Simulating calls for service for an urban police department

Police departments in the United States strive to schedule officers so that a number of benchmarks are met. The police administration is often asked to justify to local governing bodies the size of the police force. To assess the effects of force size and scheduling strategies on the ability to meet the benchmark goals, we develop a discrete-event simulation for the calls for service (CFS). Using actual call data from an urban police department in the United States, we fit distributions for call rates and service times for input to the simulation. The output of the model includes statistics related to the response delay, cross-sector calls, and officer utilization. The simulation model verifies intuitive notions about policing and reveals interesting properties in the system.