Is Conflict Contagious? Evidence from a Natural Experiment

The fact that conflicts tend to cluster in space is well documented. It remains unclear, however, whether this clustering is a result of contagion or of unobserved shocks that are correlated across space. We present new evidence for contagion by exploiting a natural experiment that increased the intensity of one conflict but had no direct effect on a second ongoing conflict in the same area. In particular, we analyze a ruling by the Supreme Court of the Philippines, which disallowed a proposed peace treaty with the Moro-Islamic Liberation Front (MILF), a Muslim separatist insurgency, and led to an escalation of conflict with the MILF in provinces with a large Muslim population. Though the ruling had no direct bearing on the conflict with the New People's Army (NPA), a communist guerrilla group, we find that it also led to a substantial increase in conflict with the NPA in the same provinces. We test several mechanisms and conclude that contagion was most likely the result of strategic escalation by the NPA in an attempt to exploit the local weakness of the armed forces.