A Computational Approach to Reducing Inequality in the Criminal Legal System

Cosponsored by the School of Information and the Goldman School of Public Policy

In this talk, I will show how the responsible use of data science and information technology creates new avenues to improve outcomes in the criminal legal system. In particular, I will focus on two such interventions I designed: first, an algorithm that removes race-related information from crime reports, allowing prosecutors to make a race-blind decision about whether to charge someone with a crime; and second, a series of behavioral nudges designed to help low-income public defender clients avoid incarceration imposed for missed court dates. Each intervention has been validated in real-world pilot projects. I will explain how each intervention works, discuss findings from each pilot, and touch on my larger body of research that investigates how technology can improve policy outcomes for marginalized communities.