Sustaining the Peace: Determinants of Civil War Recurrence

The article is an empirical research about why some civil wars recur while some others do not. The authors attempt to investigate determinants of civil war recurrence in the post-conflict environment. They argue that dual sovereignty and agency-related factors can revive an ended conflict. They also argue that how the initial conflict ended affects probability of civil war recurrence in the future.
First, they argue that to the extent that dual sovereignty persists in the post-conflict environment, civil war recurrence is possible. By dual sovereignty they refer to a condition “when an opposition group has the organizational capacity and popolar support to initiate and sustain an armed challenge to the incumbent regime’s claim to sovereign authority in the nation” (p.173). When the authors compare government victory versus rebel victory, they propose that dual sovereignty is more likely to persist after government victory. They argue that rebels can avoid annihilation by blending into the civilian population and wait until another ripe moment to rebel arrives. Rebels are likely to resume rebellion when after they rebuild organizational capacity, gain resources to sustain a war and mobilize support. Therefore, the authors argue, “any battlefield defeat short of annihilation is not likely to preempt the rebels’ capacity to revive their armed challenge at some time in the future”. On the other hand, the authors believe that a rebel victory is not likely to be followed by a civil war recurrence because the ruling elites cannot blend into the population and prefer exile. Their model suggests that “persistence of dual sovereignty into the post-conflict environment is the primary structural precondition” for the civil war recurrence. 
On the side of agent, the authors argue that “for a resumption of civil war to be preferred, the expected utility of resuming the war, must be greater than the expected utility of sustaining the peace” (p.175). Their model suggests that any factor that “(a) decreases the probability of victory, (b) decreases the payoffs from victory, (c) increases the rate at which the costs of conflict are absorbed, (d) increases the duration of the war, (e) increases the payoffs from maintaining the status quo should increase the actors’ incentive to sustain the peace rather than resume conflict” (p.175).
Here are their main findings:
a) How the previous war ended affects the probability of renewed civil war. Settlements are not as stable as military victories by either side. Moreover, they find that “the probability of civil war recurrence decreases by .27 if it ends in a rebel victory, compared to a government victory, and by .54 if it ends in a peace agreement supported by peacekeepers as opposed to a government victory”.
b) Rebel victories end dual sovereignty more decisively than government victories. 
c) Settlements supported by peacekeeping operations are more stable than those not supported with peacekeeping and the former is even more stable than government victory. 
d) They do not find any support to the contention that conflict type (whether it is an ethnic war, ideological revolution etc.) affects probability of civil war recurrence. 
e) The longer the initial civil war lasted, the less likely that this civil war recur in the future. 
f) The larger the government’s army, the less likely civil war is to recur. Larger government armies function as a strong deterrent for rebels and discourages them from resuming conflict. 
g) They find support for revenge argument  (Walter 2004) which suggests that “the greater the human costs that the protagonists suffered in the previous war, the less likely they are to be able to live together in a new postwar environment and the more likely they are to resume combat if they sense threat from their former enemy” (p.185).   
In conclusion, the authors provide empirical evidence about what determines civil war recurrence after the initial war ends. In addition to domestic characteristics, outcome of the initial war is a significant determinant. A civil war ending in a rebel victory or in a peace agreement supported with peacekeepers are more preferable.