3
and consequently may present a unique type of challenge to states. Moreover, many scholars have
suggested that the incentives for continuing the dispute drive the inability of groups and states to
settle self-determination claims. However, I argue that bargaining over autonomy is just more
difficult because of who the actors are and how they interact. When states and self-determination
groups attempt to bargain over autonomy, they both do so in the context of internal divisions that
can mire or stop the bargaining process.
This paper presents a theory to explain this propensity to fail to reach agreement focusing on
how internal divisions make bargaining harder between states and self-determination groups. I
argue that the number of internal factions within states and groups make bargaining harder by
increasing the amount of uncertainty between the two sides, decreasing their ability to remain
cohesive throughout the bargaining process, and decreasing their ability to avoid internal deadlock
or spoiling. Specifically, I predict that agreement is most likely when there are a moderate number
of factions in the state and either a low or very high number of factions within self-determination
groups. Empirically, I show that new autonomy agreements are most likely when states have a
moderate number of factions and self-determination groups have either a low or high number of
internal factions. Moreover, I show that states with ample fungible resources are more likely to
reach an agreement, and that groups with local institutions that can coordinate their demands are
also more likely to succeed in their quest for greater autonomy.
The remainder of the paper is divided into seven sections. The first section describes the
politics of self-determination and presents a basic bargaining model upon which my theory rests.
Section two briefly presents alterative approaches for understanding autonomy bargaining failure.
Section three specifies mechanisms that can prevent agreement throughout the bargaining process
that result from internal division in the states. Following that, I discuss these mechanisms for self-
determination groups. The fifth section provides a set of predictions and testable hypotheses
Convention