In armed conflicts across the globe, the loci of violence change over time, including across state borders. Nonetheless, both academic and policy analyses are typically still guided by static units of analysis and hence fail to capture spatial change in conflict. What explains change in the territorial scope and location of violent events in a single setting of armed conflict? We argue that shifts in two factors contribute to patterns in spatial change across conflicts: the relative strength of the state actor and whether there is a change in the conflict’s dominant actors. To demonstrate our argument, we build a typology of spatial change in armed conflict based on conceptualising conflict as a fluid multi-actor phenomenon. Using the UCDP Georeferenced Event Dataset, we construct a “conflict shape” in the form of a yearly changing polygon as a dynamic spatial unit of analysis. We apply the typology to five cases to show the validity of these factors: the conflict in the Afghanistan-Pakistan borderlands, the conflict in the Horn of Africa, the Islamist insurgency in Nigeria, the conflict in Colombia, and the conflict in Iraq/Syria. The analysis demonstrates the typology’s utility for analysing conflict geographies, conducting within-case and cross-case comparisons of conflicts, and explaining common patterns across various conflicts.

Discussant: Maria Gargiulo