Social Identity and Criminal Justice Contact

• In Event: Policing

Abstract

We use a census of over seven million defendants from the North Carolina judicial system, each person who was arrested from 2013 through 2019, to understand the impact of social and demographic factors on arrest rates. We build a composite index of disadvantage that incorporates information about a defendant’s age, race, and gender in addition to information about their residential address. We use geocoded residential addresses to map each individual to granular neighborhoods defined by household income levels. By comparing this five-fold index with U.S. census-derived information, we can estimate the odds of arrest for any distinct demographic group across the state.

Using this novel measure, we document how demographic disparities in the judicial system emerge across the state. We go beyond previous empirical works by demonstrating the unique role that different institutional hurdles play in exacerbating demographic disparities in the judicial system. What role do institutional actors play in generating these disparities? How do their collective actions help us understand social control in the criminal legal system? We focus on crimes that may be highly targeted at certain demographics, particularly blacks, through legislative intent (e.g., protesting, gang-related crimes, cocaine v. methamphetamines). First, we compare how differential rates of arrest for different types of crimes contributes to entry-level disparities. By focusing on crimes that are extremely common, we are able to compare different regions of North Carolina and understand baseline rates of contact. Next, we document how prosecutors routinely dismiss certain charges and how this culling effect shapes the distribution of pre-trial defendants. Finally, we describe how this pre-trial defendant population shapes jury-decision making and judicial sentencing.

By making comparisons across different institutional pivot points, we can distinguish between differential levels of criminal activity on the part of members of the public versus differential levels of policing on the part of the authorities. We find strong evidence for the differential policing hypothesis, showing clearly how powerfully criminal justice contact and outcomes related to state action rather than to differential rates of criminal behaviors.

Authors

• Marty Alexander Davidson, University of Michigan

• Kaneesha Johnson, University of North Carolina at Chapel Hill

• Frank R. Baumgartner, University of North Carolina, Chapel Hill