Assessing the Counterfactual: The Efficacy of Drug Interdiction Absent Racial Profiling

Abstract

This Article investigates the costs and benefits of racial profiling in the context of drug interdiction. I begin by reviewing the empirical economic and civil rights literature regarding the existence and rationality of racial profiling and then build an explicit model of a trooper's decision to search a stopped vehicle. Estimating the model using stop and search data from a portion of Interstate 95 in Maryland, I find that the Maryland State Police use the motorist's race as a factor in deciding which stopped vehicles to search. This result persists even after controlling for many other descriptive variables that impact the trooper's decision to search. I then introduce an additional model that controls for race's role in the search decision and estimates the counterfactual: the change in the amount of drugs the police would find if they ignored race as a factor in the search decision. Applying that model, I find that race is the strongest predictor of identifying drug traffickers, but that racial profiling comes at significant cost, as Black motorists who are subject to search are also more likely to be innocent.

Keywords: racial profiling, selection model, empirical, police, criminal justice

JEL Classification: C10, C35, K42, J78