Maps of mayhem

Abstract

Disaggregated studies of civil violence attempt to predict where violence is most likely to break out within states, but have been limited by a near-exclusive focus on political, economic, and accessibility-based factors in explaining local patterns of violence. These factors are important, but the calculus of military conflict does not focus solely on lootable resources or population distributions. Both states and insurgents try to exert control over geographic territory in order to increase their resource base and political legitimacy. Historic evidence suggests that groups use violence to contest control over strategically important locations that allow them to effectively attack and defend territory. I use GIS and social network analysis to operationalize strategic location based on the network of roads and population settlements that make up a country. I find that during conflicts, locations with high degree and betweenness centrality in the road network – in other words, locations that control access to other areas within the state – are significantly more likely to be fought over, even after controlling for a wide range of variables suggested by previous literature and testing for reporting bias. These findings expand on the previous body of literature studying disaggregated violence and show that the calculus of violence during civil conflict encompasses strategic considerations as well as economic, political, or topographic factors.