The bias fighters

Psychologists are testing ways to reduce unconscious racial prejudice--not just in the police, but in all of us.

Photo Illustration by Lesley Becker/ Globe Staff/-

It may not be easy to admit, but chances are you’ve been guilty of racial discrimination. Perhaps just in small ways: that time you got disproportionately mad at someone for walking too slowly or laughing too loudly on the bus, or that time you talked down to a call-center guy because of his accent. Maybe it didn’t register as prejudice when it happened. Maybe you’ve kept your biases a secret, even from yourself. Regardless, a mountain of research tells us it’s a near-certainty: Racial stereotypes infect the way most of us think, make decisions, and interact with others.

This bias, unconscious or otherwise, has consequences—not just in our daily interactions, but in matters of life and death. Racial prejudice among police officers has been at the top of the public agenda after the fatal shooting of an unarmed black teenager, Michael Brown, by a white officer in Ferguson, Mo. Though it’s always difficult to know for sure that bias has played a part in any individual case, Brown’s killing echoed that of other unarmed young black men—Trayvon Martin, Amadou Diallo, Oscar Grant—all of whom died under circumstances that made people suspect they’d still be alive if they hadn’t been black.

In the quest to prevent more tragedies like Michael Brown’s killing, one focus has been the need to reduce the influence of racial prejudice on how police officers do their jobs. Beneath that lies a broader question: If even the most well-intentioned, egalitarian-minded people among us operate with some degree of racial bias, is there a way for any of us to limit its impact on how we behave?

To date, psychology has learned an immense amount about how bias affects us, and comparatively little on how to fight it. But this is starting to change, thanks to a squad of bias-fighting researchers who have devoted their careers to the problem. “The knowledge that implicit bias could actually be changed is relatively recent,” said Calvin Lai, a doctoral student at the University of Virginia who studies “de-biasing.” An important early glimpse came in a 2001 experiment by University of Massachusetts Amherst’s Nilanjana Dasgupta and University of Washington’s Anthony Greenwald, in which they found they could temporarily reduce people’s prejudices by showing them pictures of black icons like Martin Luther King Jr. in conjunction with infamous white villains like Jeffrey Dahmer. “Before that,” Lai said, “it was largely assumed that these stereotypes...were inflexible.”

Marchers in Washington protested the killing of black teenager Michael Brown, who was shot by a white police officer in Ferguson, Mo.Nicholas Kamm/AFP/Getty Images

The hunt for ways to eradicate prejudice from people’s thinking has gathered momentum since then, generating a pile of lab results on what works and what doesn’t. This spring, the Journal of Experimental Psychology: General published the results of an unusual research contest, in which Lai tested suggestions, submitted by more than 20 experts, for how to reduce unconscious bias against African-Americans, and ranked them based on their impact. On the law-enforcement front, de-biasing work has begun to make its way into the real world, with some experts working directly with police departments around the country to train officers, and set department policy, in a way that minimizes racial discrimination.

Should such efforts succeed, the lessons learned could transform not only the work of police officers, but also teachers, doctors, judges, and anyone else whose biases might result in unfair treatment based on race. Its proponents see it as a new avenue in the fight against discrimination—one centered on not legal requirements or moral leadership, but rather the human mind and how it reacts to the world.

One obstacle to eliminating bias is that only some of it is harmful—in fact, “bias” is just another form of the rational, shortcut-based thinking that allows us to use what little we know about the world to navigate it efficiently. This is just the kind of thinking required by police officers, who often find themselves in tense, high-stakes situations that require snap decisions based on reflexes and instinct.

Unfortunately, psychology research over the years overwhelmingly indicates that a majority of people in America, police officers included, have one problematic bias in particular: Their reflexes consistently incline them to assume the worst about black people, while privileging whites. A psychological measure called the Implicit Association Test, or IAT, has been used to find antiblack prejudice among 75 percent of whites and, surprisingly, 50 percent of blacks. In an illustration of how this plays out in our reactions, University of Colorado psychologist Joshua Correll had untrained civilians play a simple video game in which players had to decide whether to shoot at people based on whether they were armed or unarmed; what he found was that participants were more likely to mistakenly shoot an unarmed target who was black, and more likely to holster their weapon when faced with an armed target who was white.

It’s tempting to think that simply being told that you harbor biases is enough to get them under control. Not so, says Yale University psychologist John Dovidio, who has spent decades studying what he calls “aversive racism” among people who profess to believe in egalitarian principles but still engage in racist behavior. “Making people aware of it is not sufficient,” Dovidio said. “There’s no ‘insight cure’ for it....You actually have to teach people ways to suppress their biases.”

How do you do that? The 16 methods that were submitted to Calvin Lai when he announced his contest ran the gamut, building on ideas that had been percolating in the field over the past decade. One approach involved prompting people to empathize with specific individuals they might otherwise be prejudiced against; another tried to instill a sense of common humanity and an appreciation for the values of different cultures; a third asked them to think about the injustices that whites have visited upon black people throughout history.

When Lai tested them, he found that only a few approaches were actually effective at reducing bias, which he measured by administering the IAT to the study participants before and after. The most effective, Lai found, involved exposing people to so-called counter-stereotypical images. In one intervention, which echoed the breakthrough study by Dasgupta and Greenwald from 2001, this took the form of showing people photos of widely admired black celebrities like Bill Cosby alongside notorious white evildoers like Charles Manson. In another effective approach, test subjects listened to stories, told in the second person, about a white assailant attempting to hurt them and a black man coming to their rescue. The emotional pull of the experience seemed to be key. Researchers found that making the story longer and more vivid—changing it from “With sadistic pleasure, he bashes you with his bat again and again” to “With sadistic pleasure, he beats you again and again. First to the body, then to the head. You fight to keep your eyes open and your hands up. The last things you remember are the faint smells of alcohol and chewing tobacco and his wicked grin”—was doubly effective at reducing bias.

Another approach that worked well when Lai tested it involved telling participants to imagine a scenario in which they were playing a game of dodge ball in which everyone on their team was black while everyone on the opposing team was white. A similar effective intervention also had participants imagine themselves navigating a highly threatening post-apocalyptic scenario, before being shown photos of their “friends,” who were mostly black, and their “enemies,” who were all white.

Though Lai was encouraged to find that eight of the ideas he tested worked, one caveat is that nobody knows how long their effects last. “A lot of the studies that have found that implicit bias can be changed have looked at it in the context of an hour at most,” Lai said. “It’s not at all clear whether this is a temporary change that’s just reflecting malleability, or if it’s a permanent change.” He said he and his adviser, Brian Nosek, who also worked on the contest study, are currently running tests in which subjects return 24 hours after “treatment” to see how strongly their original bias has returned.

At least one recent experiment, carried out over 12 weeks, offers hope that lasting change is possible. By alerting a group of psychology students to their prejudice—90 percent of them showed antiblack bias at the beginning of the intervention—and teaching them a range of de-biasing strategies they could employ on their own time, University of Wisconsin-Madison psychologist Patricia Devine and her team showed that prejudicial attitudes could, with sustained effort, go down and stay down for at least two months. In a 2012 paper published in the Journal of Experimental Social Psychology, Devine concluded that unconscious prejudice could be unlearned, like a bad habit, through “the power of the conscious mind.”

It’s one thing to succeed at getting people to do better on a test by subjecting them to lab experiments, and another thing entirely to apply the lessons learned to real life. At the moment there are at least two teams of experts trying to do just that in law enforcement.

One of them is led by Lorie Fridell, an associate professor at the University of South Florida who has developed a suite of training programs to carry out in departments around the country. The challenges of such programs are apparent from the outset: Attending a de-biasing session can make people feel like they’re being accused of racism and need to change, a message that doesn’t go down easily. In light of that, Fridell’s philosophy is that police officers need to be told the biases they harbor are normal and don’t reflect in any way on their character. “When we walk into a room, my trainers and I, they are somewhere between defensive and hostile....Their arms are crossed and often they’re glaring at us,” she said.“As soon as they come to understand that we’re talking to them about how their mind works, and that even well-intentioned people have biases...they start to relax and open up.”

Fridell’s training methods involve, among other things, briefing officers on the science of unconscious bias, and encouraging them to pause from time to time and ask themselves if they’d be doing whatever they’re doing if the person they’re dealing with belonged to a different race. While Fridell is well established in her field—she has received numerous grants from the Department of Justice—one important disclaimer is that she has never produced any empirical results to show that her methods actually reduce discriminatory policing. (In Fridell’s view, this goes with the territory: There’s no way to accurately measure biased policing, because it’s impossible to look at the actions of an individual police officer and know for sure whether it was influenced by bias or not.)

Generating some kind of evidence is going to be crucial going forward, according to Phillip Atiba Goff, an associate professor of psychology at UCLA who also heads an effort to help police departments drive down racial discrimination. Goff, who cofounded the Center for Policing Equity and is currently a visiting scholar at the Harvard Kennedy School of Government, has conducted research on possible training methods by working with police departments in Chicago, Denver, Las Vegas, and many other American cities.

Goff is collecting before-and-after statistics on things like traffic stops, arrests, and use of force by the officers he has worked with. Though he hasn’t published any results yet, he says his experience so far has led him to focus more on real-world situations than on the abstract question of internal bias. “There’s a difference between bias and discrimination, and as a culture we’re really bad at making the distinction,” he says. What Goff aims to do, based on that insight, is identify types of encounters where bias is likely to turn into active discrimination: important decisions like whether to arrest someone, or when to fire a weapon. Goff finds that discrimination is particularly likely to arise in the face of “identity threat”—when officers feel like their intelligence, masculinity, or moral legitimacy is being attacked.