Proactive Policing: Resource Allocation for Crime Prevention with Deterrence Effect.

The article focuses on a novel proactive policing resource allocation model that incorporates the deterrence effect of police presence on crime, using a multinomial logit (MNL) discrete choice framework to capture offenders' location choices. Departing from traditional reactive models like the maximal covering location problem (MCLP), this approach explicitly models how police deployment influences spatial crime patterns, including crime displacement and diffusion of crime control. The authors establish the NP-hardness of the problem, develop tractable mixed-integer linear and exponential cone program reformulations, and extend the model to dynamic settings with both static and mobile police resources. A data-driven case study on surveillance camera allocation in New York City demonstrates the model's effectiveness in reducing overall crime costs compared to existing deployment and MCLP, while addressing endogeneity issues through novel instrumental variables and the Berry, Levinsohn, and Pakes (BLP) estimation method.