Spoiler alert? : the effects of pro-government militias on post-civil war peace agreements

Abstract

Why do some peace agreements last longer than others? The literature speaks of “spoilers”—parties excluded from the negotiations who turn to violence to undermine the agreement—and identifies the risk that opposition groups excluded from negotiations will become spoilers. But the spoiler does not always fight for the opposition. The government party in these conflicts has erroneously been assumed to be unitary. In fact, pro-government militias—armed, organized groups that support the government but are not part of the regular armed forces—are important actors. This project questions the unitary government assumption that is common in the literature.

I propose to analyze these militias as if they were bureaucracies within the state: either they are delegated power, or they seize autonomy. These two models of bureaucratic behavior illuminate the relationships between militias and their government, and suggest how to manage militias in post-conflict situations.

My project proceeds in two stages. A statistical regression analysis finds that peace agreements fail more often when they are concluded while at least one militia was active. Importantly, militias that are closely tied to their governments, and militias that target noncombatants, are especially detrimental to the likelihood of peace. Case studies illuminate these findings. Two successful peace agreements and one failed agreement illustrate how militias act as spoilers and how negotiators used different approaches to address the spoilers.

These findings advance theoretical and practical knowledge of militias and peace processes. There are of course further questions. When do peace agreements really produce a better state of affairs? When do “spoilers” have legitimate grievances? I express no judgments on these questions. I aim only to shed light on how peace might be achieved, on the assumption that some- times it is worth bringing about.