Most of the specificity contained in the most important article of the Washington Treaty concerns geography – a major issue of contention, to be sure, in light of how several European countries still had colonies in adjacent regions and even further afield when the treaty was signed (see Coker 1982). Indeed, Article VI further delineates the territorial scope of Article 5 and notes that the attack could be “on the forces, vessels, or aircraft” of members in Europe, on the Mediterranean Sea, or in the North Atlantic area north of the Tropic of Cancer. But there is no automatic obligation to do anything even when an attack occurs. No details are given on what assistance is necessary. Mira Rapp-Hooper (2015: 16) finds that alliance treaties have become much vaguer since the Second World War relative to their pre-war antecedents. Still, alliances require at once specification and ambiguity. Written treaties ironically allow for both.
The reason why anyone would want to have a signed agreement is to avoid a future misunderstanding by managing expectations regarding how the signatories are to behave across a range of possible scenarios. Such expectations can be managed not only by defining the conditions under which the main provisions of an alliance treaty become operative, but also by leaving vague as to what might constitute an attack. Clarity can convey to allies and adversaries an intent to advance certain interests internationally and to highlight a willingness to defend militarily a common set of values. Nevertheless, enough ambiguity can allow states sufficient wiggle room to extricate themselves from an alliance obligation if so desired. Adversaries know well enough that a direct attack of some sort would precipitate an alliance reaction, but the vagueness keeps them in the dark as to which scenarios would produce a particular kind of reaction and which would not. The downside, of course, is that adversaries might still exploit this ambiguity in order to see what they can get away with. Not having a public treaty creates too much ambiguity over interests, but no alliance treaty can be completely unambiguous too. Excessive clarity is impossible because an alliance treaty cannot cover every single contingency, especially when looking further into the future. But even if a high level of clarity is possible, it would be impractical and even useless because states will simply adapt or exploit new opportunities as technology evolves, political priorities are adjusted, and threat perceptions change. With too much precision, adversaries would know how to work around a treaty to their own advantage – for example, whether to attack in ways that stop short of clear red lines. When we explore entrapment risks in the next chapter, we will see that states often design their treaty commitments to mitigate such concerns.
Another reason to have a written agreement is that it indicates who is in and who is out. Put another way, the choice is not only whether to sign a treaty with one state, but whether to sign the same treaty with multiple states. Military alliances can be either bilateral or multilateral, with contrasting benefits and shortcomings for each. John Ikenberry (2005: 146–7) argues that the United States opted for bilateral alliances in East Asia because it saw less need to give up policy autonomy to partners that were much more differentiated in size than in Europe. Bilateral alliances are thus easier to manage and so can provide a strong state with more flexibility and greater control over its weaker counterpart (see also Cha 2016). In contrast, the United States preferred multilateralism in Europe because it had a much more ambitious agenda that went beyond simply deterring the Soviet Union – one that required the partnership of “roughly equal-sized states” in order to consolidate centrist democratic governance. A multilateral arrangement like NATO would allay concerns over domination because it provided those partners with more opportunities to articulate their policy demands and to restrain the United States (Ikenberry 2005: 146–7). More cynically, some argue that racial prejudices have shaped alliance decisions – as Christopher Hemmer and Peter Katzenstein (2002), for example, allege in their explanation for why no NATO equivalent exists in East Asia. In their reading, US decision-makers saw East Asian leaders as culturally alien and lacking the racial fitness necessary for multilateral cooperation. This argument may be taking it too far. In Europe, geography and the ground threat posed by the Red Army encouraged a common front, whereas the maritime environment and the difficulties of projecting power over water lessened such a need. Moreover, countries have a say in whether they prefer bilateralism or multilateralism. The United States was, and has been, in fact keen on connecting its bilateral allies in East Asia (Izumiwaka 2020). Unfortunately for Washington, many of them were too suspicious of Japan, so formal multilateral defense pacts were invariably stillborn (Robb and Gill 2019: 161–3). For their part, Japanese leaders themselves were reluctant to build regional security institutions (Izumiwaka 2020: 26–9) Deciding against multilateralism does not imply a total aversion to fostering wider defense ties, but it can indicate a wish to avoid being exposed in the disputes of others. As such, bilateralism lends clarity to the limits of alignment within a wider bloc of states, although states are increasingly opting for multilateral arrangements over bilateral ones (Kuo 2021).
Regardless of the format, a treaty alliance allows for greater efficiency in security cooperation. Of course, greater efficiency does not necessarily imply actual efficiency. At a minimum, uncertainty over intentions and differences in capabilities will put bounds on how credible their alliance will be. That said, a written alliance gives signatories enough confidence that, if they so choose, they can pursue further military coordination, including the drawing up of war plans. This in turn can foster a degree of institutionalization that enables the alliance to weather variable conditions. Arguably, decision-makers sign alliances in part because they anticipate that threat perceptions will shift. Alliances help “lock-in” cooperation so as to mitigate any adverse consequences those changing threat perceptions might have. Interests between prospective allies are already divergent enough at the time of signature for the treaty to be able to manage those differences. Whether by attaching escape clauses or by injecting some ambiguity in the treaty language, states can discourage undesirable behavior on the part of their ally. By treading the fine line between uncertainty and clarity, states can take their security relationship to the next level. Perhaps that is one reason why arms transfers are not a perfect substitute for treaty alliances: arms transfers complement alliances more often than they substitute for them. At least since 2001, according to one study, “the United States sells over twice as much to allies as to nonallies” (Thrall et al. 2020: 113).
Predicting Future Military Alliances
Some readers may be frustrated by the lack of resolution here. Yet that is the point: military alliances are finicky creatures in international politics. Common explanations of why alliances form tend to overpredict them or to overstate their benefits. Many states face common threats but still refrain from signing an alliance. Unequal alliances may not involve the degree of concession-making often ascribed to them. In either case, the purported goal of the alliance does not seem to require having an actual written treaty. The argument that states need a signed alliance in order to specify their commitments is not sustainable. Alliance treaties can be deeply ambiguous, sometimes on the most important points, and yet that very ambiguity is paradoxically worth conveying on paper. States can leverage equivocal treaty language to disclaim any responsibility if they decide against saving a beleaguered ally, but they can still use this language to keep the adversary off balance or to induce it to back down in a crisis. It can be hard to specify the conditions under which they form because threats are what states decide them to be, and the anarchic condition in global affairs makes clear communication difficult and sometimes even undesirable. The ambivalent character of military alliances can thus be infuriating: the treaties underpinning the alliances allow for anything to happen even when the highest of stakes are at play.
That theories of alliance formation are not, and cannot be, deterministic makes it hard to consider the prospect of new alliances. And indeed, if alliance treaties offer states enough ambiguity to sidestep their responsibilities, then why do not more military alliances exist? More specifically, why have Taiwan and the United States not yet rekindled a treaty alliance? Why have China and Russia so far stopped short of signing an alliance treaty? If alliance treaties are sufficiently vague by design, then why not sign as many of them as possible to hedge one’s bets?
The problem is that states need to consider the balance between their shared interests and the differences they have, in addition