Computational Methods for Police Oversight and Reform Under Incomplete Data
Analyses of police misconduct rely heavily on self-reported law-enforcement data that suffers from unobserved confounding, mismeasurement, and selection issues. I first show how these challenges have led to distortions in prior policing research, including a high-profile retraction. Next, I demonstrate how nonparametric sharp bounds can help researchers address these challenges without relying on implausible, often implicit assumptions.
I introduce an algorithm for obtaining such bounds for any structured discrete system and essentially any estimand, assumptions, and arbitrarily incomplete dataset(s). It flexibly accommodates the wide variety of oversight tasks and data environments across America’s 18,000 law enforcement agencies and allows analysts to fuse police administrative data with novel sources, including traffic sensors and mobile location data. The algorithm outputs a fail-to-reject region capturing both fundamental lack of identification as well as sampling uncertainty.
Finally, I propose an approach for targeting future data collection under budget constraints, such as costly manual review of body-worn camera footage, to optimally narrow the expected fail-to-reject region.
This lecture will be will also live streamed via Zoom.
Dean Knox is a computational social scientist and an assistant professor at the Wharton School of the University of Pennsylvania. His research focuses on the management and reform of police organizations. It develops causal-inference methods for analyzing incomplete or distorted police administrative records, and it leverages computer-vision and speech-analysis techniques to study human interactions with new data sources such as body-worn camera footage. Dean has advised the DOJ, the ACLU, the NAACP LDF, and local organizations on the use of data analytics in civil rights. His work has appeared in Science, the Proceedings of the National Academy of Sciences, Nature Human Behavior, the Journal of the American Statistical Association, and the American Political Science Review. It has been recognized with Science magazine’s inaugural early-career award for interdisciplinary research and an Andrew Carnegie fellowship for social impact.