might explain why one civil war lasts two years and another lasts four, but they cannot explain
why some last twenty or thirty years.
In this paper, I seek to expand our understanding of the duration of civil war by moving
beyond the narrow assumption that civil wars are two-actor phenomena and explicitly examining
the effect of multiple parties on the dynamics of conflict. I argue that more parties involved in a
conflict make conflicts more difficult to resolve through negotiation and therefore of longer
duration. This difficulty arises through three separate, but related factors. First, it is harder for
parties to identify the bargaining range of agreements that all prefer to continued warfare. This
problem arises due to two mechanisms—the introduction of additional actors shrinks the
bargaining range and increases the likelihood that at least one actor will overestimate its
probability of victory. Second, in multi-party conflicts each actor has incentives to hold out to
get the best deal as the last signer. Third, shifting alliances between parties prevent the
emergence of negotiating blocks that could make reaching agreement easier. I test this approach
using new data on parties to all civil wars begun since World War II and show that the greater
the number of actors involved in the fighting, the longer the conflict will be.
This article proceeds in three stages. First, I briefly discuss existing approaches to war
duration and termination to situate the argument in this paper in context. Second, I present the
theoretical argument for the effect of veto players on the duration of civil war, culminating in the
main argument presented in this paper, that more veto players lead to longer civil wars. Third, I
test this approach empirically using a new dataset containing monthly data on all civil wars
begun since 1945.
3
Convention