What Police Departments and the Rest of Us Can Do to Overcome Implicit Bias, According to an Expert

Jennifer Eberhardt is a MacArthur “genius grant” winner and psychology professor at Stanford University who studies implicit bias. TIME spoke with her about her new book, Biased: Uncovering the Hidden Prejudice That Shapes What We See Think and Do, as well as her research, her work with police departments and how implicit bias can affect us all.

You open your new book, ­Biased, with the story of your son worrying that a fellow black passenger will blow up a plane. Why? To show how deep implicit bias is and how it can affect everyone, even a black child. This is something that every­body has to grapple with. We’re living in a society where we’re absorbing images and ideas all the time and it takes over who we are and how we see the world.

As a child you moved to a largely white school, and you couldn’t tell the girls apart. Our brains get attuned to what we’re surrounded by. And so for me, I’m really good at recognizing black faces, being able to distinguish one from another. But then I moved to this other neighborhood where all of a sudden I’m surrounded by white people with whom I had never had any real meaningful interaction before.

And even though I wanted to have friends and all of this in this new neighborhood I really couldn’t tell their faces apart. I had been in really segregated spaces. I was attuned to different features, like [shades of] skin color. So it took a lot of practice in that environment before my brain was able to sort through [using hair and eye color].

How is unconscious racial bias not just racism? When people think about racism they’re thinking about bigots. But you don’t have to have a moral failing to act on an implicit bias.

The brain doesn’t like chaos, you write, so it works to categorize things. How is that a precursor of bias? The brain needs to sort ­everything—the food we eat, the furniture we use, whatever. We also sort people. That sorting can lead to bias; once we have categories, we have beliefs and feelings about what’s in those categories.

You won a MacArthur “genius” grant for your work on bias at Stanford University. Which study did you personally find most compelling? There was a study where we exposed people to faces subliminally—a set of black faces and a set of white faces. We then showed them a blurry image of an object, which got more and more clear. Some were crime-­related, like guns or knives, and some were cameras and staplers. We found that being exposed to black faces for milliseconds leads people to pick out guns and knives sooner. That the association of blackness with crime can affect what we see in this literal way was pretty revealing.

Could you explain to me your studies about African Americans and apes? Just as blacks are associated with crime, blacks also are dehumanized and they continue to be associated with apes. We conducted a study similar to the one we had conducted with crime. This time people were exposed to images of black faces or white faces. And then we gave them line drawings of animals to look at. Again, they started out blurry and then became more clear. And we found if white or black people had been exposed to black faces beforehand they were a lot quicker to detect blurry images of apes. There’s this tight association between blacks and apes and there’s a racial imagery there that’s affecting our visual perception.

You work with police departments. How can lab work explain police behavior? If a white person was placed in the identical situation to Philando Castile [who was shot by police after he said he had a gun], we don’t know if the outcome would have been different. But in the laboratory we can create identical conditions, except for race. And we can then look to see what is the causal impact, what’s the role that race is playing in producing that behavior?

Why do you resist the idea that shootings can be blamed on the racism of one particular cop? I feel like it’s myopic. It could be that the person was implicitly or explicitly racially biased. But if we’re in a context where there are tense police-­community interactions, we want to look at how they affected the people in that interaction. So, in Oakland, police changed their foot-­pursuit policy about 10 years ago. If you lose sight of the person you’re chasing, you’re supposed to step back and set up a perimeter. Otherwise you’d be following them into a situation where you’re trapped. You have to act quickly and you’re afraid. Those are the conditions under which bias is most likely to affect decision-making. Oakland went from having eight or nine officer-­involved shootings a year to six in two years.

You analyzed 28,000 police stops in 2013 and 2014. What surprised you most in all those interactions? Handcuffing. It was one of the big issues we heard about in the community, especially for black men. We looked at the data [from Oakland], and sure enough, even when no arrest was made, 1 out of 4 black people were handcuffed. And 1 out of 15 whites. Police were seeing it as an officer-­safety issue. But it’s traumatic and was having an impact on the community.

What can people do about their own implicit bias? There are certain conditions under which we become more vulnerable to it: when we’re thinking fast and moving fast. We can slow down and make a shift so we’re less likely to act on bias.

Can you give me an example? I did a little informal consulting with Nextdoor.com. Most of the people go to the platform because they’re trying to find a good plumber or to sell something or to alert people to various events in the neighborhood. But then there’s also the “suspicious black man” posting. Nextdoor wanted to reduce that racial profiling.

In the tech world, they really prize being able to do everything fast, without friction. But they added friction to the platform. For the crime and safety tab, you can’t just write. There’s a black man, suspicious. You have to identify some behavior that is actually suspicious. And then be specific about what that person looks like so it doesn’t sweep all black people in the same category. You know that sign, If you see something, say something? They changed it to, If you see something suspicious, say something specific. It’s trying to get people to stop and think. By slowing people down, getting them to think about what they were posting, they were able to curb the profiling, they say, by about 75% on the site.

How old is your son now? He’s 16. He’s now become the target of those perceptions, of his own perceptions really. He was riding a bike and noticed that a jogger who was coming towards him sees him and she veers into the middle of the road. And he was trying to figure out why that was, and he came to the conclusion that she was afraid of him. And so we had a discussion about what that fear was about. It reminded me of the time when he was on the plane and he had the same ideas in his head. It came full circle, and it was the kind of starting point for him living his life as a black man in this country.