The sources of social violence in Latin America: An empirical analysis of homicide rates, 1980–2010 Mauricio Rivera Department of Political Studies, Centro de Investigacio ´n y Docencia Econo ´micas (CIDE) Abstract The study of social violence in Latin America has stood at the periphery of cross-national research despite the region being one of the most violent in the contemporary world. This article provides a comprehensive review of theories of crime and presents an empirical analysis of social violence in Latin America from 1980 to 2010. The literature often emphasizes one theoretical approach over others and existing explanations are seen as competitive rather than com- plementary. Yet, the empirical findings of this study support different explanations and illustrate how considering different theoretical approaches helps improve our knowledge on social violence phenomena. The results from dif- ferent estimation methods reveal that youth bulges, female workforce, and post-conflict states are positively associ- ated with social violence, as measured by homicide rates. The results also show that states’ efforts to strengthen judicial system capacity and increase school attendance can promote peace. Moreover, while drug producers and/ or transit countries are not systematically related to social violence, money-laundering countries experience higher homicide rates, suggesting that not all dimensions of drug-markets increase violence. Whereas Latin America as a whole has experienced few episodes of civil wars in the past decades, the findings suggest that several factors affecting the onset of civil wars also influence other forms of non-political violence such as social violence. This echoes earlier calls in the literature on the necessity of bridging conflict and criminology research. Keywords democratic regimes, homicide rates, Latin America, social violence Introduction Social violence has been commonplace during the last decades in Latin America. 1 Estimates from the United Nations suggest that more than one million people died as a consequence of an unlawful purposefully inflicted killing by another person between 2000 and 2010 (UNDP, 2013). Paradoxically, the region is one of the most violent worldwide and yet there is a lack of systema- tic cross-national research on the subject (see Bergman, 2006, for a review). This lack of attention is puzzling for several reasons. Fox & Hoelscher (2012) show that institutionalized democracies experience lower levels of social violence – as measured by homicide rates – than ‘hybrid’ regimes. From a theoretical perspective, it is thus surprising that social violence has been endemic despite most Latin American countries being democratic, suggest- ing that political-institutional arrangements cannot fully account for variation in homicide rates in the region. From a policy perspective, the lack of attention is puzzling not just because of the severity of the violence, but also because of its effects on economic development and dem- ocratic governance. The consequences of social violence extend far beyond the direct deaths and injuries, and it has profound detrimental effects on development. The esti- mated material costs of social violence are 3.6% of the 1 Social violence is defined here as ‘acts of violence between individuals or small groups of individuals’ (Fox & Hoelscher, 2012: 431). Corresponding author: mauricio.rivera@cide.edu Journal of Peace Research 2016, Vol. 53(1) 84–99 ª The Author(s) 2015 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0022343315598823 jpr.sagepub.com at CIDE - Parent on February 17, 2016 jpr.sagepub.com Downloaded from