Fearon–Kalyvas Model: Toward a Unified Model of Battles and Violence in Civil War

Fearon (1995) and Kalyvas (2006) have arguably made the most influential contributions to conflict studies in recent decades. However, scholars tend to use their models separately—Fearon’s for conflicts between armed groups and Kalyvas’s for violence against civilians—overlooking how they inform each other. We propose a model that unifies these classical theories. The new model, Fearon–Kalyvas model, highlights the crucial role of the relative efficacy of battles and violence. When fighting has better prospects than violence, armed groups attack their adversaries to avoid intermediate territorial control and violence against civilians. Therefore, unlike in Kalyvas (2006), intermediate territorial control does not necessarily result in violence against civilians. Theoretically, this suggests that the “cost of peace”—maintaining control through violence—can result in a bargaining failure. Empirically, our model implies a selection bias: territorial control is endogenous to the efficacy of violence, and this endogeneity can bias naïve regression estimates.