We analyze the first large-scale, randomized experiment to measure presidential approval levels at all outcomes of a canonical international crisis-bargaining model, thereby avoiding problems of strategic selection in evaluating presidential incentives. We find support for several assumptions made in the crisis-bargaining literature, including that a concession from a foreign state leads to higher approval levels than other outcomes, that the magnitudes of audience costs are under presidential control prior to the initiation of hostilities, and that these costs can be made so large that presidents have incentive to fight wars they will not win. Thus, the credibility of democratic threats can be made extremely high. We also find, however, that partisan cues strongly condition presidential incentives. Party elites have incentives to behave according to type in Congress and contrary to type in the Oval Office, and Democratic presidents sometimes have incentives to fight wars they will not win.

W hen states come to believe that other states are hostile to their interests, they often reorient their foreign policies by realigning alliance commitments, building arms, striking first, mobilizing troops, or adopting policies to drain the resources of states that menace them. This article presents a crisis bargaining model that allows threatened states a wider array of responses than the choice to back down or not. Two implications are that (1) "cheap talk" diplomatic statements by adversaries can affect perceptions of intentions, and (2) war can occur because resolved states decline to communicate their intentions, even though they could, and even though doing so would avoid a war. The model relates the content and quality of diplomatic signals to the context of prior beliefs about intentions and strategic options. In simulations, this form of diplomatic communication reduces the likelihood of conflict.

We conduct a survey experiment to examine the effects of international compromise, war, and foreign government rhetoric on presidential approval. We find that, in certain conflicts, popular approval tracks fairness heuristics-leaders seeking to maximize voter approval prefer equitable divisions of disputed goods and are risk acceptant for divisions below this threshold. Moreover, aggressive rhetoric by a foreign leader increases domestic leaders' expected approval from war, decreases the value of compromise, and provides them with powerful incentives to fight harder. Thus, leaders motivated by popular approval have preferences that are inconsistent with the non-satiated, risk-averse preferences defined over shares of an objective good-that is, with those that much of the rationalist literature on conflict assumes. Fairness heuristics and the rhetorical framing of disputes during the conflict process may be at least as important as material factors in understanding why some disputes result in war.

Many scholars and policymakers argue that deterrence strategies have no significant role to play in counterterrorism. The case against deterrence rests on three pillars: terrorists are irrational; they value their political ends far above anything deterring states could hold at risk; and they are impossible tofind. Each pillar is either incorrect or its implications for deterrence have been misunderstood. Under certain conditions, deterrence is preferable to the use of force. Analysis of the structure of terrorist networks and the processes that produce attacks, as well as the multiple objectives of terrorist organizations, suggests that some deterrence strategies are more effective than those of the past. In particular, many terrorist groups and elements of terrorist support networks can likely be deterred from cooperating with the most threatening terrorist groups, such as al-Qaida. Although the use of force against multiple groups creates common interests among them, an appropriate deterrence strategy could fracture global terrorist networks. The current policy of the U.S. and Philippine governments toward the Moro Islamic Liberation Front (MILF) and the Abu Sayyaf Group illustrates the potential of this approach and the risks of using force. Not only can groups such as the MILF be deterred from cooperating with al-Qaida, they may even be coerced into providing local intelligence on operatives linked to it.

States often negotiate with each other over more than one issue at the same time. This article presents a model of multidimensional international crisis bargaining. Unlike unidimensional bargaining, with two issue dimensions states can send costless signals about their resolve that have dramatic effects on other states' beliefs and actions. One reason is that when states claim a willingness to fight over an issue they in fact are not willing to fight over, they may lose the opportunity to get what they really want without conflict. As a result, when there is a chance that adversaries may each be willing to fight over two issues, the states can even sometimes convey with certainty when they will fight for both issues. The model also leads to some surprising comparative statics, for example, decreases in the probability that the target is willing to fight can increase the probability of war.