How to explain and how not to explain contemporary criminal violence in central America

Abstract

The level of violence in Central America measured in homicides is ten times higher than the world average. This evaluation does not refer to the formerly all too typical description of Central America as the site of revolution, civil war, or military dictatorship, but rather the situation after the “third wave of democratization” well into times of peace. Consequently, it has to be violence in peace. Although peace replaced Central America’s civil wars as well its political violence, it did not oust violence per se. Criminal violence not only continued, it intensified and produced higher homicide rates than during the civil wars. Criminal violence therefore cannot be a transitional phenomenon. Hence, the question arises: even though the region shares historical-structural similarities, why is the level of violence in terms of homicide in three countries (Honduras, El Salvador, and Guatemala) particularly high, and at the same time in two countries (Costa Rica and Nicaragua) exceedingly low? With only Costa Rica as the counter-example, this difference would not be a puzzle-as Central America’s Switzerland and as a democratic welfare state, Costa Rica resembles Western European developed countries. In particular, the puzzling case is Nicaragua, as it is characterized by low levels of violence but shares many of the same structural features of its northern neighbors El Salvador, Guatemala and Honduras (notably inequality, social exclusion, poverty and a violent past). This contribution examines the causes of criminal violence on the isthmus. These causes will be identified through a stringent causal comparison of all five countries. The independent variables have to be absent in those cases in which the dependent variables do not appear. Regarding the rigor of this comparison, it is compelling that those factors which are to be confirmed as causal for criminal violence are obliged to be present in all three countries characterized by high levels of violence. The same factors must not occur, in contrast, in either of the two countries with low levels of violence. The present contribution provides a conceptual model combining political science with criminological approaches and presents empirical evidence to support the model to finally solve the puzzle of violence in Central America.