A new study from WSU Spokane says Spokane police officers appear less likely to shoot unarmed black suspects than unarmed white ones, even though most officers subconsciously associate black people with weapons.

Share story

A new study from WSU Spokane says Spokane police officers appear less likely to shoot unarmed black suspects than unarmed white ones, even though most officers subconsciously associate black people with weapons.

The study, by Lois James, Stephen James and Bryan Vila at WSU’s Sleep Performance Research Center, measured how long 80 Spokane patrol officers took to shoot armed black and white suspects in a simulator and how they reacted to unarmed suspects of each race. It also used an implicit-bias test developed at Harvard University to measure how strongly officers associated images of black people with weapons.

Officers were not told that the research was about race.

Although 96 percent of officers showed implicit bias toward black people on the test, that underlying bias didn’t appear to influence shooting behavior. Officers incorrectly shot unarmed black suspects in just 1 percent of simulations and shot unarmed white suspects 14 percent of the time. They also took 0.23 seconds longer, on average, to shoot armed black suspects over armed white ones.

“In spite of these implicit biases or perhaps because of these biases, officers are trying to reverse them or overcorrect,” Lois James said. Her study is titled “The Reverse Racism Effect.”

Since the 1970s, other published research has suggested implicit bias results in disproportionality in police shootings. University of Chicago psychologist Joshua Correll published a study in 2007 documenting that both police officers and community members are quicker to shoot armed black suspects than armed white suspects. That conclusion was based on showing people photos of armed or unarmed suspects and having them press a button for “shoot” or “don’t shoot.”

James says her study differs because it uses realistic, videogame-like simulations similar to the scenarios officers use to train. That provides a more realistic look at likely behavior in the field than simple button-pressing, she said.

Despite intense public scrutiny over officer-involved shootings, they remain poorly tracked by the federal government. The Washington Post tracked fatal officer-involved shootings nationwide in 2015 and found black people were shot at three times the rate of white people, adjusted for population where the shootings took place. Black men, who make up 6 percent of the U.S. population, were shot in 40 percent of cases where the shooting victim was unarmed.

James said while it’s useful to track police shootings and collect demographic information, it’s difficult to generalize about future behavior based on those results. Looking at the racial makeup of people shot by police relative to the total population doesn’t provide much useful information, James said, without knowing the demographics of people whom officers contact.

“That’s the major limitation, unfortunately, with data that’s gathered in officer-involved shootings in the field,” she said.

The study notes that the time difference in shooting armed black vs. white suspects is too small to be the result of conscious decision. It suggests that this effect is likely “rooted in people’s concerns about the social and legal consequences of shooting a member of a historically oppressed racial group.”

James has published two other studies on officer shooting behavior using realistic simulations. Both concluded that officers are less likely to shoot unarmed black suspects, but neither used an implicit-bias test to measure underlying bias.

According to Harvard’s website, 85,742 people took the race and weapons implicit- bias test online between 2001 and 2006. A majority, 72 percent, showed a bias toward associating black people with weapons, though 19 percent showed little or no bias, and 10 percent associated weapons with white people.

Most Spokane police officers tested showed a moderate or strong bias.

“I was surprised that it was as high as it was,” James said. “I wasn’t surprised that there is a tendency to associate African-Americans and weapons, just because that tendency tends to be shown in Americans in general.”

Officers in the Spokane study came into James’ lab four times between August 2012 and November 2013, before widespread protests against police shootings began in Ferguson, Mo., and elsewhere. The sample group was overwhelmingly white (95 percent) and male (88 percent).

Because Spokane is overwhelmingly white, the results from her study may not be generalizable to other cities with more racially diverse populations, James said.

“It gives us a preliminary look at what might be going on, but I don’t think it’s fair to generalize this to the police profession at large,” she said.

James plans to continue her research by repeating the same tests in other cities across the country.