The brazilian Armed Forces and public security on the border: a spatial regression discontinuity design approach

Abstract

Brazil’s Border Strip plays a strategic role in national defense, public security, and regional development and is the target of various control and patrol policies. This study investigates the impact of the Armed Forces’ actions on public security in municipalities within the Border Strip. To achieve this, we analyzed the variation in municipal homicide rates between 2010 and 2019 and applied the regression discontinuity design (RDD) to estimate the effects of the Armed Forces’ presence by comparing municipalities inside and outside the strip. The results indicate that the presence of the military affects violence; however, these effects are not spatially uniform. In the Northern Arc, there was a significant reduction in homicide rates in municipalities within the Border Strip compared to those outside this region. In contrast, in the Central and Southern Arcs, the estimates were either not statistically significant or, in some years, counterintuitive. Furthermore, we tested alternative classifications for treatment and control groups, including cut-off points that consider the location of the municipal seat and differentiated exposure to the national security apparatus. These tests corroborate the positive impact of homicide reduction exclusively in the Northern Arc. The analysis supports the hypothesis that the presence of the Armed Forces has a positive effect on public safety, though this impact is concentrated in the northern arc municipalities.