I examine racial disparities in police use of force using new data from New Jersey.
I find that blacks and Hispanics are more likely to have more severe types of force
used against them conditional on force, that these disparities persist after adjusting
for an exhaustive set of factors and using new methods to limit selection bias, and
that they increase with force severity. I then extend empirical Bayes methods to
estimate department-specific racial differences, finding significant variation across
New Jersey’s hundreds of departments. Finally, I observe that officer diversity
cannot predict these departmental disparities, though income and inequality may.