A hierarchical SLX model application to violent crime in Mexico.

In this paper, we develop a Bayesian Hierarchical Spatial Lag of X (SLX) spatial econometric model to examine the relationship between contextual factors associated with violence levels in Mexico over the period 2005-2010, a period in which violence in the country reached record levels. We provide local spillover estimates of relevant covariates at both levels of the hierarchy, i.e., the municipal and state levels in Mexico. We also provide a methodology to compare relevant models regarding the proper specification of the spatial weight matrix in these models. [ABSTRACT FROM AUTHOR]