Using data from the Center for Policing Equity’s National Justice Database, I am building hierarchical logistic regression models to decompose the variance of racial disparities in police use of force into location-based components. Specifically, I am interested in understanding whether the variance in racial disparities in the probability of use of severe force is higher across census tracts than across precincts within one city. This understanding will guide future research efforts into the practices that are most likely to curb these disparities in a city. This work is advised by Dr. Amelia Haviland.