The Counterinsurgency Dilemma: Examining Opportunity and Willingness in Insurgency and Civil War

Abstract

[enter Abstract Body]As an alternative to the structural models that currently dominate the civil war literature, we propose an interactive model of onset based on Most and Starr’s (1989) reformulation of the ecological triad. Using the Global Terrorism Database (GTD), we identify 116 periods of low intensity, oppositional group violence short of civil war which constitutes a risk set of actual escalation opportunities. After examining variation in the state’s behavioral response to these escalation opportunities, a logistic regression model predicts, from 1 to 7 years in advance, which of these environments will produce a civil war. Our results provide several contributions to the civil war onset literature. First, we find that oppositional group violence is a widespread, constant, if not random, feature of state-society relations. Compared to strong states, weak states do not face any greater frequency or levels of oppositional group violence short of civil war. They may in fact experience less. Second, we find that most states facing oppositional group violence maintain respectable human rights conditions. A small minority of states respond with harsh repression and, almost without exception, civil war develops only where the state responds with harsh repressive violence. After controlling for the behavior of the state during actual opportunities for insurgencies to escalate, we find that state capacity, poverty and conditions favorable to irregular warfare have no explanatory power in predicting where civil wars will develop. Despite an abundance of low intensity armed struggles against a highly diverse group of regimes around the world, we find an extremely strong and robust regularity: where repression is low – insurgencies don’t grow.

Keywords: insurgency, civil war, repression, weak state