Using Agent-Based Models to Investigate the Presence of Edge Effects around Crime Generators and Attractors


Crime clusters in space, with several potential explanatory mechanisms for the phenomena proposed within the field of environmental criminology. Of these explanations, the concept of crime generators and crime attractors set out by Brantingham and Brantingham (1995) has garnered considerable research attention. Relatedly, the study of crime concentration has also found that boundary areas between spaces experience more crime events than internal zones, commonly referred to as edge effects. There is, however, a dearth of research exploring the relationship between crime generators, crime attractors and edge effects. In an attempt to explore this gap in knowledge, this chapter utilises agent-based models to test whether these theoretical concepts can be considered in conjunction, with the aim of exploring whether the mechanisms which underpin generators and attractors can also lead to the emergence of edge effects. Whilst results of this study suggests that they cannot, they do identify clear differences in the spatial distribution of crime both inside and outside these spaces. To that end, in our experiments simulated crime generators produced increases in crime around facilities which decay over distance and high levels of clustering within them, while conversely crime remains consistent outside simulated crime attractors with only the facility itself seeing increased offending. We discuss how these findings contribute to theory development, and how they may support empirical studies that seek to better understand the mechanisms that underlie real world concentrations of crime.

Agent based modelling for criminological theory testing and development