Carl Lieberman, a PhD candidate in the Industrial Relations Section, recently published a working paper, Variation in Racial Disparities in Police Use of Force, that examines racial disparities in police use of force, with an emphasis on exploring the heterogeneity missed in overall levels.
    
Using novel data from New Jersey's hundreds of police departments, Lieberman first documents that Black and Hispanic subjects are significantly more likely to have more intense types of force used against them, conditional on any force, than White and Asian/Pacific Islander subjects, even after adjusting for a rich set of incident-level characteristics. These disparities tend to increase with the severity of force considered. He then extends empirical Bayes methods to estimate the distribution of departmental disparities, finding that racial disparities can vary tremendously, with some departments not having any observed racial disparities in use of force. Finally, he observes that these departmental disparities are difficult to predict using observable characteristics, which may indicate that one-size-fits-all interventions may be sub-optimal and that intangibles such as department culture may be important.