Violence is a significant development constraint that generates economic problems, limits public and private investments, and damages the country’s infrastructure. This paper offers an explanation of violence through an empirical analysis of Colombian departments that takes into account categories of violence and variables of economic development and the deterrence of violence. We use different datasets to measure violence and economic development, and we employ panel fixed-effects regressions and a dynamic panel model for a sample of 32 Colombian departments between 1993 and 2007. We find that the aggregate-level production per capita, education, deterrence variables and employment rate show a negative effect on violence, whereas the GINI coefficient and lagged rate show a positive effect on violence. Moreover, the objective conditions and their interrelationships have been important in the trends of violence in Colombian departments.
