Sources used in this document:

Conflict is costly, but alas, Morrow states, conflict is not the only way nations measure extracted costs. War's costs are unpredictable, of course, but other nation's behavior patterns are unpredictable during negotiations. There may be some predictability in terms of specific interactions between specific nations, based on past behavior, but no national actions can be predicted with unerring accuracy.

First and foremost, Morrow stresses, it is critical to examine variation across and within cases, rather than trying to predict the outcome of individual cases. Ultimately, despite his theoretical construction, Morrow takes a dim view of theoretical predictability given that offensive-defensive theory does not necessarily predict even the conflicts of the past such as World Wars I and II. However, despite the lack of predictability of international actors and resultant war, even from an offensive point-of-view, one of the costs of initiating a crisis is the chance that it will escalate to war. If the parties involved can anticipate the magnitude of that chance, then coercive diplomacy becomes less attractive as the chance it will result in war rises.

Of course, the full effects of all factors on the probability of a crisis and of it escalating the crisis are never certain. (Morrow 2000) Morrow suggests to mitigate potential unpredictability, nations should examine not pre-historical or pre-existing measures of resolve, in other words, the current actions taken during a crisis, rather than pre-crisis balances of capabilities of actions (Morrow citing Fearon 1994b). "Actions taken in a crisis include the effects of private information as revealed by those actions during the crisis, while pre-crisis acts reflect only the observable factors and judgments about the other side's private information." In other words, as a crisis transpires, more is at least seemed to be revealed about the actors in question. Thus the states involved change their strategies accordingly Morrow concludes his article with an overall critique of the supposed universality of offensive-defensive theory, particularly is failure to account for democratic society's better ability to achieve a sense of peace between one another -- a tendency, obviously, not a universal postulate. Still, he states that such a theory as offensive-defensive theory does not explain wars taken for limited ends, merely assuming that wars are undertaken willingly by at least one of the participants though strategic responses to expectations about the likely course of a war may not be as straightforward as the theory assumes. "Offensive advantages may make war less likely in some situations, while defensive advantages increase the risk of war in others." (Morrow, 2002) Ultimately, Morrow ends his article with an optimistic view of democracy's ability to stay in a state of peace, and a dim view of political scientists in the theoretical discipline of international relations to predict strategy where knowledge of other actor's motivations, munitions, and resolve are in constant flux and subject to imperfect data.