An agent-based approach to modelling chronic offenders

Police departments are required to ensure the safety of the general population using limited resources. Chronic offenders are a major drain on police resources. The precise definition of a chronic offender may change between jurisdictions but the general concept refers to a class of offenders who commit many crimes in short intervals. Unlike other offenders who typically stop committing crimes early in life, chronic offenders continue committing crimes late in life. Understanding this class of offenders allows police departments to modify their operational strategies and make better use of their resources. We develop an agent-based model of how chronic offenders move between incarceration and freedom. Under appropriate limits we convert the agent-based model to a system of differential equations which is analyzed using established differential equation methods.