Neighborhood Context of Arrest Probabilities in Baltimore

• In Event: Enhancing Police Legitimacy and Procedural Justice in Black-American Communities

Abstract

This paper examines how neighborhood racial composition and disadvantage shape police responses when citizens ask for help during a 911 call in Baltimore, Maryland. I term these calls citizen initiated calls to distinguish them from police-initiated contact (such as a stop and search). I expect that the probability of arrest following a 911 call will be lower in more disadvantaged or minority neighborhoods, as police are less likely to take calls for help seriously in areas that they view as “less deserving” of justice. Specifically, I used 2016-2018 Baltimore Police Department Calls for Service and Arrests Data and American Community Survey data (2011-2015) to conduct multi-level modeling where individuals are nested within census block groups. The main model is a multi-level logistic regression that predicts the probability of arrest following a 911 call. Preliminary analyses indicate there is a lower probability of arrest in block groups with higher proportion of Black residents, after controlling for type of call. This finding aligns with benign neglect theory such that police are more likely to neglect victims of crime in neighborhoods with relatively high shares of Black residents than crime victims in majority White neighborhoods.

Author

• Madisen Placzkowski, University of Maryland, College Park