Why do rebel groups in multi-party civil wars choose to fight one another? Such conflicts divert crucial resources from the fight against the government. If rebel groups seek accommodations that can only be granted by governments, or if they seek to seize the state itself, allocating resources toward fighting another rebel group appears sub-optimal. What accounts for this behavior?
Although military cooperation among rebel groups in multi-party civil wars could help rebels defeat or extract concessions from an incumbent government, violent conflict among rebel groups is empirically prevalent. Why do rebel groups in multi-party civil wars choose to fight one another? This article models the strategic dilemma facing rebel groups in multi-party civil wars as an alternating-offer bargaining game of incomplete information with an outside option. The game-theoretic model explores the relationship between the status quo distribution of power among rebel groups, the costs of fighting, and the likelihood that one rebel group will opt to unilaterally end bargaining over a set of goods, such as access to supply routes, natural resources, and control over civilian populations. We show that the likelihood of violent conflict between rebel groups is lowest when the status quo distribution of benefits reflects the existing distribution of power.