Skip to main content

Verified by Psychology Today

Jay Richards Ph.D.

Race in America: The Invisible Hand of the Implicit Mind

With Anthony Greenwald, co-author of Blindspot: The Hidden Biases of Good People

Mahzarin Banaji & Anthony Greenwald
Source: Mahzarin Banaji & Anthony Greenwald

Over the remainder of this century, judgements about the prominence and impact of race in American society will need to take into account a series of recent critical events. The outright social rebellions in Ferguson and Baltimore, the racially motivated massacre in Charleston, and the ever-continuing series of unarmed black men, women, and children being killed by the police will continue to have important ramifications. The shocking truth is that these events have occurred while the residents of the White House were an African American family. Once, undisguised expressions of prejudice and racial antagonism were rife throughout American society, but since the Civil Rights Era racial vitriol has virtually withered away.

Today only a small minority of Americans endorse any form of anti-black sentiment. If old-fashioned racism is clearly not a viable cause, why are outcomes for Blacks increasingly worse than those for Whites in so many important many dimensions of life? And why is the current state of affairs in race relations—epitomized by policing, incarceration, and unemployment—viewed so differently by Black Americans and White Americans?

I believe that some important answers to these questions can be found in the unconscious biases that the great majority of us unknowingly carry with us. In their new book, Blindspot: Hidden Biases of Good People, Dr. Anthony Greenwald, professor of social psychology at the University of Washington and Dr. Mahzarin Banaji, a Yale University social psychologist, share the results of 30 years of psychological research to provide a deeper understanding of our current racial gaps.

According to their research, otherwise “good” people who wouldn’t ever consider themselves to be racist, sexist, agist, etc, nevertheless, have hidden biases about race, gender, sexuality, disability status, and age. These biases come from a part of the mind that functions automatically and efficiently, and does its work outside our conscious awareness. If asked if we held these beliefs or attitudes, we would often disown them, but they nonetheless have a powerful and pervasive impact on our decisions and behavior.

I had an in-depth conversation with Dr. Greenwald about the often-surprising insights from Blindspot.

JR: What inspired you to write Blindspot?

AG: In the mid-1990s, my co-author Mahzarin Banaji, Brian Nosek (another researcher from University of Virginia), and I created the Implicit Association Test (IAT) to test people’s unconscious biases and stereotypes. The IAT has produced some very robust and very intriguing results. So many people were interested that we felt we had to get something out that was informative, readable, and that would point out some of the implications of this kind of research.

JR: The IAT is not just another pencil and paper questionnaire. Can you explain what kind of test it is and how it is able to measure biases that an individual is not aware of having?

AG: Yes, but the quickest way to learn about how the IAT works it to take one of the tests. The race test is on the Project Implicit website and only takes a few minutes. There are also printed IAT examples in Blindspot that you can take and score.

In a nutshell, the IAT is a two-part task involving responding to a series of words and faces that appear on a computer screen. The words are either pleasant or unpleasant and the faces are faces of black or white people. On the first part of the IAT you are asked to make the same response (push the same key) when either a white face or a pleasant word appears on the screen and to push a different key if a black face or an unpleasant word appears. You try to do this as fast as you can without making errors. In the second part, you have new instructions. Now white faces and unpleasant words are keyed together, and you respond to black faces and pleasant words using a different key. The difference between the time it takes to do the two trials is a measure of preference. If, like many people, you are faster when white faces and pleasant words are keyed together than when black faces are keyed with pleasant words, you have an automatic bias in favor of viewing white faces, and white people, more favorably than black people.

When I created and tried this task in about 1995, I was pretty amazed how much faster I was on one than on the other.

JR: This is one of those aha moments in science when the scientist tries the invention on himself.

AG: I found that I could put white faces and pleasant words together much faster than I could put black faces and pleasant words together. I told myself that this was just a matter practice. But the time difference didn’t change with more practice. I've taken the test literally a hundred times over the last 20 years and my scores haven't changed very much. I thought this was really interesting, because my test results were telling me that there was something in my mind that I didn't even know was there before.

JR: What a surprises readers most about what's in the book?

AG: The thing that has been most challenging for readers and others who have taken the IAT, is the pervasiveness of the biases that are revealed in the research we do. When I say pervasive, I don’t just mean the number of people who hold these biases. There’s also a very wide range of different implicit attitudes, such as liking whites more than blacks, young more than old, Americans more than Asians, and a whole lot more. The extremity of the data is also surprising. For example, the Implicit Association Test shows that 70% of people prefer younger people over older people, and this implicit age bias is held just as strongly in people 70 or 80 years old as it is for people in their 20s and 30s.

JR: In our recent conversations, you have referred to psychology undergoing an Implicit Revolution. Can you tell us about this development?

AG: Yes and this revolution is in part responsible for the origin of the Implicit Associations Test, which is an earlier form of our Implicit Attitudes Test. It began in the early 1980s when cognitive psychologists were studying memory, and discovered new methods (or actually resuscitated some older methods) to demonstrate that people could remember things they were not aware of remembering. This took the form of performing “judgment tasks” which indicated that they had taken something from an experience, but didn't remember the experience itself. This kind of memory was called implicit memory, a term which was popularized in the late 1980s by Dan Schacter who is a professor at Harvard.

Mahzarin and I were starting to be very interested in this research and we thought we should be able to apply it to social psychology. So we started to develop a means of measuring implicit attitudes and stereotypes. We spent several years trying to find a method that would work with human subjects, which at the time were mainly college sophomores from Ohio State University, University of Washington, Yale and Harvard. We were successful and saw that having an understanding of the implicit aspect of our minds had vast potential.

This implicit research has been so successful, in fact, that it has led to a paradigm shift in psychology. And it is still gathering strength 25 years after it started in the field of memory. About 5 years ago, I decided that we needed a name for this paradigm shift, so I started calling it the Implicit Revolution. This is not yet a catch word that you will find everywhere. In fact, I haven't even published anything trying to proclaim it as a label for what's going on now and it wasn't even included in Blindspot. But I think it's a real thing.

JR: And what do you mean by “implicit”?

AG: The mind does things automatically that feed into our conscious thought and provide the basis for judgments. The result is that we make conscious judgments that are guided by things that are outside of our awareness. We only get the end products, and we don't recognize the extent to which those products have been altered by our past experience. That's where those biases and stereotypes come in.

JR: I have heard this referred to as different levels of consciousness is that language that you would use to describe it?

AG: Yes these levels have been described in different ways, but what is important is the idea that there are levels. There is a slower, automatically operating level that is outside of awareness, and a higher attentional level that can operate deliberately and rationally with conscious intention. That's the distinction that actually defines the Implicit Revolution. We are elevating this lower level—the implicit level, the automatic level, the intuitive level—to a prominence that corresponds with the importance of the work it does.

JR: So if I understand you correctly, when we’re perceiving things, those thoughts and perceptions are really end-products of unconscious processes? We’re not really aware of the “sausage-making” that went into creating these end products of thought and perception?

AG: That’s a great metaphor. Another example I like to use to explain this distinction is that of a Google search. When you look something up in Google, advertisements just sort of pop-up on your computer screen that relate to what you were looking for. Every time we enter a query into a search engine there are very rapid and invisible processes that we couldn't even begin to follow. All we see is the end-product which shows up on the screen. That distinction between the behind the screen level which operates very rapidly and what see on the screen, that we can read and interpret and make use of, corresponds to the two levels were talking about now in psychology.

JR: Stereotype is a term that is central to your work. We use it a lot, but I am not sure that we always have a clear idea of what it means. How do you use the term stereotype in your work?

AG: The term stereotype originated as a psychological term in the writings of the journalist Walter Lippmann. It came from a printer’s term that referred to a block of metal with a page of type engraved on it that could be used to stamp out numerous successive copies, each identical to the other. Walter Lippmann used stereotype to refer to the mind stamping out a social image for everyone in a certain category, such as age, ethnicity, gender, or others we now attach the term stereotype. When a stereotype is used to understand people, everyone in a social category is seen as sharing the same properties. To the extent that we see all women, all older people, all disabled people, all Italians as having as having shared characteristics we are using this identical mold that Lippmann was describing like the one in the printing process. Stereotypes effectively obliterate the differences among people in each category, and instead focus only on the qualities they share.

JR: I’ve heard stereotypes typified as being a form of lazy thinking. What do you think about the age old statement that stereotypes have a kernel of truth?

AG: I think that they often do. I have a stereotype that Boston drivers are a bit out of control. While I think there’s a real kernel of truth to it, I don't want to think that all Boston driver are wild people and you should try to keep off the road in that city The kernel of truth is usually an average difference between one group and another group. For example, there is obviously truth to the gender stereotype that men are tall relative to women. But that doesn't mean that every man is taller than every women. The problem with stereotypes is when we ignore the individual differences between people within the category. So yes, there's a kernel of truth to stereotypes, but we lose truth when we allow them to dominate our perceptions to such a degree that we don't see the individual differences between people.

I’ve got to say one more thing about the idea that stereotypes are mental laziness. That is totally correct. When we use a stereotype, it is our mind operating automatically and giving us something that is sometimes useful and sometimes not. But often don't really bother to ask ourselves whether it's useful or not. We should be aware that our mind operates this way. It is a very normal way of operating and does a lot of good work for us. But we need to be wary that sometimes it will do work that actually gets in the way of what we are trying to do.

JR: You know there was an interesting idea in chapter 5 of your book about stereotypes that I have never really run into before. It’s the paradoxical idea that applying stereotypes could actually bring you to the point where you are able to picture the individuality and uniqueness of a person, which is the exact opposite of stereotyping. Can you explain that?

AG: Yes it is bit of a tough idea, and one that is doesn't really exist yet in social psychology. In that chapter we explored how we can combine categories like race, religion, age, etc. to come up with very unique creations, because these combinations form pictures in our minds. For example, in that chapter we suggested picturing in your mind a Black, Muslim, sixtyish, French, lesbian professor. Now, most of have never met anyone with all those characteristics, but we can string together labels like types of occupation, sexual orientations, etc, and combine them to construct a category of person that makes sense to us. We have no difficulty creating a pretty good mental picture of that kind of person, even though we may have never known a person like that in your entire life.

JR: Your book is based on a lot of research. The Implicit project has over 2 million people who have participated.

AG: Actually over 16 million people. We started in 1998 and there are now 14 different versions of it on the website now. Most of them have been running for more than a decade. We know that the Implicit Association Test has been completed over 16 million times The one that has been completed more than any other is the race attitudes test, which measures pleasantness and unpleasantness associated with the racial categories black and white. That test has been completed between 4 and 5 million times.

JR: One enjoyable aspect of Blindspot is interactive activities, visuals, and practical examples that help engage people in these ideas and concepts. Early in the book demonstrate the idea of the blind spot. Can you tell us what that is and how the blind spot helps us understand this whole area of stereotypes and implicit bias?

AG: The blind spot is an old perceptual demonstration that involves looking at a page that has two dots drawn about 5 inches apart on a white page. When you close one eye and focus on one dot and then move the page within 7 inches of your eyes, the other dot disappears. Then, if you switch which eye is open and which is closed, the dot that disappeared becomes visible and the other dot disappears. That’s the blind spot. When you're experiencing this blind spot in the demonstration, the background is continuous, and there’s an illusion of a hole in your vision. That’s because your brain actually fills it in the blind spot with whatever else is in the neighborhood. The blind spot becomes a metaphor for a piece of mental apparatus that is not actually seeing what's going on.

JR: We are hard wired to have a visual blind spot.

AG: Right, but the mental blind spot we are referring to is not merely a single compensatory apparatus. It is actually a whole array of mental operations, which we can't see happening. They are happening out of sight. This is very important stuff. The marvel of the Implicit Association Test is that it actually gives us a way of seeing the portions of the mind in which these things are happening.

JR: The racial IAT findings say that many Americans have preferences for white faces relative to black faces, which is easy to extend to being a preference of white people over black people. But what are we to make of this? To some people the fact that you like different faces on this test would not be a very unimportant piece of data.

AG: You might think “Okay I have this preference according to the IAT, but isn’t that just a different way of measuring what I would say if you just asked me questions about my racial preferences?” But that's wrong. The biases revealed by the IAT, wouldn’t come out if I were simply answering questions. If you asked me questions about my racial biases, I would deny that I have I any kind of racial preference. And not because I’m lying, but because I’m not aware of the automatic associations that the IAT reveals. This pattern actually applies the majority of Americans and people in other countries also.

JR: There’s an example in your book of someone wrote to you and said there's no way that they really liked Martha Stewart more than Oprah Winfrey, even though your tests said they did.

AG: Yeah. This happens all the time. There is a very understandable source of resistance to believing that what the IAT is measuring has any validity. We can understand this theoretically in terms of the two levels we discussed earlier. The IAT measures something that is going on automatically at the lower level, outside of our awareness. Survey questions, however, where you answer with words or check marks reflect conscious thoughts that are happening at the higher level. We now understand that these two levels of mind do not necessarily have to agree with each other. Then it becomes a question of how to deal with this discrepancy.

One of the common questions we often get is whether or not the unconscious attitudes measured by the IAT have a significant effect on our behavior. The answer is yes. The automatic associations we make at this lower, unconscious level will generate conscious thoughts that reflect those associations, even though we don't even know that we have them. This can then alter the judgments that we make consciously.

My wife told me about a radio story she heard about a black attorney named Bryan Stevenson who works for the Equal Justice Initiative. He was in the courtroom with a client, who happened to be white, sitting at the defense desk before the trial started. The judge walked in and came up to Mr. Stevenson and said “Hey, what are you doing sitting at the defense table? You shouldn't be here until to your lawyer is here.”

JR: That's amazing!

AG: Yes. Bryan Stevenson laughed it off. The judge laughed it off. But it was a very serious thing, reflecting the automatic operations in the judge's head which told him that a black person sitting at the defense table, even one wearing a suit, is not the lawyer but the defendant.

JR: Wow. In one of the Appendixes in Blindspot, you describe a significant change over the decades in how people answered straightforward questions about race. The kind of blatantly negative views of black people are no longer popularly endorsed, as they were before the Civil Rights era. Isn’t the IAT is telling us that these more blatant expressions of racism may have changed without a corresponding change in implicit negative associations many people may continue to hold toward black people?

AG: Yes Mahzarin and I have been very careful to say that what the IAT measures does not deserve to be called racism. The IAT is measuring automatic preferences for whites relative to blacks. This is a preference one can have if one likes both whites and black, if one dislikes both whites and blacks, or indeed if one likes whites and does not like blacks. But this is not racism. It is a mental association that happens automatically. It is related to discriminatory behavior, but is not necessarily hostile discriminatory behavior. This is something that occurs much more subtly.

JR: One of the interesting findings you describe in your book is that many African Americans also have an unconscious preference for whites.

AG: That is true. Among African-Americans in the United States there is close to an even split between those who have a preference for white faces relative to black and those who have preference for black relative white. Yet if those same people are asked if they feel warmer toward whites versus blacks, the African-Americans will very strongly make it clear that they feel more warmly to black people than white people. Interestingly it seems that many African Americans are not governed by political correctness like whites, many of whom think that if they feel more warmly toward one race than another that they shouldn't express this feeling. But not among black people. African Americans do show different patterns on the race IAT than whites, but it's not exactly the opposite. They are very balanced and on average show very little net preference one way or the other. But what is similar is the distinction between what their words say about preference and what the IAT says about their preferences. What they honestly believe about themselves often differs from their implicit preferences, as is often the case with whites.

JR: I'm wondering if your book has sparked public controversy.

AG: That's interesting. Our scientific work has been controversial in that there are people who are very much against the idea of the using reaction time as a way of measuring of the kind of attitudes that were in the past measured by survey questions that had verbal responses or used checkmarks. We experience much more controversy from within our field than we do within the general public, including readers of Blindspot. There has been almost no strong opposition to the conclusions of the book, and many people are finding that these ideas lead them to understand that it is necessary to do something to prevent the operation of the unconscious biases. But we have some scientific colleagues who want to fight about all this.

JR: The science in Blindspot suggests how resistant many of these implicit biases are to changing. But the fact that Barack Obama was elected twice to the presidency seems to reflect some major changes. Some people are even saying that the age of race is over and that we are in a post-racial era.

AG: I share the view that I know a number of political scientists hold, which is that Barack Obama managed to get elected president despite the fact that he was black. This had, in part, to do with other things going on in the country. Republicans were starting to lose political support due to issues like immigration and the financial catastrophe of 2008. These forces just managed to overcome the loss of votes that Obama experienced due to the fact that he is black. I've actually done research on this topic that has been published in scientific journals.

JR: In black society we sometimes talk about something called the black tax. That’s the additional amount black people pay for things because they earn less money, they aren’t offered fair deals, or the obstacles to success are harder for them. So what was Barack Obama’s black tax? What did being black cost him in terms of election percentage points?

AG: The estimates from the study we did are that there was close to a 5% decrease in votes for Obama because of his race. And others have made similar calculations. There’s no doubt Barack Obama would not have been elected in a presidential election conducted just by white voters. Obama would have lost by a huge landslide, perhaps as much as 60% to 40% in favor of his opponent.

JR: I am wondering what your IAT research can do to help us navigate the many significant race issues that have been in the headlines recently—things like unjustified police shootings of African-Americans? In those cases, the officers almost always say that they felt that their lives were endangered, but most African-Americans—and maybe most people—look at the situation and think how could that be possible?

AG: To answer that question, we need to distinguish between the different kinds of situations in policing. For example, when police find themselves confronted by someone who is possibly carrying a gun, it may not make a difference whether that person is black or white. They may assume that no matter who that person is, if they are reaching for something that could be a gun, the police officer may indeed feel that there's a real threat. That’s a very important type of situation, but not one that I've studied. Nor am I prepared say exactly how the IAT applies to it.

The kinds of policing situations that I study are much more common, such as profiling. Say a police officer is following a car and decides to stop it because a taillight is not functioning. It is well known from stop and frisk studies that it makes a difference whether the driver is white or black. That’s the kind of thing that can results from automatic processes that the police officer may not necessarily be aware of. I'm not saying that there are no police officers who engage in deliberate profiling of blacks for stops. I think that does happen. But I think the more significant problem is the implicit profiling that operates more automatically. If the police officer has more suspicion that something illegal is going on if the driver is black, then it seems to me that there could be an implicit, automatic.

JR: I was surprised to find out from your book that some of the best documented bias is found in medical practice, where African-Americans are more frequently given the less preferred medical interventions. And the people showing this bias in medical care are among the best trained people in the country.

AG: It is very difficult to suspect that doctors are producing health care disparities, which often show up in the unequal treatment of whites and blacks. It’s very difficult to treat this as something that is covered by conscious intent to provide less satisfactory treatment to black patients. So it becomes plausible that something is operating on a more automatic level of basic stereotypes that the doctors may not be aware of. Many medical professionals are interested in this. In training sessions related to medical disparities they often have a hard time getting their minds around the idea that there could be something their minds that is causing them to provide less care than they would want to provide. It’s something that will someday be solved by training, but not the kind of training that is easy to do. Psychologists need to provide more continuing education on the implicit revolution in order to get people to understand the extent to which their minds can operate automatically.

JR: This Implicit Revolution is a major paradigm shift for us. Most of us have gotten over the idea that the earth is round and that it goes around the sun. But this is a big one for people who have a strong sense of personal independence and like to think that they are the master of their fate.

As we wrap things up, I wonder what you would consider to be the important take home message that you would like people to get from Blindspot?

AG: It is sort of a know thyself message. In this book, we were trying to show what psychology has learned recently about how our minds function and what we can do to better align our behavior with our conscious beliefs, as opposed to our unconscious biases. Part of the secret to doing that is to simply do things that cause your mind to do more than merely operate automatically. You can do this by monitoring closely what you're doing.

JR: You offer a challenge in the title of your book by saying that these are the hidden biases of good people. These are people with good intentions who see themselves as good, but some of your research might challenge that assumption.

AG: You have to realize that part of the reason for that subtitle is that the two authors of the book regard themselves as good people and they have these biases. And we believe that we’re not alone in thinking that we are good people and we’re not alone in not wanting to be governed by these biases. There are so many such people that if they all and bought the book I'd be very wealthy indeed.

JR: One thing I often remark on in teaching students or trainees about dealing with offender populations, antisocial personalities, and psychopaths is that good people want to be good and they also want to be seen as good. In contrast, with criminally oriented personalities, you often find that they don't want to be good and they don't be seen as good. So I think that wanting to be good goes a long way toward starting to be good. This process of knowing yourself is something that you should engage in whether you’re involved in the race conversation or not. I highly recommend your book and your research as a starting point to that process of knowing yourself—knowing where you are and where we are here in America.

AG: I want to thank you for making that point. Those of us who do want to see ourselves as good people should be interested in learning how our mind’s automatic operations may get in the way of our intentions. That's a great point to end on.

JR: Thanks, Tony. I really appreciate your generosity with your time and also giving readers a chance to share in the debut of some new breakthrough concepts that you introduced during our interview. I ‘ll certainly be looking for more about the Implicit Revolution. Having these ideas more commonly understood will prepare the way for many positive changes.

AG: Thanks for this conversation appreciate you taking interest in our work.


Click here to listen to the full interview with Anthony Greenwald about his book Blindspot.

About the Author

Jay Richards, Ph.D., is a forensic psychologist, whose specialty is the evaluation and treatment of violent offenders. He is the author of Silhouette of Virtue.