Applying Synthetic Control to Evaluate the Impact of Data Driven Approaches to Crime and Traffic Safety on Violent Crime

• In Event: Analytic Innovations in Violence Reduction

Abstract

Data Driven Approaches to Crime and Traffic Safety (DDACTS) is a problem-oriented policing strategy, using high density traffic enforcement to reduce violent crime and traffic accidents in areas where the two overlap. In this research we evaluate the impact of a local implementation of DDACTS on violent crime using recent advances in synthetic control methods. Based on the application of survey methods, the synthetic control method produced a high quality counterfactual estimate of what violent crime trends in the DDACTS hotspots would have looked like had the program never been implemented. A synthetic control constructed from non-target areas of the study site suggested null effects for the program, potentially due unaddressed threats to validity. Synthetic controls constructed from a different city suggest that the program resulted in an initial increase in aggravated assaults, and a decrease in robberies over the course of the intervention period. Substantive and methodological implications will be discussed.

Authors

• Jason Rydberg, University of Massachusetts Lowell

• Edmund F. McGarrell, Michigan State University

• Alexis Norris, California State University, San Bernardino

• Giovanni Circo, Michigan State University