We rely on a policy experiment in the New York City Police Department (NYPD) to address the well-known problems of omitted variable bias and infra-marginality in traditional outcomes tests of racial bias in police stops. The NYPD designated specific areas as impact zones and deployed extra officers to these areas and encouraged them to conduct more intensive stop, question, and frisk activity. We find that the NYPD are more likely to frisk black and Hispanic suspects after an area becomes an impact zone compared to other areas of the city.
John MacDonald & Jeffrey A. Fagan,
Using Shifts in Deployment and Operations to Test for Racial Bias in Police Stops,
American Economic Association Papers & Proceedings, Vol. 109, p. 148, 2019; Columbia Public Law Research Paper No 14-619
Available at: https://scholarship.law.columbia.edu/faculty_scholarship/2655