Are You An Unconscious Racist?

What does it mean to be a racist?

• A person who thinks their “race” or ethnic group is better than everyone else’s by virtue of genetic superiority, religion, customs, food, way of life, or beliefs. 
• A person who fails to hire an applicant with the best qualifications if that person is from a different ethnic or religious group from the employer’s. 
• A person who is part of an institution that requires him or her to systematically target and discriminate against African Americans or other minorities. 
• A person of any race, ethnicity, or religion who feels more comfortable with others who are like themselves. 

Or: 

• A person whose score on the Implicit Association Test (IAT) reveals that he or she is unconsciously biased against Black people. 

Some of the above? All of the above? 

Throughout the first decade of this century, surveys repeatedly found that prejudiced attitudes—notably the once-common beliefs that Blacks were inferior to Whites, women inferior to men, gay men and lesbians inferior to straights—had declined sharply, especially among young people. Surveys, of course, supposedly assess what you think. But what if they assess what you think othersthink you should think? What if they simply reflect your awareness that it isn’t cool to reveal your actual negative feelings about another group? Self-report data is inherently plagued with this problem. Thus, most social psychologists who study prejudice and discrimination focus on what people do, not what they say they might do. For example, when researchers have sent identical résumés to potential employers, varying only a name that indicates gender, or implies race (a Black-sounding name or membership in an African American organization), or mentions religious affiliation, many employers have revealed a bias in whom they choose to call for an interview.¹

Of course, whether or not you choose to tell an interviewer that you would never willingly hire a [fill in the target person], you know what you feel about “those people.” But some researchers have set their sights on capturing the prejudices that they believe lurk below awareness, hoping to identify implicit, unconscious negative feelings—not only in people who know they are prejudiced but don’t want to admit it, but also among people who believe they are unprejudiced. 

We can’t let our wish for a method or a finding to be right block our ability to evaluate it critically, and to change our minds when the evidence dictates.

Nearly twenty years ago, a team of eminent psychological scientists, including Anthony Greenwald and Mahzarin Banaji, developed the Implicit Association Test, which measures the speed of people’s positive and negative associations to a target group.² You sit at a console or your computer and are shown a series of faces you must sort as quickly as you can—pressing a left key for a Black face, say, and a right key for a White face. Now you have to do the same for a series of positive or negative words—press the left key for positive words (such as triumph, joy, honest) and the right key for negative words (such as devil, maggot, failure). Once you’ve mastered these sorting tasks, the faces and words are combined: Now, as quickly as possible, you must press the left key when you see a Black face or a positive word and the right key when you see a White face or a negative word. You are given a rapid set of combinations: black + triumph, black + poison, white + peace, white + hatred. The pairings get harder as you go along. Many people respond more quickly when White faces are paired with positive words and when Black faces are paired with negative words. That speed difference is said to be a measure of their implicitly racist attitudes toward African Americans because it’s harder for their unconscious minds to link African Americans with positive words. 

When the research first appeared all those years ago, my colleague Carole Wade and I were disinclined to report it in our introductory psychology textbook. It was unclear what those microsecond “associations” meant; it seemed a leap to call it a measure of prejudice; at best it seemed simply to be capturing a familiar cultural association or stereotype, in the same way that people would be quicker to pair bread + butter than bread + avocado. A person of any age might be aware of negative associations between old people and mental decline without being prejudiced against old people in general. One team got an IAT effect by matching target faces with nonsense words and neutral words that had no evaluative connotations at all. They concluded that the IAT does not measure emotional evaluations of the target but rather the salience of the word associated with it—how much it stands out—and negative words attract more attention. When they corrected for these factors, the presumed unconscious prejudice faded away.³

So Carole and I figured that the IAT would travel the route of other hot ideas that cooled off in the face of failed replications or more plausible interpretations. Indeed, had the measure’s originators simply said they had found a modest but interesting association between various groups and words culturally linked with them, that might have been that. Instead, over the years, the success of the IAT grew so rapidly, spilling into the public arena with such an enormous splash, that textbook authors (let alone the public, college administrators, and politicians) could no longer avoid it. To date, more than 17 million people have taken the test online (at “Project Implicit”), and it has also been given to students, business managers, employees, and countless others to identify their alleged prejudices toward Blacks, Asians, women, old people, people with disabilities, and other groups. I asked a young friend, one of the least prejudiced and most open-minded humans on the planet, if he knew about the IAT, and he said yes, he’d taken it online. “What in the world for?” I asked. “To see if I have an unconscious bias,” he said. 

Carole and I were awakened from our somnambulance by Malcolm Gladwell, who set off our skeptical buzzers on high alert. Gladwell, who is biracial, took the IAT and learned that he was prejudiced against Black people. In an interview with Oprah Winfrey, he said: “The person in my life [his mother] who I love more than almost anyone else, is Black, and here I was taking a test, which said, frankly, I wasn’t too crazy about Black people, you know?” A gay activist said she was stunned to learn that “her own mind contained stronger gay = bad associations than gay = good associations.” But the pièce de résistance was that one of the developers of the IAT, Mahzarin Banaji, a woman of color who was born and raised in India, reported that she herself “failed” the racial IAT, revealing anti-Black associations that she consciously repudiates.⁴

How do I know what my unconscious is thinking? By definition, it’s unconscious! Clearly, the test knows more than I do, even if I do love my mother! 

Now this is curious. Why jump to indict oneself instead of saying, “Uh oh, maybe something’s wrong with the test?” Banaji and Greenwald might have considered the possibility that they got too enthusiastic too soon, claiming more for their test than it warranted. However, by the time their 2013 book for the public appeared, Blindspot: The Hidden Biases of Good People, they had much invested in the IAT; and, in their admirable zeal to educate people about the persistence of prejudice, they began claiming more and more about the test’s significance and relevance. For their part, many laypeople accepted the IAT’s findings because of the great power attributed to methods that purport to uncover what our brains are doing without our knowledge. How do I know what my unconscious is thinking? By definition, it’s unconscious! Clearly, the test knows more than I do, even if I do love my mother! 

If the IAT were being used solely as an instrument to generate discussions of what prejudice is and is not, few would object. Having a “hidden bias” is not, in and of itself, a sign of prejudice; it’s a sign of having a human brain. What social psychologists call the “in-group bias”—a feeling of comfort with, and preference for, people who are like us—is a universal phenomenon, undoubtedly one that evolved to aid human survival by binding us to our groups. But it does not inevitably produce “out-group hostility” or discrimination against those who are unlike us. 

Unfortunately, as the IAT’s reputation grew, claims about what this test was revealing began to outstrip the data. The IAT was no longer said to be capturing an “association”; not even merely a “bias”; but a prejudice—especially racism. And not just racism, but discrimination—the willingness to act on that prejudice. The Project Implicit website warns: “When we relax our active efforts to be egalitarian, our implicit biases can lead to discriminatory behavior, so it is critical to be mindful of this possibility if we want to avoid prejudice and discrimination.” 

And so we come to the crux of the matter: does the IAT really capture unconscious prejudices? Can the test predict whether people will actually behave in a biased or discriminatory way? The evidence is now pretty clear that the answers to both are “no.”⁵ When people are asked to predict their responses toward different groups on the IAT, they are highly accurate—regardless of whether they were told that implicit attitudes are true prejudices or culturally learned associations. People’s scores aren’t reliable, either; they might score “highly biased” one week and get a different result two weeks later. And as for the IAT’s ability to predict behavior—the ultimate measure of any test’s scientific validity—meta-analyses of hundreds of studies on many thousands of people find that the evidence linking IAT scores with behavior is weak to nonexistent. “The IAT provides little insight into who will discriminate against whom,” wrote Frederick Oswald and his colleagues in a 2013 meta-analysis, and it provides no more information than just asking people if they are biased.⁶ They concluded that the correlation between people’s IAT scores and their behavior is so small as to be trivial. Greenwald and Banaji countered that statistically small effects can have “societally large effects.”⁷

It confuses normal cognitive biases with bigotry. And it locates the problem of discrimination in people’s unconscious minds, not in the systemic patterns of racism that deserve our far greater attention and search for remedies. 

In the final analysis, I think what is most problematic about the IAT is that it directs people’s attention to their supposed unconscious feelings, leaving many puzzled and worried that they might be awful racists without knowing it, and without knowing what they are supposed to do about it. It confuses normal cognitive biases with bigotry. And it locates the problem of discrimination in people’s unconscious minds, not in the systemic patterns of racism that deserve our far greater attention and search for remedies. 

In his meticulous investigation of the IAT, Jesse Singal concluded that “after almost 20 years and millions of dollars’ worth of IAT research, the test has a markedly unimpressive track record relative to the attention and acclaim it has garnered. Leading IAT researchers haven’t produced interventions that can reduce racism or blunt its impact. They haven’t told a clear, credible story of how implicit bias, as measured by the IAT, affects the real world. They have flip-flopped on important, baseline questions about what their test is or isn’t measuring. And because the IAT and the study of implicit bias have become so tightly coupled, the test’s weaknesses have caused collateral damage to public and academic understanding of the broader concept itself…[S]crutinizing the IAT and holding it to the same standards as any other psychological instrument isn’t a sign that someone doesn’t take racism seriously: It’s exactly the opposite.” 

How ironic that this well-intended effort to illuminate a dark side of our natures now obfuscates the very thing we’re trying to understand. And it’s a story with an all-too-familiar lesson for scientists and other skeptics: we can’t let our wish for a method or a finding to be right block our ability to evaluate it critically, and to change our minds when the evidence dictates.