Unemployment and crime in the Northern-border states of Mexico: a spatial-Bayesian autoregressive vector approach

Despite the recent episodes of criminal activity intensification in Mexico, it is striking the scarcity of studies analyzing its relationship with unemployment rate fluctuations. The aim of this research is to evaluate the impact that an external shock of a surprising increase in unemployment rates would have upon crime activity, particularly, robberies in Mexico’s northern Border States. To this end, a spatial Bayesian VAR that follows LeSage and Cashell (2015) is applied. Results suggest a heterogeneous response of robberies across northern Border States. For instance, in some states such as Chihuahua, the number of robberies seems to deploy transitory effects, while Tamaulipas appears to be less sensitive to own and neighbor state unemployment fluctuations.