Felons justifiably killed by police represent a tiny fraction of the total population. Of the 183 million whites in 1998, police killed 225; of the 27 million blacks, police killed 127. While the rate (per million population) at which blacks were killed by police in 1998 was about 4 times that of whites, the difference used to be much wider: the black rate in 1978 was 8 times the white rate.

Participants in an innovative Washington State University study of deadly force were more likely to feel threatened in scenarios involving black people. But when it came time to shoot, participants were biased in favor of black suspects, taking longer to pull the trigger against them than against armed white or Hispanic suspects...

[WSU researcher Lois] James’ study is a follow-up to one in which she found active police officers, military personnel and the general public took longer to shoot black suspects than white or Hispanic suspects. Participants were also more likely to shoot unarmed white suspects than black or Hispanic ones and more likely to fail to fire at armed black suspects.

“In other words,” wrote James and her co-authors, “there was significant bias favoring blacks where decisions to shoot were concerned.”

When confronted by an armed white person, participants took an average of 1.37 seconds to fire back. Confronted by an armed black person, they took 1.61 seconds to fire and were less likely to fire in error. The 24-millisecond difference may seem small, but it’s enough to be fatal in a shooting.

This behavioral ‘counter-bias’ might be rooted in people’s concerns about the social and legal consequences of shooting a member of a historically oppressed racial or ethnic group.