The Washington PostDemocracy Dies in Darkness

Researchers have discovered a new and surprising racial bias in the criminal justice system

(Amy Cavenaile/The Washington Post; iStock)

For a nation as diverse as the United States, the judiciary is quite male and white. In theory, this shouldn’t matter. Judges are supposed to be impartial — as boring as “umpires,” in the famous words of Chief Justice John Roberts.

In practice, of course, it's much messier. People can’t help but see the world through the lens of their own experiences. A recent report by two sociologists showed that white federal judges are about four times more likely to dismiss race discrimination cases outright. Another analysis found that they are half as likely as black federal judges to rule in favor of people alleging racial harassment in the workplace. 

“Since African American judges have likely experienced discrimination themselves, they can recognize more complex and subtle forms of racial harassment,” the authors argue.

But in criminal cases, the evidence has been blurrier. Just because black judges might understand more about the black experience doesn’t mean that they’re more sympathetic toward black criminals. In fact, a surprising new study shows that judges in Louisiana — white or black — actually give longer sentences to juvenile offenders of their own race.

“It was a little bit shocking at first,” says Briggs Depew, an assistant professor at LSU and co-author on the paper, which was released as a draft by the National Bureau of Economic Research on Monday.

Depew and his colleagues, Ozkan Eren and Naci Mocan, found that when a black judge rules on a black defendant, or when a white judge rules on a white defendant, the sentences are about 14 percent longer. Also, the judges are about 5 percent more likely to recommend jail time instead of probation for defendants of the same race.

The study hinges on the fact that these cases are assigned randomly. There’s no telling whether a black judge or a white judge will handle someone’s hearing. The researchers also controlled for the age and gender of the defendants, the place where the crime was committed, and the average tendencies of each judge. They focused on cases in which the defendants already plead guilty, so the judges were only in charge of deciding the terms of the punishment. 

In other words, there are few alternative explanations for this pattern of racial bias. Judges in Louisiana, it seems, are systematically treating similar defendants differently. And, unexpectedly, they are more lenient on defendants who are of a different race.

The researchers have a few theories for why this might be.

First, it could have something to do with the race of the victims. There wasn’t any information about that in the Louisiana data, but the researchers note that national statistics show white people are more likely to be victims of crimes perpetrated by white people. Correspondingly, black people are more likely to be victims of black crimes. Judges may be more sympathetic toward victims of their own race, and may mete out more severe punishments to those who perpetrate such crimes, regardless if they are white or black.

The researchers have some preliminary evidence supporting this idea. When they focused on victimless crimes, they found that the judges seemed to be a lot more even-handed. But the same-race bias was large for crimes with victims, implying that something here was important. (There wasn’t enough data for this difference to be statistically significant, but it’s suggestive.)

Second, this could be tough love. Since judges may more knowledgeable about people of the same race, they might feel more comfortable assigning longer sentences. Out of unfamiliarity, a white judge might prefer to err on the lenient side when dealing with black defendants. A black judge could have more insight into whether a black defendant could benefit from more jail or probation time. 

Third, judges could be worried about seeming prejudiced. White judges might go easy on black defendants to ward off accusations of racism, and black judges might treat black defendants more harshly to avoid the appearance of racial favoritism. In laboratory tests, researchers have shown that white trial judges, like most white Americans, hold implicit biases against black people — but also that judges can overcome those prejudices with conscious effort. The racial patterns in Louisiana, might be the result of overcompensation.

A study like this tends to raise more questions than it answers. The important message is that diversity matters. As my colleague Max Ehrenfreund noted last week, President Obama's nominations to the federal judiciary have been remarkably non-male and non-white. Obama explained to The New Yorker in 2014 that he wanted minorities "to be able to see folks in robes that look like them." But the consequences of a multiracial, multicultural bench will be more than just symbolic.

As America's judges start to look like, well, the rest of America, the law will bend in interesting ways. A striking example is Supreme Court justice Sonia Sotomayor, a Latina who grew up in a Bronx housing project. Just on Monday, she scolded fellow justice Samuel Alito for being ignorant, apparently, of how arrest warrants can tangle up the lives of the poor.

“Personal experiences affect the facts that judges choose to see,” Sotomayor once said in a 2001 speech (the same in which she made her "wise Latina" comment). "My hope is that I will take the good from my experiences and extrapolate them further into areas with which I am unfamiliar."