Stay, Split or Overthrow: Theory and Evidence on Secessionist vs Centrist Conflict

This paper proposes a simple model aimed at predicting under what conditions on fundamentals one should expect that a conflict, conditional on occurring, takes the form of a secessionist conflict or centrist conflict. We show that the relative group size of the involved opposition group, the relative intensity of perceived cultural differences, and the time discount factor, all affect the relative likelihood of secessionist conflict over centrist conflict in a clear manner, leading to testable pre-dictions supported by the data.