

About the Author

Michael Blanding is a writer based in Brookline, Massachusetts

Post A Comment

In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.