Civil War: Spatiality, Contagion, and Diffusion

The spatial diffusion of violent conflict has been a major concern in recent literature on war. Specifically, students of international conflict have come to devote increasing attention to civil war and how various other forms of internal conflict disperse through a region. In this paper, we assess the effect of spatiality and distance on the proliferation of civil conflict and the prospects for diffusion of civil war either through clustering or contagion from state to state. We employ methods of spatial analysis, including join count statistics, to evaluate spatial dependence in the occurrence of civil war in the postCold War period, from 1991-2004. In addition to discerning the existence of dispersion in the incidence of civil war, our analysis moves toward allowing us to distinguish between clustering and contagion (and the effects of different measures of distance). Our findings provide evidence that domestic conflict does indeed spread through spatial dispersion. Paper presented at the conference, “The Spatial and Network Analysis of Conflict,” September 25-27, 2008. An earlier version was presented at the 2007 Annual meeting of the American Political Science Association, Chicago. INTRODUCTION: THE DISEASE AND ITS CONSEQUENCES Over 35 years ago Norman Z. Alcock wrote The War Disease (1972). He developed a model that was based on the idea that because at least one war (interstate war, which was his concern) was taking place somewhere in the system, the various mechanisms based on the security dilemma and its attendant uncertainty and fear, would serve to generate more war among what we would now call “relevant dyads.” In this way, the “war disease” would continue to spread within the social system of nationstates. One major theme that Alcock presented was the idea that war generated more war. He was, thus, concerned with the consequences of war. As one of the present authors noted that same year (Starr 1972), the study of international conflict was then (as it is now) heavily weighted toward the study of the causes of conflict, while much less attention was devoted to the consequences of conflict. However, the two areas of study are intimately connected in that one of the major consequences of conflict may be more conflict. This has been well demonstrated in the study of war (for two overviews see Most, et al. 1989, and Starr and Siverson 1998), looking at several forms of diffusion. More recently, students of civil war (e.g. Murdoch and Sandler 2004; Sambanis 2002; Gleditsch et al. 2008) have also been concerned with the consequences of civil war; (although, see also Starr 1994). Again, one of the most importance consequences of civil