Exploring Our Unconscious Biases

Travelers walk down the halls at Miami International Airport in Miami on Friday, May 22, 2015.

For some time now, I’ve been aware of Project Implicit. The university-led collaborative administers web-based tests that purport to reveal whether a person is unknowingly biased about a wide range of issues.

The ongoing study was started in 1998 by three social scientists, Tony Greenwald of the University of Washington, Mahzarin Banaji of Harvard University, and Brian Nosek of the University of Virginia. Over the years, I’ve steadfastly avoided taking the test; truth be told, I didn’t want to know how prejudiced I might be.

Much has been written on the effects of implicit bias and how the often-unconscious attitudes and beliefs that nearly all of us hold foster our comprehension of race, gender, class, ethnicity, and a host of other social constructs. The Kirwan Institute for the Study of Race and Ethnicity at the Ohio State University has undertaken perhaps the best and most comprehensive discussion of this subject. In fact, it was at a Kirwan-organized conference many years ago that I first heard about Project Implicit and grew wary of what its mind-probing test might reveal about me.

But I couldn’t hide forever behind the veil of blissful ignorance. I recently took 1 of the 14 tests on the website. The areas of study include a wide range of subjects intended to measure a person’s bias with respect to sexuality, religion, weight, and, of course, race, among several others.

If I was going to take one of the tests, I resolved to pick the most sensitive one I could find. As a dark-skinned African American man, I chose the “Light Skin – Dark Skin” test to measure my predilections based on the gradations of people’s skin tones.

To be sure, I can think of no topic more divisive and upsetting to black Americans than what Harvard sociologist Jennifer L. Hoschchild and Yale University political scientist Vesla Weaver described as the “skin color paradox” in their 2007 article published in the scholarly journal Social Forces. Hoschchild and Weaver made the case that:

Dark-skinned Blacks in the United States have lower socioeconomic status, more punitive relationships with the criminal justice system, diminished prestige, and less likelihood of holding elective office compared with their lighter counterparts. This phenomenon of “colorism” both occurs within the African American community and is expressed by outsiders, and most Blacks are aware of it. Nevertheless, Blacks’ perceptions of discrimination, belief that their fates are linked, or attachment to their race almost never vary by skin color.

In other words, while black Americans recognize a discrepancy in the way the larger world views them in racial terms, they do not tend to judge themselves or others along a skin-tone grade. More often, Hoschchild and Weaver say, black Americans see all people of color as oppressed, compared with their perception of white Americans as relatively privileged. This is especially true in how black Americans form their political views: “Given that skin color is connected with attitudes and life outcomes in myriad ways, one would expect it also to be associated with political beliefs and identities. We … found a surprise: skin tone seems almost entirely unrelated to the political views of ordinary residents of the United States. We call this the skin color paradox.”

I’m not convinced that Hoschchild’s and Weaver’s views predominate the discussion. Another set of scholars—Stephanie Irby Coard of the New York University School of Medicine, Alfiee M. Breland of Michigan State University, and Patricia Raskin of Columbia University Teachers College—examined skin color attitudes among 113 African American college students to find a marked preference among respondents for “skin color of a medium tone, rather than exhibiting a self-preference for either lighter or darker skin tones.”

Coard, Breeland, and Raskin refused to read too much psychological meaning into their or others’ studies of skin-tone attitudes. “The majority of studies continue to infer, rather than measure, Black identity,” they write. “[A]dditional research is needed on how to more accurately assess skin color.”

Little wonder I was reluctant to wade into such turbulent waters. All my life, I’ve witnessed family and friends wrestle with colorism—the idea that gradations of skin pigmentation are somehow linked to self worth or public esteem. I’ve fought against believing this was true, and it would hurt me deeply to discover that I subconsciously bought into such notions.

With all this angst rattling in my head, I took the test. It began innocently enough, with a series of questions that allowed me to state whether I had any known biases toward skin tones. I answered as honestly as possible but feared my conscious choices tilted toward skin tones similar to my own chocolate-colored skin.

The test then asked me to click on a set of faces paired with words such as “good” and “bad” and concepts such as “joy,” “peace,” “agony,” “terrible,” and “hurt.” Clearly, I thought, the idea would be a measurement of associations.

Not exactly. The test seemed to judge how quickly I made the calculation and measured my response times over the face and value associations. Thus, my subconscious mind would emerge due to the speed of my answers, not whether I answered in a politically correct manner. Very clever, I thought.

The entire test took about 10 minutes. I felt drained afterward, fearful of what I might learn about myself. I expected to show a strong to moderate bias for dark skin. According to the test, however—like some 17 percent of those who had taken it before me—I had “little to no automatic preference between skin tones.”


The study’s overall findings note that most people have a preference for light skin, with 27 percent having a strong automatic preference and 27 percent having a moderate automatic preference. Relatively few respondents have a preference for dark skin; only 2 percent had a strong automatic preference, and 4 percent had a moderate automatic preference.

The test placed me squarely in the middle, in a place of genuine racial ambiguity. I find this ambiguity as uncomfortable as if I’d discovered myself to be a virulent racist. I know better than to insist that I’m unaffected by all the cultural cues that compel nearly everyone to judge others and themselves by race and skin color. Indeed, I’m certain that I have racial biases, even if I’m not actively aware of them.

“Implicit bias confirms that race and gender matter—even among those who consider themselves nonracist and nonsexist,” john a. powell, director of the Haas Institute for a Fair and Inclusive Society and a professor of African American Studies and Ethnic Studies at the University of California, Berkeley, wrote recently in a quarterly journal of the National Committee for Responsive Philanthropy. “Researchers have begun to recognize that most cognitive and emotional responses to our environment happen without our awareness.”

With all these studies in mind, I’m a bit flummoxed by the test. I know enough about implicit bias to have expected mine to be stronger in one direction or another. I am left with more questions than answers. Ultimately, I am unsure why the test failed to tell me what I didn’t really want to know.

Sam Fulwood III is a Senior Fellow at the Center for American Progress and Director of the CAP Leadership Institute. His work with the Center’s Progress 2050 project examines the impact of policies on the nation when there will be no clear racial or ethnic majority by the year 2050.