View article

This essay describes a generic discrete-event simulation (DES) model of a prison system. The model tracks individual entities through a prison “career,” beginning with admission and ending with release from custody. Career stage transitions are modeled as discrete events that occur in real time. With each career transition (or state-change) the model pauses to update the appropriate subpopulation databases. Entities carry information tags into and through the model. The tags can be used to create arbitrary subpopulations for forecasting or analysis. We demonstrate the model by forecasting the population of a large corrections system conditioned on two policy interventions.