International relations scholar James Morrow (2000) argues that states write down their alliances because doing so serves to signal commitment. Recall how “defensive alliances” really aim to deter war rather than to fight a war on the defensive. But to deter a war, a state, or a group thereof, must communicate to its adversary which actions are unacceptable, what consequences would follow if those actions are undertaken, and, significantly, how resolved it is to follow through on imposing those consequences, provided that it has the will and capacity to do so. One prerequisite to deterrence is assurance: the notion that promises will be kept and good behavior will not be exploited. Assurance is key to defensive alliances just as much as deterrence, with respect to both allies and adversaries. States need to determine whether an ally will truly come to their aid in the event of an attack. They will look for, and oftentimes ask for, statements or gestures that indicate that the ally has every intention of adhering to its pledges.
Unfortunately, communication in international politics is not straightforward precisely because the anarchic environment creates uncertainty: states can say one thing but do another, without worrying too much about being held to account by a higher authority. There is no one thing that all states can say or do to achieve deterrence or assurance. Otherwise, every state would then be doing that one thing. Statements of interests can thus be muddy.
The solution to this problem is to take actions that are costly enough that not every state would perform them. Costs come in two forms: ex ante and ex post. Ex ante costs refer to sunk costs that states must pay before performing an action; they can include buying armaments and deploying military personnel abroad. Ex post costs reflect measures that may not be felt immediately but nevertheless can constrain future decision-making options (Fearon 1997: 70). According to this rationalist perspective, only in rare cases when interests are completely and obviously aligned – that is to say, identical and harmonious – would states not have to pay costs to demonstrate their support for one another. As may be possibly true of Israel and the United States, “the shared interest carries the entire relationship, and therefore that relationship need not be negotiated formally” (Morrow 2000: 64).
According to Morrow (2000), treaty alliances help to signal interests and, as an ex post cost, to make binding commitments. Indeed, “alliances are institutions” that help to define “the rules of the game” by creating greater predictability in what states are trying to achieve in their foreign relations (Keohane 1988: 184; North 1991: 3). To begin with, states usually have to ratify their international agreements through their domestic legislatures. Treaty ratification can be particularly difficult and time-consuming in multiparty democracies where opposition groups may have reasons for opposing the foreign policy initiatives of the executive. A country’s leadership cannot embark on a treaty alliance capriciously lest it loses domestic support over the issue (Morrow 2000: 72). In order to enhance its international reputation, the state leadership will presumably only endure the pains of treaty-making if it believes that it has enough shared interests with another state to do so. Simply put, an alliance treaty – once signed and ratified – indicates to others that the signatory states have common interests.
An alliance treaty also generates ex post costs by way of tying hands and, therefore, creating commitments. Signing a treaty might only be worth the cost of the paper on which it was written at the outset, but it conveys the state’s intent to come militarily to the aid of another under certain conditions. The public nature of such a treaty creates reputation costs that the signatory state would incur if it decides to renege on promises made to its ally. That signatory state might come to be seen as duplicitous and an unreliable partner for others that confront similar challenges. Faced with greater distrust, the unreliable state would be forced to make more concessions in order to assure potential allies in the future. The ex post costs can also be domestic, especially in a democracy. If domestic legislators ratify a treaty, then they may not wish to see the executive violate the agreement when its obligations are operative, because the national interest, in their view, would be harmed as a result. No such constraints purportedly exist in the absence of a treaty. As such, leaders may feel that they have to fulfill their alliance commitments (Gaubatz 1996; Leeds and Savun 2007; Leeds et al. 2009).
Morrow’s view has become standard in international relations theory, but it has certain conceptual weaknesses. The first weakness touches on whether shepherding an alliance treaty through domestic legislative processes allows the state to signal its interests more clearly to others. States are certainly selective concerning the treaties they sign and ratify, suggesting that, if they have done so, then they were interested enough in striking those agreements. That said, foreign policy interests are not static. If such a signal is sent with the formation of a military alliance, then its strength can easily dissipate over time as new political leaders take power, resources become more or less available, and geopolitical events have the effect of changing citizens’ attitudes and policy priorities (see, e.g., Gartzke and Gleditsch 2004). Even the example of the United States and Israel highlights that interests are not at all harmonious despite what leading theories of alliance formation might imply. President Obama and Israeli Prime Minister Benjamin Netanyahu clashed over issues like Jewish settlements in the West Bank and Iran’s nuclear program (Gilboa 2013: 20–1). Such disagreements should presumably create more pressure for a treaty. Further, whether legislators, and their voters, care about foreign policy, much less military alliances, is an empirical question with no clear answer. In the United States, for example, Congress has historically deferred to the president on foreign policy issues, at least before the Vietnam War. Even afterwards, the White House still “matters more than Congress,” although that is not to say that Congress has no influence on foreign policy, given that it controls the purse strings (Lindsay 1992/3: 608). Finally, there is a question about how reputation costs are generated. Suppose that legislators do care about foreign policy issues and have the capacity to punish the executive when it undertakes actions that are contrary to the national interest. Setting aside the implicit assumption that national interests transcend partisanship and other parochial interests (Trubowitz 1998), legislators and voting publics might still forgive the violation of an alliance treaty if they believe that the risks and costs of compliance outweigh its benefits. After all, fulfilling a treaty commitment could mean a devastating, unwelcome war. As discussed in Chapter 2, reputation concerns are but one of many inputs that figure into a decision to fight, if they matter at all.
So why, then, have written alliances? It may seem that “[a]lliances require specification because the allies need to clarify their degree of shared interests, both to each other and to others outside the alliance” (Morrow 2000: 64). Confirming this view, as discussed in the next chapter, is that states do use treaties to attach escape clauses or specify key conditions about the scope of their commitment, thereby preventing their ally from abusing their military support. However, this view is incomplete. There is not much that is specific in the Washington Treaty, for example. Article V, so often seen as the gold standard of alliance commitments, provides that NATO countries:
agree that an armed attack against one or more of them in Europe or North America shall be considered an attack against them all and consequently they agree that, if such an armed attack occurs, each of them … will assist the Party or Parties so attacked by taking forthwith, individually and in concert with the other Parties, such action as it deems necessary, including