The Economics of Crime and Punishment: A Computational Approach
Loading...
Authors
Appert, John
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Agent-Based , Complexity , Crime , Simulation
Alternative Title
Abstract
Emergent characteristics of crime rates and law enforcement are observed empiricallyin cities. For example, spatial clustering of crimes is observed. It is difficult to explain
this clustering with Becker’s crime model using representative agents. We extend
Becker’s model using an agent-based approach to explain this phenomenon. First,
we develop a grid model of a city with agents located in housing. We allow those
agents to decide whether to burgle a house in their spatial location. A government
agent allocates resources to fines or law enforcement. We show how these agents’
interactions lead to endogenous criminals and clustering of high-crime areas. We then
demonstrate that this model also produces results found in empirical literature and
how this method could be used to evaluate policy in real-world cities with relaxed
assumptions about human decision-making.