Research on race in crime alerts: What can the data tell us?

The question of whether to include racial information in crime alerts has been a focus of important discussions on our campus. In response to requests that the University of Minnesota remove racial descriptors from crime alerts, administrators have changed the crime alert policy to reduce the use of suspect descriptions. This change is critical because racialized crime alerts can reinforce stereotypes of people of color — particularly African-Americans — as violent and prone to criminality.

As social scientists interested in the problem of racial inequality, we understand that these stereotypes do not just hurt feelings — they pose a deeper threat to the well-being of people of color and to society as a whole. Systematic research confirms these concerns. For example, racialized crime alerts strengthen negative stereotypes about black people.

In a 2008 study by Harvard University researcher Scott Akalis and his colleagues, participants who read a single crime alert about a black suspect viewed blacks as more hostile and dangerous than whites. Other studies find that including racial information in crime alerts may actually make it harder to identify the correct suspect.

In this respect, a 2009 study by Kansas State University researcher Ginger Loggins found that racial labels may confuse those who read crime alerts by drawing their attention away from more specific details about the suspect — leading them to incorrectly identify actual suspects.

These consequences are not merely academic. Perceptions of discrimination and exclusion can seriously interfere with the education of students of color, and the well-being of people of color in the broader community.

Indeed, a well-established body of research links racial discrimination to lower academic performance and serious mental and physical health consequences, such as depression, anxiety, traumatic stress, hypertension and increased substance use.

The University’s new policy moves us away from including racial information by default in crime alerts, and we commend the administrators for their decision.

Nonetheless, we have some concerns. The University will now “only use a suspect description when there is sufficient detail that would help identify a specific individual or group.”

However, it is not clear what will be considered sufficient detail to warrant a suspect description. We urge the University to implement this new policy in a transparent manner, by clarifying the criteria for including or excluding a suspect description.