The killing fields. A Bayesian analysis of crop eradication and organized crime violence in Mexico

Both qualitative and quantitative approaches have been used to study the violence in Mexico during the "war on drugs'' under Felipe Calderon's government. In reference to the latter, although works now exist that analyze the problem via statistical models, there remain aspects yet to cover. In an effort to fill these gaps, the present article aspires to study the effects of eradicating illicit crops on generating violence. As a hypothesis, we propose that violence by organized crime does not occur randomly but rather in response to a spatial and temporal pattern, which is influenced by the eradication of illicit crops. With the object of testing this hypothesis, we undertook regional modeling and selected the most relevant array of covariates and developing parameters for a hierarchical Bayesian spatiotemporal model, with space and time effects being separable. The model captures different sources of variation through a hierarchy of parameters for considering predictive variables and random effects. The findings show a pattern of violence that correlates with the hectares of eradicated illicit crops, allowing us to see that as the measure of eradicated hectares rises, violent events are increasingly perpetuated by organized crime. (C) 2021 Elsevier B.V. All rights reserved.