Lawrence T. White Ph.D.

Culture Conscious

The Truth About “They All Look Alike to Me”

Own-race bias in eyewitness identifications

Posted Aug 10, 2012

Most of us have heard someone say, “How should I know if I’ve seen the guy before? They all look alike to me.” We’re inclined to think the clueless person is a bigot, but might there be some truth to the idea?

For decades, psychologists have studied various factors that can affect the accuracy of eyewitness identifications. One of the most firmly established findings is that eyewitnesses are usually better at recognizing and identifying members of their own race or ethnic group. When a witness and a suspect belong to different racial groups, the chance of a mistaken identification goes up by about 50%.

This phenomenon is known as own-race bias or the other-race effect—and it’s been demonstrated in dozens of laboratory experiments. It’s also been documented in actual criminal cases. In one analysis of 77 mistaken identifications, 35% of the cases involved Blacks who were misidentified by Whites, whereas only 28% of the cases involved Whites who were misidentified by Whites.

When scientists first learned about the other-race effect, they guessed that racial attitudes were responsible, that prejudiced witnesses were the ones who produced the effect. But that’s not the case. Studies have found that racial attitudes don’t predict performance in cross-race identification tasks; prejudiced and non-prejudiced people are equally likely to fall victim to the other-race effect.

If racial attitudes can’t explain the effect, what can? Cognitive psychologists have pointed to the fact that faces are not all alike; they differ from each other in terms of specific features like width, length, size of nose, and color of eyes. Interestingly, a feature that varies a lot in faces of one race doesn’t always vary a lot in faces of a different race.  Black faces, for example, show more variability in skin tone, but White faces show more variability in hair color. In short, races have different kinds of physiognomic variability.

The implications for eyewitnessing are clear. If a culprit is White, witnesses will be better off noticing and remembering the culprit’s hair color. Noticing skin tone is less helpful because Whites don’t vary much when it comes to skin tone. If the culprit is Black, witnesses will be better off noticing and remembering the culprit’s skin tone. Hair color is less helpful because Blacks don’t vary much when it comes to hair color.

Growing up, we learn which features can help us distinguish members of our own group, so Whites tend to focus on hair color and Blacks tend to focus on skin tone. This strategy works fairly well until we’re put into a situation in which we want to identify the face of a person who is racially different. This is exactly what happens when you witness a crime, the perpetrator’s race is different from yours, and the police ask you to look at a lineup.

Choosing the right person in a lineup isn’t easy, and the task is made even more difficult when everyone in the lineup is racially different from the witness. Because we have little practice distinguishing among faces of people of other races, we fail to notice and remember the features that would be most helpful. The paucity of our interracial interactions makes us less competent witnesses—and sometimes we make mistakes. “They all look alike” because we unwittingly look at the wrong things.


Greene, E., & Heilbrun, K. (2011). Psychology and the Legal System. Belmont, CA: Wadsworth.

Scheck, B., Neufeld, P., & Dwyer, J. (2000). Actual Innocence. New York: Random House.