Recognizing Unconscious Racial Bias

Lecture Examines Stereotypes and Their Impacts

By Sara Frueh | Oct. 15, 2019

We tend to think of the process of seeing as fairly objective — that our eyes are similar to cameras, neutrally taking in light and turning it into pictures. But research has shown that biases buried beneath our awareness can powerfully influence how we see.

Take, for example, an experiment conducted by Stanford University social psychologist and NAS member Jennifer Eberhardt, who delivered the 2019 Henry and Bryna David Lecture last week at the National Academies of Sciences, Engineering, and Medicine. Eberhardt and her colleagues showed undergraduate students pictures of the faces of either black men or white men — photos flashed so quickly that the participants’ eyes couldn’t consciously detect their content — or no photos at all.

The students were then shown photos of objects that gradually transitioned from blurry to in-focus, and were told to push a button at the moment the object became clear enough to identify.

For photos of neutral objects — such as staplers or cameras — all groups of students were able to identify them at the same point in time. But for images of crime-relevant objects such as guns or knives, the result was far different: Students who had been shown black faces identified the images of weapons much sooner than those who had been shown white faces or who had not been shown any photos.

The experiment illustrated not only the power of unconscious biases, but also the depth of the stereotypic association people have between blackness and crime — an association that Eberhardt has studied for nearly 20 years.

“African Americans are so associated with crime that the mere presence of a black face can lead us to see weapons more clearly,” said Eberhardt.

“Without even thinking about it consciously, our eyes betray us. It’s as though racial bias works as a visual tuning device, bringing into focus that which is most consistent with the stereotypes that we have, and blocking out the rest,” she added later.

Bias’s Real-World Consequences

Eberhardt’s lecture offered an overview of research on how biases work beyond the lab, shaping our real-world behavior. “Bias can be triggered and can have a devastating impact even when we’re not aware of it, even when it’s our intention to be fair,” she said. 

One study Eberhardt conducted using data collected by the New York Police Department on stop, question, and frisk practices found that NYPD officers were far more likely to stop and frisk black people based on “furtive” movements — a term without an objective definition — than they were to stop white people, even though the black people stopped proved no more likely than whites to have a weapon. 

“Bias can be triggered and can have a devastating impact even when we’re not aware of it, even when it’s our intention to be fair.”

Another study found that during sentencing in murder cases where a victim was white, black defendants whose facial features were more stereotypically black were more than twice as likely to receive death sentences as were defendants with less stereotypically black features.

To study whether bias might be shaping disparities in treatment earlier in life, during schooling, Eberhardt and her colleagues gave teachers files describing a hypothetical student’s misbehavior, and asked how they would discipline the student. Half the teachers were given a file for a student named Greg — a stereotypically white name — and the other half were given an identical file for a student named Darnell, a stereotypically black name. 

When a single incident of misbehavior was described, there was no difference in how teachers said they would discipline the students. But when a second incident was added, race had a large effect: Teachers wanted to discipline Darnell more harshly, and could more easily imagine suspending him in the future. For Greg, the teachers interpreted the two disruptive incidents as unrelated, while for Darnell, they saw the two incidents as a pattern of misbehavior that needed to be shut down.

Loosening the Grip of Unconscious Bias

What can we do about racial bias? “Although racial bias can touch our lives in so many ways, we’re not doomed to be under its grip,” said Eberhardt. “We’re all vulnerable to bias, but we’re not acting on that bias all the time. Bias is triggered by the situations we find ourselves in, and as researchers we know a lot about what those situational triggers are.”

For example, lack of time is a huge trigger, she said; when we don’t have time to think things through, we’re forced to rely on our automatic associations to make decisions. Using subjective standards rather than objective standards can also bring biases to life. Emotional states such as fear can trigger bias as well, as can the cultural norms of the groups surrounding us.

Knowing the triggers for biases can help point the way toward mitigating them, as evidenced by Eberhardt’s work with the online platform Nextdoor. The site, which was designed to help neighbors communicate with one another, discovered that it had a problem with racial profiling: Residents would look out the window and see a black man in an otherwise white neighborhood, and then go on Nextdoor and report the person as suspicious, even when there was no evidence of criminal behavior.

The company’s co-founder sought Eberhardt’s help, and they developed a way to curb profiling by slowing people down. Now, before Nextdoor users can alert their neighbors about suspicious behavior, they see a pop-up checklist prompting them to identify the behavior that makes a person suspicious, reminding them to provide a detailed description of the person (not just their racial group), and explaining what racial profiling is and that it is prohibited.

In doing so, Eberhardt said, Nextdoor was attempting to modify the common advice “If you see something, say something” to “If you see something suspicious, say something specific.”  Adding the checklist led to a drop in profiling by over 75 percent.

Eberhardt’s presentation was supported by The Henry and Bryna David Endowment, which grants awards to innovative research in the behavioral and social sciences by selecting a leading expert to give a lecture on their research and then publish an article based on their lecture in Issues in Science and Technology, a joint publication of the National Academies of Sciences, Engineering, and Medicine and Arizona State University.

Watch the video

Relevant National Academies reports: