Bridging Concepts and Applications in the Spatial Structure of Crime

Abstract

This study combines four core concepts—spatial concentration, spatial dependence, spatial clustering, and sigma/beta convergence—into a cohesive framework for analyzing crime patterns and processes. Using the case of Mexico City census blocks, results indicate that extreme victimization has remained confined to a small fraction of blocks, while lower-intensity crime has expanded more broadly. Divergent sigma convergence outcomes highlight the sensitivity of findings to raw versus logarithmically transformed metrics. Negative beta coefficients consistently reveal a catch-up dynamic across blocks, albeit conditioned by time-varying factors captured by year dummies. By uniting these concepts, the approach equips criminologists with a robust analytical toolkit, offering more precise, evidence-based insights for crime prevention and policy.