Replication Code for: Measuring Racial Discrimination in Algorithms

This Stata and R code replicates the analysis in Arnold, Dobbie, and Hull (2021). The data for this paper contains confidential information about criminal defendants in New York City and so is restricted-use. Accessing the data can be done by entering a data sharing agreement with the New York State Division of Criminal Justice Services and Office of Court Administration. Inquiries can be sent to:<br>DCJS Research Request TeamOffice of Justice Research and Performance, New York State Division of Criminal Justice Services 80 South Swan St., Albany, NY [e-mail]<br><br>An abstract of the project follows<br><br>There is growing concern that the rise of algorithmic decision-making can lead to discrimination against legally protected groups, but measuring such algorithmic discrimination is often hampered by a fundamental selection challenge. We develop new quasi-experimental tools to overcome this challenge and measure algorithmic discrimination in the setting of pretrial bail decisions. We first show that the selection challenge reduces to the challenge of measuring four moments: the mean latent qualification of white and Black individuals and the race-specific covariance between qualification and the algorithm’s treatment recommendation. We then show how these four moments can be estimated by extrapolating quasi-experimental variation across as-good-as-randomly assigned decision-makers. Estimates from New York City show that a sophisticated machine learning algorithm discriminates against Black defendants, even though defendant race and ethnicity are not included in the training data. The algorithm recommends releasing white defendants before trial at an 8 percentage point (11 percent) higher rate than Black defendants with identical potential for pretrial misconduct, with this unwarranted disparity explaining 77 percent of the observed racial disparity in algorithmic recommendations. We find a similar level of algorithmic discrimination with regression-based recommendations, using a model inspired by a widely used pretrial risk assessment tool.<br>