An Examination of the Spatial Network Dynamics of Violence Against Civilians in Civil Conflicts

What explains the proliferation of civilian targeting behavior between militant organizations across space? Targeting civilians in conflict is a high-cost and potentially counterproductive strategy in civil conflict. Despite this it remains a prolific strategy across areas and contexts. Existing research on violence against civilians (VAC) has identified multiple mechanisms within a conflict environment that may produce or increase the spread of VAC. I argue that existing methodological approaches suffer from a lack of precision and fail to properly account for VAC’s existence within a complex system of conflict where multiple mechanisms may be present and simultaneously affect one another. I rely on existing spatial econometrics literature to suggest that these mechanisms should be disaggregated from one another rather than analyzed cumulatively. I build off of the work of Lis, Spagat, and Lee who find VAC to be spatially dependent and suggest emulation or retaliation as possible mechanisms. While both mechanisms may be plausible, I show that by partitioning the spatial analysis using network analysis, spatial proximity only appears to be a relevant factor between actors without a conflict history in the previous year, which suggests that it is more likely that emulation is the primary mechanism driving spatial diffusion rather than retaliation. Additionally, I suggest that this may point to a larger conversation on precision in the study of spatial dynamics of conflict.