bear those costs.
Second, factions in civil war face a constant threat of being defeated militarily
and therefore not being able to bargain further. In governmental bargaining this situation only
arises when political parties are faced with elections.
Rather than seeing these differences as making it impossible to apply a veto player
framework to civil war negotiations, I use both the similarities and differences between
governmental bargaining and civil war negotiations to develop a theoretical model of multi-party
bargaining in civil war. This model emphasizes the strategies that actors in conflict will use both
on the battlefield and at the negotiating table to try to get the best deal possible based on the
number, preferences, and strength of veto players participating in the conflict. I analyze how
these various strategies affect the duration of these conflicts.
Before turning to the presentation of the theory there is one remaining question that I
need to address. I have argued that, contrary to the predominant assumption in the literature, civil
wars often involve more than two parties. In the next section, I discuss what factors lead some
civil wars to become multi-party.
Why Civil Wars are Multi-Party
Using a veto player framework to analyze civil war negotiations is only useful if there is
variation in the number of veto players across civil wars. Earlier, I discussed several examples of
multi-party internal conflicts. A closer look reveals that a significant number of civil wars
contain more than two parties. The Peace Research Institute of Oslo/Uppsala Armed Conflict
Database (ACD) identifies 90 internal armed conflicts that contain more than two combatants at
the same time.
This number represents more than 30% of the conflicts in the dataset. Some of
these conflicts contain as many as ten different groups fighting at the same time.
9
A clear example of the costliness of governmental inability to implement policy can be seen in MacIntyre’s 2002
discussion of the ability of Asian states to respond to fiscal crises.
10
The Armed Conflict Dataset and codebook can be found at http://www.prio.no/cwp/armedconflict/.
7