Escalation and delay in long international conflicts

Why do escalations in long international conflicts sometimes hasten the pace
of negotiations? And why is it sometimes the case that the resulting terms of
agreement were deemed unacceptable to one or both sides before the escalation? We analyze these issues in a game-theoretic setting with asymmetric information, in which the delay a party exercises before it makes an acceptable offer is served to signal credibly its true stand, of which the other side is initially uncertain. The implications of escalation are twofold. First, escalation makes both sides more eager to settle than before, as an agreement would end the increased level of hostilities. Thus escalation increases the “peace pie” — the overall gains from agreement, and one side or both may eventually yield to terms they wouldn’t accept at the outset, in order to put an end to the increased level of damages that they suffer. Second, escalation shortens the delay each side needs to exercise in order to credibly portray its genuine stand to the opponent. The combination of these “constructive” effects may tempt one or both sides to escalate. However, it turns out that the larger is the overall increase in violence implied by escalation, the higher are also the chances
that the aggressor will eventually regret its decision to escalate.