Iris Malone

Between 1970 and 2012, 963 armed groups initiated violent militant campaigns, but only 25% of these challenges resulted in a full-blown insurgency. Why do some proto-insurgencies become civil wars, but others do not? The dissertation leverages within-state variation in the conflict propensity of armed groups to advance a new theory about why and under what conditions insurgencies form.

The project argues that asymmetric information about armed group capabilities causes a state to misallocate counterinsurgency resources, risking miscalculation and civil war onset. To test the dissertation's theory, I employ a multi-method approach incorporating fieldwork interviews, formal theory, case studies, and machine learning techniques. The project tests these hypotheses on both an existing dataset recording 47,000 distinct, geocoded militant attacks and an original cross-national dataset on the organizational characteristics of 963 militant groups active between 1970 and 2012.

For scholars, these findings advance scholarly understanding of why civil wars begin and the effect of uncertainty on conflict. For policy-makers, this research will improve risk assessments about the conflict in fragile states. If states better understood what types of armed groups are most likely to launch insurgencies, they could improve their efforts to prevent future conflicts.