Twentieth Century Fox/Everett Collection

Gregory Peck, far left, as a journalist who goes undercover as a Jew for an exposé on anti-Semitism in Elia Kazan’s Gentleman’s Agreement, 1947. Also pictured are Celeste Holm, John Garfield, and Robert Karnes.

In his ill-fated attempt to win a Senate seat in 1858, Abraham Lincoln squared off against the incumbent, Democrat Stephen Douglas, in a series of seven debates that centered on the issue of slavery. Lincoln, of course, was the candidate in favor of equal treatment under the law for the black populace. Here is an excerpt from those debates, in which Lincoln reveals his feelings about black people themselves, aligning himself as much as possible with the political correctness of his time:

There is a physical difference between the white and black races which will ever forbid the two races living together on terms of social and political equality. And inasmuch as they cannot so live, while they do remain together, there must be the position of superior & inferior. I am as much as any other man in favor of having the superior position assigned to the white race.

This doesn’t sound like the man who freed the slaves, but Lincoln didn’t see an inconsistency in advocating for civil rights while holding a belief in white supremacy. As he put it, “I do not perceive that because the white man is to have the superior position the negro should be denied everything.”1

Today, of course, things are different. Today we have a president whose father was black, and remarks that are taken to indicate racial, ethnic, or gender bigotry can cause great damage to a public figure’s career, for example former Harvard president Lawrence Summers, radio host Don Imus, Nobel Prize winner James Watson, actor Mel Gibson, radio talk show host Laura Schlessinger, news analyst Juan Williams, and comedian Michael Richards. Yet our society still exhibits a stubborn tendency to discriminate, not just on the basis of race, but also gender, religion, ethnic group, and body weight, just to name a few. If open statement of prejudice by well-known people provokes immediate public disapproval, privately held contempt can be pervasive and strong.

In their new book, Blindspot: Hidden Biases of Good People, social psychologists Mahzarin R. Banaji and Anthony G. Greenwald examine the nature of today’s social biases, and the difficulty we face in erasing them. The authors’ central point is that most of us are biased toward various groups. Moreover, though some of us are aware of being prejudiced, and some of us publicly express bigoted views, the authors assert that far more of us hold prejudices seated in a deep level of our minds that is inaccessible to our conscious awareness. The attitudes lurking in that blind spot, they say, have an important part in perpetuating discrimination.

Banaji’s and Greenwald’s view aligns the study of prejudice with a larger movement that has transformed academic psychology in recent years. “A quarter century ago,” as the authors put it,

most psychologists believed that human behavior was primarily guided by conscious thoughts and feelings. Nowadays the majority will readily agree that much of human judgment and behavior is produced with little conscious thought.

Banaji and Greenwald attribute this revolution in psychology to new research methods aimed at revealing mental processes that are said to be beyond the reach of personal introspection. With regard to our understanding of bias, the most telling of these is the Implicit Association Test, or IAT, first described in a 1998 article in the Journal of Personality and Social Psychology. The authors have based Blindspot on data drawn from the IAT. They know the test well: it was Greenwald himself who invented it.

The field of psychology as we know it today arose from the work of people like Wilhelm Wundt, who, in 1879, was denied German government funding for the first laboratory of psychology, but started one anyway, in a small classroom; and William James, who, around the same time, set up an informal psychology laboratory in two basement rooms of Lawrence Hall at Harvard. They and other pioneers debated the function and importance of the unconscious, but while Freud took that idea and ran with it in his clinical work, scientific psychology soon grew to largely ignore the unconscious in favor of mental processes of which we are well aware. Thus, Banaji and Greenwald tell us, when psychologists began studying discrimination in the 1920s, they did it by asking people directly about their prejudices.

The questionnaires were not subtle. One study, developed by psychologist E.D. Hinckley, introduced a psychological tool called the “Attitude Toward the Negro” scale. In that study, Hinckley asked his subjects to indicate whether they agreed or disagreed with thirty-two statements. Banaji and Greenwald don’t reveal any of the results but the statements themselves reveal much about prejudices in American culture at that time. For example, statement number six: “The feeble-mindedness of the Negro limits him to a social level just a little above that of the higher animals.”


Observing the clear pictures Americans seemed to have regarding the nature of people they had never met, in 1922 Walter Lippmann co-opted the term “stereotype” from the printing business. In printing it referred to a process by which multiple metal plates were created, enabling the mass printing of newspapers and books on several presses at once. For human psychology, it was, as he explained, a way of dealing with complexity. “The real environment is altogether too big, too complex, and too fleeting for direct acquaintance,” he wrote. “And although we have to act in that environment, we have to reconstruct it on a simpler model before we can manage with it.”2 That simpler model is the stereotype.

The “recognized starting point for modern understanding of stereotypes,” the authors tell us, was Gordon Allport’s 1954 book The Nature of Prejudice. Allport wrote: “The human mind must think with the aid of categories…. Once formed, categories are the basis for normal prejudgment. We cannot possibly avoid this process. Orderly living depends upon it.”3

Imagine what life would be like if we didn’t categorize, if we treated each chair, each apple, each taxi we encountered as an individual, a blank slate whose character and purpose we had to decipher anew. We wouldn’t get very far, nor would we have lasted long on the ancient savannah if we paused to consider the intentions of each individual lion we encountered. Instead, once we’ve seen a single lion eat a human, we form a prejudice against the entire species.

The problems that occur when a human mind fails to see the world in categories were sadly and perhaps uniquely dramatized by the case of a shopkeeper in London in the 1980s who had a stroke that affected the lower part of his occipital lobe. His motor skills and cognitive function seemed unimpaired, except in one respect: if shown two objects that had the same function, such as two brooms or two beer mugs, he failed to associate them. As a result he had great difficulties in everyday life, even when attempting simple tasks such as setting the table, or when reading—for deciphering the printed word involves an understanding that a and A, though different, are in some sense identical.4

Lippmann recognized that although shortcuts are useful, even necessary, when applied to people, stereotypes can be dangerous. Today, many polls report that Americans are in favor of a tolerant society in which people are judged on their merits, and think it is wrong to assess people according to the social and ethnic categories to which they belong.5 Why then does discrimination based on class and ethnicity appear to be endemic in American society?

The authors’ answer is that biases that are often hidden influence our judgments and actions in many of the situations that we typically encounter in life. Such biases are so prevalent, they assert, that even people who in their conscious thoughts abhor prejudices commonly hold them—so prevalent, even, that members of negatively stereotyped groups are commonly biased against their own social group.

One example Banaji and Greenwald offer concerns race: they report that about 75 percent of Americans have an unconscious, automatic preference for whites over blacks. A similar percentage, they say, are prone to stereotype by gender. And significant numbers also show bias with regard to “sexual orientation, and age as well as body weight, height, disability, and nationality.”

Though their research on prejudice has been accepted by many psychologists, Banaji and Greenwald are aware that many people may be skeptical of their claims. “The influence of Freud notwithstanding,” they admit, “it is hard for human beings, endowed with the capacity for conscious thought, to accept that the beliefs and preferences that so define us can be shaped by forces outside our awareness.” The evidence that the unconscious associations they speak of do exist comes from Greenwald’s IAT. To date, Greenwald and his colleagues have administered the test over 14 million times through the Internet (you can take it at

After introducing us to the idea of unconscious bias, the middle chapters of Blindspot are something of a hodgepodge in which the authors describe the IAT and the significance of what it tells us. In all those chapters, we never stray far from a discussion of the IAT. To get a feeling for the test, imagine holding a deck of cards face up and sorting the cards as quickly as you can into two piles, with hearts and diamonds on the left, and clubs and spades on the right. Now imagine repeating the task, this time placing hearts and spades on the left, and diamonds and clubs on the right.


It is probably no surprise that if you actually performed these two tasks, it would take you longer to get through the second one. That’s because you naturally make a mental association between diamonds and hearts, and between clubs and spades, based on their colors. That is the basis of the IAT—you can sort things faster if you are sorting them in a manner consistent with your mental associations than if you are sorting them against dispositions you have already acquired. As Banaji and Greenwald put it, the effectiveness of the IAT “relies on the fact that your brain has stored years of past experiences that you cannot set aside when you do the IAT’s sorting tasks.”

In the race version of the Implicit Association Test, you are asked to sort a mix of recognizably African-American and European faces, along with a mix of pleasant (“gentle,” “heaven”) and unpleasant (“hurt,” “anguish”) words. You are told to sort the European faces and unpleasant words to the left, say, and the African-American faces and pleasant words to the right—as quickly as possible. You are then asked to sort them again, this time with the European faces and pleasant words to one side, and the African-American faces and unpleasant words to the other. If you find the first task is more difficult (and thus takes more time), then, just as you associate red with diamonds and hearts, so, too, do you associate African-Americans with unpleasantness. It is the comparative timing of the two tasks that is telling. Since developing the race IAT, Greenwald has developed IATs that probe attitudes toward many other social groups.


Banaji and Greenwald say the IAT reveals that the division is highly permeable between the culture “out there” and the content of our minds “in here.” Thus, “Whether we want them to or not, the attitudes of the culture at large infiltrate us.” Hence the potential for bias against even one’s own social group. In one dramatic example, Banaji and Greenwald tell us of a gay activist who took the IAT and was stunned to learn that “her own mind contained stronger gay=bad associations than gay=good associations.” Or consider Malcolm Gladwell, who after taking the IAT told Oprah Winfrey in an interview: “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.”

Given the diversity as well as the ubiquity of the biases the IAT has revealed, it seems that few of us are immune from holding some mental associations of which our conscious minds would not approve. One unwanted bias the authors say affects almost everybody is bias against the elderly. We may love our patient, warm, and gentle grandparents, but apparently, in the case of 80 percent of us, our unconscious, to paraphrase Gladwell, isn’t too crazy about old people. Only 6 percent of those who took the ageism IAT showed a stronger association between old=good than between young=good.

The research on ageism is especially interesting because “the elderly” is unlike other social groups in that most of us end life as members of it, but none of us started life that way. You’d think, therefore, that our attitudes toward age would change as we grow older, but the IAT results surprisingly suggest that this is not so. Instead, the IAT reveals that even the elderly have a strong anti-elderly bias.

How do the elderly deal with the conflict between knowing they are old and preferring the young? Banaji and Greenwald provide the answer through a University of Kansas study showing that older Americans “had automatic associations of self=young that overpowered any self=old identification.” Apparently, Banaji and Greenwald tell us, eighty-year-olds are sincere when they say, “Inside I feel like I’m eighteen.”

In an apt image, the authors say the IAT is a mirror in which people see a reflection of an inner self that they do not recognize. Technically, that mirror reflects “automatic preferences” and “implicit bias.” But to have such preferences and biases is not necessarily to act upon them. So if the IAT is to be more than an exercise in academic wheel spinning, a key question is: Do mere associations and unconscious biases affect people’s behavior?

A large body of evidence, Banaji and Greenwald report, suggests that the answer is “yes.” For example, a statistical analysis of 184 studies on racial bias showed that the IAT “predicted discriminatory judgments and behaviors significantly more effectively than did the types of question-asking measures that had long been used in studies of prejudice.” That may be impressive, and the authors go into some depth regarding the mathematics of that statistical survey, but it would have been more enlightening had they instead focused on the more important issue of how the bias they measure is translated into behavior.

In one of the few behavioral studies of bias that Banaji and Greenwald do describe, young professionals were asked, “Do you care if your boss is male or female?” Their answers were “a resounding no.” Instead, they named salary, location, and the boss’s personality as factors that are important in making a choice of job. But when presented with a series of job choices, a statistical analysis of the subjects’ responses suggested that the subjects were willing to absorb a salary penalty of $3,400 on jobs with an average salary of $42,500 if it meant reporting to a male rather than to a female.

“Remarkably,” Banaji and Greenwald write, “male and female respondents were equally willing to give up a chunk of salary for the pleasure of having a male boss, even though they had sworn to having no such preference.” These job seekers, they stress—men and women alike—were not making a conscious decision, but rather acting in a manner “they presumably would have opted to avoid had they been aware of the patterns of their choices.”

Banaji and Greenwald also list a handful of other behaviors that studies suggest may be related to the unconscious negative associations measured by the IAT:

In a simulated hiring situation, judging White job applicants more favorably than equally qualified Black applicants; emergency room and resident physicians recommending the optimal treatment—thrombolytic (blood-clot dissolving) therapy—less often for a Black patient than for a White patient who presented with the same acute cardiac symptoms; and college students being more ready to perceive anger in Black faces than in White faces.

Though the IAT is usually taken on a computer, Blindspot sprinkles through its middle chapters several paper IATs that the reader can take to test his or her own unconscious associations. The authors include a “note of caution.” If you don’t want to risk discovering a surprising—and presumably unwelcome—truth about yourself, they say, “you might want to avoid this IAT.” That sounds a bit melodramatic, but, as Malcolm Gladwell put it, seeing your implicit bias reflected by the mirror that is the IAT can be a “creepy, dispiriting, devastating moment.” Indeed, to those unfamiliar with the IAT, the picture of rampant bias that Banaji and Greenwald paint could well be both surprising and disturbing.

In Dr. Seuss’s The Sneetches, a group called the Star-Bellies consider themselves naturally superior and refuse to have anything to do with the Plain-Bellies, who lack stars. An entrepreneur named Sylvester McMonkey McBean invents a Star-On machine that can transform a Plain-Belly to a Star-Belly for a small fee. This throws the social system into a bit of chaos, for it becomes difficult to decide whom to discriminate against. McBean has a remedy for that too—a Star-Off machine that can create a new elite—this time the Plain-Bellies—for a slightly larger fee.

“In the Dr. Seuss story,” Banaji and Greenwald write, “McBean exits laughing and rich, convinced that the Sneetches will never learn. But he is wrong, for by the end of the book they do come to understand that, with or without stars, ‘Sneetches are Sneetches.’”

If we hold biases that are automatic and outside of our awareness and control, can we, too, hope to eventually achieve a just and fair society? Can we overcome our biases in hiring, in providing access to education and health care, and in other social interactions? The final chapter of Blindspot is meant to address those questions. It is called “Outsmarting the Machine,” the machine being our unconscious.

In reading this chapter, I was hoping for a discussion of the factors that contribute to some people holding unconscious biases while others do not. If there is any such research, the authors do not mention it. Instead, Banaji and Greenwald tell us that they can offer “a few good weapons.”

The authors actually offer just two weapons, neither of which is in most cases practical. They call the first the “blinding method.” You simply arrange not to actually meet the person you are interacting with. They give the example of symphony orchestras that erect a partition so the judges cannot determine the gender of those auditioning. That method doubled the number of women hired by major orchestras, but it is hard to imagine anyone interviewing a prospective corporate vice-president or a union leader from the other side of a shoji screen.

Their other strategy is the “no-brainer” method, in which you blindly apply a set of fixed rules to eliminate human discretion from whatever decision is at hand, like the computer programs that diagnose disease based on symptoms and other patient data. Again, this seems to be quite limited in its applicability.

The real hope for further progress in overcoming our biases probably lies elsewhere. Banaji and Greenwald didn’t mention Walter Lippmann’s concerns about the media being the source of much of our bias, but in their final chapter they do mention that the mass media are potentially rich sources of “counterstereotypic” role models.

Film and television continue to regularly stereotype blacks, Hispanics, gays, and other social groups, but as those groups make progress, the news media, at least, cannot help but portray those “counterstereotypic” individuals. There are many blacks, for example, in high positions in today’s society, and when they appear in the press, that can only help to erode negative black stereotypes. One would hope that this erosion would diminish the obstacles facing African-Americans in the future, leading to even more well-publicized success stories. Such increasing encounters might eventually lead us away from discrimination in all its forms, especially if parents and teachers can do more to promote egalitarian values among children. It would, however, be naive to think that the accomplishments of the well known and well-to-do will by themselves soon erase prejudices.

Blindspot is a brief book, whose contribution is an in-depth description of the IAT and its implications for the struggle to attain social equality. Despite the book’s shortcomings, the authors explain their ideas clearly, anticipate questions that might arise in the reader’s mind, and take their time with points that might cause confusion. Their conversational style makes the book easy to read, and best of all, it has the potential, at least, to change the way you think about yourself and those around you.