Criminality in Brazilian Border Municipalities: An Analysis of Overrepresentation and Its Determinants.

This paper analyzes the effect of a municipality being in the "Brazilian border" on its criminality rates, focusing mainly on twin cities and the different arcs. It also identifies determinants of criminality in those municipalities. Firstly, the Propensity Score Matching Method was used to identify the existence of overrepresentation of municipalities in border areas in criminality due to their location and, next, the Spatial Tobit model was applied aiming to identify the determinants of local criminality. Our results showed that, on average, there is no effect of "being in the border" when all the municipalities in border areas and their different arcs are considered. However, when twin cities were analyzed, such effect exists and intensifies the local criminality. Regarding determinants, a difference was found between the municipalities in the Brazilian border and the country as a whole, in which economic advantages were seen to be relevant in the latter context, while the presence of the State reducing social disadvantages is vital to decrease public safety issues in border areas. [ABSTRACT FROM AUTHOR]