By Eleno Castro and Randy Kotti

Despite the growing body of qualitative evidence suggesting collusion between gangs and political parties in various parts of the world, little has been done to study quantitatively the extent to which criminal organization may affect political elections in such context. Using police data and voting results in El Salvador, we find that homicides in gang-controlled neighborhoods tend to decrease by 24 percent of the mean during electoral seasons. We also estimate that gang control is associated with a 2.75 percentage point increase in electoral participation. These effects are especially significant in the neighborhoods where political parties have a strong voting base. Consistent with the interviews we conducted, this suggests that parties negotiate with gangs to mobilize electoral participation in the areas where they are more likely to receive electoral support and thus increase their chances of winning. To conduct our analysis, we geolocated the homicides reported daily in the registry of the National Civil Police from 2005 to 2019 crossed with electoral results reported at the voting-center level across El Salvador. We exploit the sudden and exogenous decrease in criminality resulting from the 2012 truce between the government and the two main gangs in El Salvador to identify gang-controlled neighborhoods. We also use penitentiary data from the General Directorate of Prisons for robustness measures. 

Pre-publication, 2022. 51p.