Violence by Other Means: Institutional Design for Democratic Peace after Civil Conflict

Abstract

Why do some transitions work and others fail? Over a third of civil wars that began and ended in the post war period ended in democratic transitions sealed with elections. About a third of these failed after at least one election (Jung 2007). I argue that the design of government must be considered in combination with electoral systems; societies where electoral systems and policy-making institutions do not mesh run the risk of having elections ignored or overturned, prolonged political instability, or even the reinitiation of conflicts or violence. To this end, I establish baseline conditions for democratic survival, considering both that relative losers need to benefit more from sticking with the system than rejecting and seeking to overthrow it, and that relative winners must not believe that they could do sufficiently better under alternative rules to lead them to overturn democratic rules (e.g., in a self-coup). I then model the interaction between structure and process and the ability of officeholders to influence policy and the electoral rules that determine their identity or preferences. As a test, I apply the event history model to analyze the survival of peace in civil conflicts using the theoretically identified conditions in the model.