Identifying and Measuring Excessive and Discriminatory Policing

Alex Chohlas-Wood

Executive Director, Stanford Computational Policy Lab

Marissa Gerchick

Data Scientist, Stanford Computational Policy Lab.

Sharad Goel

Professor of Public Policy, Harvard Kennedy School.

Aziz Z. Huq

Frank and Bernice J. Greenberg Professor of Law, University of Chicago.

Amy Shoemaker

Data Scientist, Stanford Computational Policy Lab.

Ravi Shroff

Assistant Professor of Applied Statistics, New York University.

Keniel Yao

Data Scientist, Stanford Computational Policy Lab.

Download pdf

Abstract

We describe and apply three empirical approaches to identify superfluous police activity, unjustified racially disparate impacts, and limits to regulatory interventions. First, using cost-benefit analysis, we show that traffic and pedestrian stops in Nashville and New York City disproportionately impacted communities of color without achieving their stated public-safety goals. Second, we address a longstanding problem in discrimination research by presenting an empirical approach for identifying “similarly situated” individuals and, in so doing, quantify potentially unjustified disparities in stop policies in New York City and Chicago. Finally, taking a holistic view of police contact in Chicago and Philadelphia, we show that settlement agreements curbed pedestrian stops but that a concomitant rise in traffic stops maintained aggregate racial disparities, illustrating the challenges facing regulatory efforts. These case studies highlight the promise and value of viewing legal principles and policy goals through the lens of modern data analysis—both in police reform and in reform efforts more broadly.

TABLE OF CONTENTS