Using Project Implicit Data to Understand Racial Disparities in Health

By Jordan Axt

How can implicit and explicit racial attitudes be used to understand important life events? What role may implicit and explicit racial bias play in understanding health outcomes for minorities? Researchers Jordan Leitner, Eric Hehman, Ozlem Ayduk and Rodolfo Mendoza-Denton present one investigation of such issues in a recent paper published in Social Science and Medicine.

The authors used 10 years of Project Implicit data from the Race IAT (publically available at to find average levels of implicit and explicit racial attitudes among Black participants. These county-level averages were then used to predict health outcomes in that area. Specifically, the authors focused on the rate of circulatory-related deaths--the leading cause of death in the United States--among Black and White people.

Results showed that Black residents were more likely to die from circulatory-related illnesses when living in counties where Black people had greater anti-White implicit and explicit attitudes. That is, counties where Black people indicated liking Black people more (and White people less) were more likely to have higher Black death rates. Further analyses showed that implicit, automatic attitudes still predicted Black death rates even after accounting for demographic variables like unemployment, housing density, and income.

These results suggest a number of compelling conclusions. For one, it provides evidence that racial bias (at least implicitly) among minority populations can be used to predict health outcomes, as previous investigations have focused more on the implicit biases and health outcomes of only majority group members. Second, it suggests that implicit racial bias in attitudes works similarly for both majority and minority populations; in both cases, more ingroup implicit bias was associated with worse health outcomes.

Of course, these data are correlational, so it may not be the case that implicit attitudes themselves cause the differing death rates between counties but rather another factor related to implicit attitudes is what causes these health disparities. For example, the authors speculate that areas with higher implicit racial attitude biases may also have more strained race relations, or greater physical distance between Black and White people, and these factors might contribute to the observed disparities in circulatory-related health outcomes.

Finally, the work lends further support to the approach of using Project Implicit data to investigate larger county-level outcomes. These analyses demonstrate that Project Implicit data can be used to measure an area's implicit and explicit racial attitudes, and that such attitudes are related to meaningful life outcomes.