Case Dismissed: Police Discretion and Racial Differences in Dismissals of Felony Charges

Abstract

Prior research has produced conflicting evidence of racial profiling during traffic stops. We instead analyze rates of case dismissal against felony arrestees by race. At the arrest stage, superficial bias based on "unobservables" should be reduced due to the evidentiary requirements and non-negligible costs of filing charges. Nonetheless, we produce and empirically test a model that predicts higher rates of false arrests (and subsequent dismissals) for blacks for crimes where police are required to make snap judgments. Using data from over 58,000 U.S. felony cases from 1990-1998, our probit analysis finds higher dismissal rates for blacks charged with minor violent crimes and crimes where police have more discretion over whether to make an arrest, which points to more aggressive policing of blacks in these situations. Case dismissal rates are also elevated for both whites and blacks when blacks are under-represented on local police forces.

Keywords: Racial profiling, arrests, criminal case dismissals, judicial selection

JEL Classification: K14, D70, D63