Since the end of World War II, U.S. strategic thinking has been dominated by the doctrine of deterrence. At its most simple, deterrence refers to one state’s ability to use threats to convince another that the costs of some action—say, invading one of its neighbors—will outweigh the benefits. Such was the logic behind the Cold War concept of mutual assured destruction: if either the United States or the Soviet Union used nuclear weapons, the other would respond with nuclear strikes of its own, resulting in the total devastation of both. By making the costs of war intolerably high, both sides hoped to keep the peace. 

Yet for Washington, deterrence was never merely about protecting the U.S. homeland. As it built the postwar system of alliances that today forms an essential part of the global order, the United States developed a strategy of “extended deterrence.” According to this strategy, the United States would use its military power, including its nuclear arsenal, to defend its treaty allies—among them Japan, South Korea, and the states of NATO. The point was not only to discourage Soviet adventurism in Asia and Europe but also to reassure U.S. allies. If Germany and Japan (to take just two examples) knew that Washington would guarantee their security, they would not need to take actions—such as building a nuclear bomb—that might destabilize the international system. 

Today, the Soviet threat is gone, but the strategy of extended deterrence remains central to the United States’ global power. Washington is still, on paper at least, committed to using military (and, if necessary, even nuclear) force to protect its allies from aggression by rivals. The stationing of U.S. military forces abroad gives additional credence to this commitment, as any attack on a major ally would likely cause U.S. casualties, all but guaranteeing a U.S. military response. Today, Washington’s two principal geopolitical rivals are China and Russia. China is a rising power that is beginning to challenge the United States’ economic and technological supremacy, and Russia under President Vladimir Putin has grown more and more dedicated to undermining the U.S.-led order. Recognizing the threat posed by Beijing and Moscow, top defense officials in both the Obama and the Trump administrations have emphasized the need for Washington to maintain and even strengthen its traditional deterrence strategies.

The question, however, is whether these strategies can credibly deter the sorts of aggression that the United States is likely to face in the twenty-first century. China and Russia are not Soviet-style superpowers with dreams of world domination; they are revisionist powers that want to challenge and change aspects of the U.S.-led global order. There is little chance that China, for instance, would help North Korea try to invade and conquer South Korea, as it did in the Korean War. It is more likely to engage in smaller tests of U.S. resolve, such as seizing from Japan one of the disputed (and unoccupied) islands in the East China Sea that are known as the Diaoyu in China and the Senkaku in Japan. Although the United States has formally pledged to defend these islands, China might suspect that it is unwilling to risk a great-power war over what are effectively worthless rocks. Yet if Washington cannot credibly promise to retaliate, extended deterrence has already failed—and much greater consequences than the loss of one of the Diaoyu/Senkaku Islands could follow. 

Doubts about U.S. credibility have been heightened since the election of President Donald Trump in 2016. Trump has openly questioned the value of U.S. alliances and disparaged key U.S. allies. At times, he has challenged the logic of extended deterrence directly: in July 2018, he expressed bewilderment that the United States’ obligation to defend Montenegro, a NATO member, could lead to World War III. In addition to emboldening U.S. adversaries, such rhetoric runs the risk of undermining Washington’s ability to reassure its allies. And the more that those allies come to doubt the United States’ willingness to protect them, the more pressured they will be to provide for their own security, potentially leading to nuclear proliferation and an increased risk of preemptive or preventive war, among other consequences.

Trump or no Trump, the United States’ contemporary security problems cannot be solved by traditional military deterrence alone. Washington must commit itself to reassuring its allies that it is both willing and able to fulfill its treaty obligations. But even more important, it must begin to broaden its approach to deterrence in light of the changing nature of the threat posed by rivals such as China and Russia. Above all, U.S. policymakers must develop strategies that combine military elements with economic sanctions and other forms of nonmilitary punishment. Such a strategy would reduce the risk of a disastrous war by convincing adversaries that the United States is willing to follow through on its threats, even in an era in which China and Russia are not only wielding more powerful weapons but also showing an increased willingness to use them. 

Chinese and Japanese ships near the Diaoyu/Senkaku Islands, April 2013
Chinese and Japanese ships near the Diaoyu/Senkaku Islands, April 2013
Kyodo / Reuters


In some respects, the fact that the United States now faces problems reassuring its allies should come as no great surprise. Reassurance is hard to achieve. Indeed, in a world of Westphalian nation-states, it is downright unnatural. Persuading one country to depend on another for its security, and perhaps even its survival, runs counter to intuition, common sense, and most of human history. Although Trump’s rhetoric is often imprudent, it may simply make explicit what many already suspected about the United States’ dependability.

Reassurance is also difficult because promises to protect allies should not be unconditional. U.S. allies should not feel that they can be reckless, safe in the knowledge that Washington will bail them out if they get into trouble. In the 1960s, South Korea developed plans for so-called decapitation strikes to kill the North Korean leadership; the United States sought to ratchet down its ally’s offensive inclinations. And in 1965, Pakistan attacked India in the belief that it was protected by U.S. security guarantees. Some fear that Saudi Arabia could attempt something similar today with Iran. As Trump has pointed out in his criticisms of NATO members’ military spending, unconditional reassurance can encourage free-riding by allies, who may assume that the United States will always pick up the bill for collective defense. 

Both deterrence and reassurance require clarity of messaging about when and how the United States will back up its allies. Given Trump’s inconsistency and penchant for rhetorical brinkmanship, some of the greatest sources of worry today come from Washington. Trump waited until June 2017—nearly sixth months into his presidency—to reaffirm the United States’ commitment to NATO’s mutual-defense pledge. He has questioned whether the United States would come to the aid of an ally that has failed to meet its pledge to spend at least two percent of GDP on defense. According to anonymous aides quoted in The New York Times, Trump has even said privately that he does not see the point of NATO and would like to withdraw from it. (In June, he also criticized the United States’ mutual security treaty with Japan as “unfair.”) Earlier, on the presidential campaign trail, he mused that perhaps Japan and South Korea should have their own nuclear weapons rather than depend on those of the United States. 

There is a fine art to both deterrence and reassurance.

Somehow, to date, Trump’s words seem to have done little permanent damage. A 2018 Pew poll found growing doubts about U.S. reliability among U.S. allies (only ten percent of Germans and nine percent of French people asked expressed confidence that Trump would “do the right thing regarding world affairs.”) Yet the same poll showed that an overwhelming majority of those surveyed still preferred the United States to China as the world’s leading power. The U.S.-led world order does not appear to be crumbling. No ally has opted out of the Nuclear Nonproliferation Treaty or threatened to do so. Indeed, no ally has even embarked on a major military buildup. NATO has modestly improved its military burden sharing since 2014, but most of the United States’ security partners still spend a historically modest one to two percent of GDP on their militaries, much less than during the Cold War. Poland and the Baltic states have increased their defense spending to two percent of GDP, but they have not taken the steps, such as fortifying their borders, that one would expect if they truly feared a Russian invasion. U.S. allies may be nervous, but they do not appear to be panicking or radically changing their own national security strategies. 

Despite his rhetoric, moreover, Trump has staffed his administration with figures who are committed to the United States’ presence abroad. Neither Secretary of State Mike Pompeo nor National Security Adviser John Bolton is known for his dovishness or isolationism. The U.S. defense budget has continued to grow during Trump’s tenure, and the president has requested additional money from Congress to develop advanced weapons. U.S. troop deployments have generally remained static, and in some places, such as on the eastern flank of NATO, they have actually increased. Trump has hosted high-level meetings with the leaders of most of the countries—Japan and South Korea, Poland and the Baltic states—on the frontlines of struggles with China and Russia, assuaging their fears of being abandoned in a crisis. In these respects, counterintuitively, the transition from George W. Bush to Barack Obama to Trump shows more continuity than change. 

These have been good steps. But there is a fine art to both deterrence and reassurance: they require constant attention, as both are ultimately in the eye of the beholder. In addition to avoiding capricious threats to pull out of alliances, the United States should make its military commitments more credible in ways that do not require a major increase in combat forces stationed abroad. Washington, for example, could improve its capabilities in Poland by strengthening its logistical and headquarters assets there (as the Atlantic Council has recently recommended) and agreeing to deploy U.S. troops on a permanent, rather than rotational, basis. The most overdue policy changes, however, lie not in the realm of Department of Defense force planning but in that of statecraft—in integrating economic and military tools to develop a new and more realistic concept of deterrence.

NATO fighter jets on patrol in Lithuania, November 2014
Ints Kalnins / Reuters


The Trump administration’s two principal strategic documents, the 2017 National Security Strategy and the 2018 National Defense Strategy, both stress the United States’ blossoming rivalry with its great-power competitors, China and Russia. The NDS identifies both as “revisionist powers” that “want to shape a world consistent with their authoritarian model.” Most notably, Trump has increased the annual U.S. defense budget by around $100 billion since taking office, including in it generous funding for high-tech weapons modernization, among other priorities. 

But in the effort to strengthen deterrence, it is important to ask how a U.S. war with China or Russia would likely start. Put differently, where and how might deterrence, and specifically extended deterrence, actually fail? 

China and Russia know they are weaker than the United States according to raw military metrics. Both are thus highly unlikely to launch the kind of all-out surprise attack against a U.S. treaty ally that would require American retaliation. It is hard to imagine, for instance, that China would invade the main islands of Japan, where some 50,000 U.S. troops are currently stationed, or that Russia would attempt to annex an entire NATO country, even a small Baltic state. Both Beijing and Moscow know that such open aggression would be met with overwhelming U.S. force. 

Yet it is much easier to imagine Beijing or Moscow carrying out smaller tests of U.S. resolve. Perhaps Russia would, as it did in Ukraine, send so-called little green men—soldiers in green army uniforms without insignia—into a small town in eastern Estonia under the pretext of protecting ethnic Russians there. Putin has declared a right to protect Russian speakers anywhere they live, especially on former Soviet territory, providing him with a tailor-made pretext for such aggression. But what he might truly relish is the chance to nibble at a piece of NATO territory and put the alliance on the horns of a major dilemma. Would Article 5 of the North Atlantic Treaty, which guarantees that alliance members will defend one another in the event of an attack, require a military counterattack from NATO in such a situation? Putin might hope that NATO’s 29 members would tie themselves in knots over how to respond. In the event that NATO members, hoping to avoid a great-power war over a relatively minor incursion, failed to honor their Article 5 promises, it could lead to existential doubts about the core purpose of the alliance. 

Should Washington risk a great-power—and potentially nuclear—conflict in order to preserve its credibility?

Or, as suggested above, China might occupy one or more of the Diaoyu/Senkaku Islands. These worthless islets are claimed by both China and Japan. The United States takes no official position on who should control the islands but acknowledges that Japan now administers them and that its security treaty with Japan should therefore apply to their defense. Such a complex, muddled situation is ripe for deterrence failure. Beijing might try to seize one of the islands in order to signal, without crossing the threshold of serious aggression, to Japan and the United States that it is unhappy with some aspect of the postwar order in the Pacific. Beijing might hope that it could force Japan into negotiations and some type of humiliating compromise or drive a wedge between Tokyo and Washington that would make Japan feel more exposed, and less self-confident, on other issues in the years to come, thereby opening up East Asia and the western Pacific to Chinese domination. 

This type of limited enemy assault would raise difficult questions for U.S. policymakers—what I call “the Senkaku paradox.” Should Washington risk a great-power—and potentially nuclear—conflict in order to preserve its credibility, even over something relatively unimportant? Or should it conclude that the stakes are too small to justify such a risk? In the event of limited enemy aggression against an inherently worthless target, a large-scale U.S. response—as the traditional approach to extended deterrence would dictate—would seem massively disproportionate. On the other hand, a nonresponse would be unacceptable, and inconsistent with American treaty obligations, too. 

The way out of this paradox is through a strategy of asymmetric defense. The United States should not formally renounce the possibility of a full military response to very limited (and quite possibly nonlethal) aggression against its allies. Indeed, Lieutenant General John Wissler, then commander of the U.S. III Marine Expeditionary Force in Japan, was right to insist in 2014 that the United States and Japan could expel the Chinese from the Diaoyu/Senkaku Islands if required. As a practical matter, however, the United States needs other options—for both before and after a crisis begins. 

Most of all, Washington’s deterrence strategy should seek to avoid drawing first blood against another great power if at all possible. The United States should prepare responses to small-scale aggression that emphasize economic warfare, and sanctions in particular. At first, the main role of U.S. military force should be to create a defensive perimeter so that China’s or Russia’s appetite for expansion is not further whetted. In the event that a crisis worsens, Washington and its allies could attempt indirect military measures—for instance, targeting ships in the Persian Gulf carrying oil to China. This sort of response would, at least initially, keep the conflict far from the shores of any great power, providing more time for the belligerents to avoid further escalation. But economic warfare should be the core of the strategy, with military force in support. 

Such an approach would help convince a would-be adversary that it would have more to lose than to gain from the use of force—especially if the United States and its allies had taken proper preparatory measures to ensure that they could tolerate any reprisals. The trick would be to make sure the punishments for noncompliance were commensurate with the initial aggression, while maintaining the potential to escalate if necessary. 

U.S. and Japanese ships during a joint naval exercise south of Japan, November 2014
U.S. and Japanese ships during a joint naval exercise south of Japan, November 2014

For sanctions to be economically sustainable, the United States and its allies need to understand vulnerabilities in their supply chains, financial dealings, and other economic relationships. They should develop strategies to mitigate these vulnerabilities—for example, by bolstering their national defense stockpiles of key minerals and metals, many of which today come primarily from China. They should take steps to avoid becoming overly dependent on China for key manufactured components and goods—Washington could prevent Chinese imports from exceeding a specified percentage of certain critical sectors. European states should also continue improving the infrastructure needed to import liquefied natural gas from the United States and other countries as a backup in case energy imports from Russia are interrupted in a future crisis. 

A sanctions-based strategy would be judicious and proportionate, but it would not be weak. Indeed, if Beijing or Moscow refused to either back down or otherwise resolve the dispute once the United States and its allies had deployed sanctions, Washington could raise the stakes. Recognizing that the aggressor state’s strategic aims had become fundamentally untrustworthy or hostile, Washington could seek to not only punish the perpetrator for its specific action but also limit its future economic growth. Over time, export controls and permanent sanctions could replace temporary punitive measures. This strategy would require support from key U.S. allies to be effective—one more reason why Washington needs to respond to these kinds of crises in a way that seems judicious, patient, and nonescalatory, so as to strengthen its coalition and not scare away key partners.  


In his 2017 book, All Measures Short of War, the political scientist Thomas Wright persuasively argues that global orders do not break down all at once. They are challenged, weakened, and eroded in key regions where the interests of rival powers come into direct competition. The western Pacific and eastern European are precisely the regions where such developments are most likely today. 

But China and Russia will not be so mindless as to attack the heartland of a major U.S. ally; American deterrence has not deteriorated that much, even in the age of Trump. The tough scenarios will be in the so-called gray zones of conflict, where classic war-fighting concepts do only so much good. U.S. war plans, as best as can be deduced from the outside, are still too focused on those classic concepts, and probably too escalatory for a world in which large-scale war between nuclear-armed powers must be an extreme last resort. Only a deterrence strategy that recognizes as much—and develops plans involving all the tools of statecraft instead of just military force—can respond to the modern challenges of great-power competition and keep the U.S.-led system of alliances resolute and reassured.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now