This month, a federal grand jury indicted the police officer who fatally shot Walter Scott as he ran from a traffic stop in North Charleston, S.C., a year ago. Lawyers for Michael Slager, who was charged with murder, say Scott had grabbed Slager's Taser during a traffic stop.
Slager is white, like approximately 80 percent of the North Charleston police department. Scott was black, as is about half of North Charleston's population. Cities with similar demographics — where the population is evenly divided between white and black residents but there are only a few black officers on the force — tend to have the highest rates of fatal police shootings of black civilians, the research suggests.
Although reformers and activists have been calling on law enforcement to hire more black cops for decades, the new working paper suggests that there may be
measurable advantages to more diverse police forces, at least in certain circumstances.
"Diverse police departments are particularly able to alleviate tensions in cities ... where you have preexisting racial, ethnic tensions," said Joscha Legewie, a sociologist at New York University. He conducted the research with Columbia University's Jeffrey Fagan, a legal scholar.
Consider two hypothetical cities, both with populations half white and half black. The cities are identical except for their police forces. In the first city, there are no black police officers, but in the second, one out of eight officers is black. Because the second city has a more diverse police force, Legewie would expect there to be 18 percent fewer killings of black civilians by police there than in the first city, based on the data he and Fagan collected.
The pattern was less pronounced in cities with large white majorities, or where more than two major racial or ethnic groups divided the population.
There are only a handful of police departments where the share of black officers is at least equal to the share of the population that is black. A few examples are the departments in Austin, Dallas, Denver, Los Angeles and D.C.
Legewie and Fagan used data on police shootings from the online clearinghouse Fatal Encounters from 2013 to 2015, doing their own research to fill in missing information about the races of the deceased. They focused on incidents in which black civilians were killed, since that problem has attracted the most public attention, Legewie said.
He and Fagan based their approach on other evidence suggesting that where a population is equally divided between two racial or ethnic groups, the hostility between them can be greater than when one group is clearly outnumbered or where there are several distinct groups.
In comparing the rates of killings of black civilians by police in different cities, the researchers accounted for other factors that might create a sense of a racial competition, such as disparities in income between black and white residents, black immigration from other places and the race of the mayor. They also accounted for social conditions that could make police work more dangerous in some cities, such as poverty and rates of arrests and violent crime.
After controlling for those factors, the researchers found that in cities where the population was closely divided between races, killings of black civilians were less common when the racial composition of the police force reflected the population as a whole.
Legewie and Fagan speculate that given the dangers involved in law enforcement, police develop a sense of loyalty to their fellow officers and that their loyalty might be less likely to reinforce unconscious racial biases when they are accustomed to seeing people of different races wearing blue.
A diverse police force was also helpful in places where there were more white victims of homicides in which the suspects were black.
Homicides involving two people of different races are rare. In 2014, the suspects were black in only about 15 percent of homicides in which the victim was white, according to the Federal Bureau of Investigation. Yet in cities where more such homicides were committed, killings of black civilians by police were also more common — although less so in cities where the make-up of the police force was similar to the population's.
By contrast, there appeared to be no connection between killings of black civilians by police and the rate of homicides involving black suspects and black victims.
Legewie cautions that their research is preliminary. He and Fagan put the working paper with their results so far online just this month. It is still unclear whether police departments could indeed reduce the chances of black civilians being killed by hiring more black officers.
More from Wonkblog: