Buy Print

Buy Print

The Strategist

Against the Great Powers: Reflections on Balancing Nuclear and Conventional Power

The toughest and most important challenge for U.S. defense strategy is how to defend vulnerable allies against a Chinese or Russian fait accompli strategy, particularly one backed by nuclear threats. Here's how the United States should think about how to defeat such a strategy, and what it means for America's conventional and nuclear forces.


The fundamental problem facing U.S. national security — and indeed grand — strategy is clear: The United States seeks to extend deterrence to dozens of allies in parts of the world that are increasingly shadowed by Russia and China, each of which fields survivable nuclear arsenals and conventional forces that are more and more formidable in their respective regions. An increasingly powerful China seeks ascendancy in Asia and ultimately beyond, while Russia has recovered some of its military potency and aspires to upend or at least substantially revise the post-Cold War European settlement.1 Both China and Russia have developed strategies and forces designed to enable them to attack or suborn U.S. allies or partners and make such an effort potentially worth the risks and costs. Their aspirations place them at odds — or at least in tension — with U.S. interests in defending its alliance architecture, and their increased capability to pursue these aspirations makes them more dangerous and the possibility of war with them more likely.2

In the face of these challenges, Washington wants to deter and, if necessary, defeat attacks on its allies by Russia or China. The problem is that these alliances are, while of course important, still fundamentally secondary interests for the United States. Yet Washington wisely seeks to defend them from states that have the assured ability to conduct nuclear strikes on the American homeland, which, naturally, represents the profoundest type of peril to the nation’s ultimate primary interest: its survival as a functioning society.

In light of the mutual vulnerability of the United States on the one hand and Russia and China on the other, the disincentives to large-scale use of nuclear weapons are of the gravest and most direct sort. No one could rationally seek general nuclear war, which would be tantamount to suicide. In this context, the influence of nuclear weapons derives from the perception of a willingness to risk their use at scale — in effect, to be more willing to court destruction. Coercive leverage derives from establishing a superior position about which state is more resolute in risking nuclear Armageddon.3

But such a competition is not only about resolve in some pure or abstract sense, disconnected from events or acts. Rather, resolve is not an immutable value, but is shaped and formed by a host of factors, and thus is itself subject to manipulation. A state’s willingness to fight is, in other words, not simply a product of an unchanging judgment of the import of a given stake. It is also formed by assessments about the difficulty of and degree of risk assumed by fighting, the connection of the equity at issue to other interests, the perceived nature of the opponent as well as the scale and ambition of its aims, judgments of justice and legitimacy, and so forth. The more these sorts of interests are — or can be — implicated in a given contest, the more likely a state will be willing to risk, fight, and endure, even if the contest is initially or nominally focused on a relatively peripheral interest. The Union fought more resolutely than it otherwise might have against a South that had attacked Fort Sumter first, and the United States fought much more ferociously against a Japan that had launched a dastardly surprise attack on Pearl Harbor and conducted its ensuing aggression with notorious brutality.

In particular, the more aggressive, brazen, illegitimate, unjust, or inherently menacing one state’s behavior seems, the more likely it is that it will generate the willingness of the other state to assume some additional risk of nuclear Armageddon. Put another way, the more capable a state is of attaining its aims through means that appear less escalatory, the less it will need to rely on its resolve to risk general devastation. Conversely, the less capable a state is of pursuing its aims through less escalatory measures, the more it will need to rely on its willingness to court mutual suicide. Thus, the ability to fight successfully without having to seriously escalate is a great source of advantage because it permits one to prevail even with a deficit of resolve.

This is crucial for the United States. In a pure contest of resolve against Russia over Eastern Europe or against China over Taiwan — or even against Pyongyang over the Korean Peninsula — it is not clear that the United States would prevail. But Washington need not and should not permit such a pure contest to appear plausible. Indeed, for many years after the end of the Cold War, Washington enjoyed a situation in which resolve was largely immaterial to plausible contingencies touching on threats to U.S. allies. While Russia had survivable nuclear forces and China a modest strategic deterrent, neither had the conventional forces to mount serious assaults on U.S. allies that would enable them to push the onus of escalation onto the United States, and thus to create a more favorable contest of resolve.4 China might have reminded the United States that it could destroy Los Angeles in the pursuit of subordinating Taipei, but such a threat was not coercively useful without the conventional forces to sustain a blockade or an invasion of Taiwan. Few imagined that China would leap immediately to destroying Los Angeles when such an act, by its manifest disproportion and unreasonableness, would very likely have triggered the most fearsome sort of retaliation.


The post-Cold War period, however, is over. The increased conventional military power of Russia and China, and China’s maturing nuclear deterrent, have changed the situation.

Each is pursuing a variant of what is fundamentally a fait accompli strategy. In a situation of mutual vulnerability to large-scale nuclear attack, the fait accompli is the most attractive offensive strategy for a power that is weaker than its opponent, as China and Russia are relative to the United States and its allies. The fait accompli strategy works by moving or attacking in a way that forces the defender’s counterpunch to have to be so costly and risky as to seem not worth the benefit of reversing it. It is most insidious when the violence needed to succeed with the fait accompli is less grievous, making the very great response needed to eject the attacker seem not only too perilous but also unjust. As a consequence, in a nuclear world, advantage in the deadly competition in risk-taking between two states armed with survivable arsenals will thus accrue to the side that can take action and hold territory — and then push the onus of responding onto the other side in such a way that the sort of escalation required to remedy the situation is simply too costly and risky.

The increased conventional military power of Russia and China, and China’s maturing nuclear deterrent, have changed the situation.

In Europe, Russia’s conventional forces can now rapidly seize territory in places such as the Baltic states and eastern Poland, while Moscow’s large and variegated strategic and nuclear forces provide ample options for controlled strikes designed to “spook” NATO into terminating a war before the alliance could bring its greater strength to bear to reverse Russian gains.5 In Asia, meanwhile, China is developing a conventional military that will be able to compete for — and could be able to establish — superiority over the United States and its allies in substantial areas of the Western Pacific, as well as a nuclear force that could increasingly be used in more limited, controlled ways to attempt to deter U.S. vertical or horizontal escalation.6

The two cases are similar but differ in the greater Russian degree of reliance on nuclear weapons. The nub of the challenge from Russia lies in Moscow’s potential ability to transform its temporary and local conventional advantages with respect to the Baltic states and eastern Poland into permanent gains through the threat of a nuclear escalation that both sides fear but that, Moscow may reckon, the West would fear more. The challenge from China, meanwhile, appears likely to lie more in its potential ability to attain practical conventional superiority over the United States with respect to East Asia and to use its nuclear and strategic forces to dissuade Washington from meaningfully escalating, including to the nuclear level, in order to negate or reverse that superiority. Both, however, involve ways in which a potential U.S. adversary could use its military forces to create durable positions of advantage. Moscow or Beijing might plausibly calculate that such uses of military force would be exceptionally difficult and demanding to roll back or dislodge. This would shift the onus of escalation onto the United States and its allies and would allow for the use of nuclear and other strategic forces to deter the United States from taking the potentially escalatory and dramatic actions needed to achieve its more limited objectives, such as the restoration of an ally’s territorial integrity.


If U.S. grand strategy is to remain predicated on the defense of its allies, the United States needs to deal promptly and resolutely with this thorny set of problems. While Chinese and Russian provocations against U.S. interests have mostly been confined to the “gray zone” thus far, a perception that strategies such as these could advantageously be pursued may lead to more direct and clearer challenges, especially if the relevant regional military balances shift away from Washington and its allies.7 Sub-conventional “salami-slicing” is an attractive strategy when one fears the consequences of pushing much harder or further. If Beijing or Moscow judges it can push more ambitiously or assertively without risking a plausible and sufficiently painful U.S. response, then it is likely to do so. If the United States is resolute and clear enough, however, gray-zone problems will remain manageable — that is, if the United States retains a military advantage with respect to its allies and established partners (such as Taiwan) vis-à-vis Russia and China. If it loses that advantage, gray-zone provocations are likely to transform into far more direct and menacing assertions of power by Moscow and Beijing.

To prevent this, the United States should want a defense posture that demonstrates to potential opponents that such challenges would not succeed or, failing that, would be too costly to be worth the candle. Ideally, this would entail a U.S. ability directly to defeat outright any aggression against its allies or interests, as essentially was the case during the unipolar period that followed the collapse of the Soviet Union. This standard may be difficult to achieve, however, given the dramatic growth of Chinese power and the more modest, but still significant, recovery of Russian military power, the proximity of plausible points of conflict to them and their distance from the United States, and the diffusion of U.S. military effort and focus across multiple theaters. It may be especially difficult to do so rapidly or without requiring significant forms of escalation that may seem a bridge too far for U.S. decision makers absent evidence of a much higher degree of Chinese or Russian malignity or ambition.

The United States should therefore aim to field a military posture of conventional forces that makes attack against U.S. allies and territory at best futile and at minimum a necessarily very brazen, destructive, and aggressive act.8 Compelling the adversary to conduct aggression in this way is far more likely to catalyze U.S. and allied resolve to pursue the kinds of military actions necessary to defeat such an assault, for instance through a much larger counteroffensive, including conventional strikes into an adversary’s territory. This conventional posture should be designed not only to achieve the important but limited aim of repelling an adversary’s attack and denying the fait accompli but also to shift onto the opponent the onus of more dramatic forms of escalation — above all to the nuclear level. That is, U.S. conventional military operations need to be ferocious enough to degrade a capable opponent’s ability to pursue and consummate its attack on U.S. allies, but they should also be framed and implemented in such a way as to compel the other side to have to face the choice of conceding or dramatically escalating. If China can take over Taiwan quickly, cleanly, and with relatively little damage, this is likely to make a large and ferocious U.S. counteroffensive seem disproportionate, thereby lessening the probability that it would happen and that other states would support it. Conversely, if Beijing can hope to conquer Taiwan only through a massive, bloody, and patently aggressive offensive — and might well fail at that — then the sorts of U.S. actions needed to help Taiwan are likely to seem much more reasonable and palatable, thereby increasing the likelihood that Washington would take such actions and that others would support those efforts.


This is primarily a challenge for U.S. and allied conventional forces. But U.S. (and allied) nuclear forces also play a central role. The U.S. nuclear arsenal should be designed to demonstrate to potential U.S. opponents — most importantly Russia and China — that dramatic forms of escalation against key U.S. interests, including but not exclusively nuclear escalation, would be too costly and risky to pursue and ultimately would be self-defeating. This involves demonstrating to potential foes that attempts to transgress core American interests, or to use nuclear weapons for military effect, or seeking to favorably manipulate the fear of escalation to Armageddon would not redound in their favor — and would ideally work against them. This should contribute both to deterring them from using nuclear weapons as a way to reverse a limited conventional defeat but also from crossing fundamental American and allied red lines short of employing nuclear weapons. This, of course, rules out the adoption by the United States of a “no first use” pledge, which would be especially inadvisable given the growth of Chinese conventional military power.9

At the same time, however, U.S. nuclear strategy should seek to avoid unnecessarily or inadvertently triggering a large-scale nuclear war. That is, U.S. nuclear forces should both exercise significant and ideally decisive yet targeted coercive influence but avoid prompting escalation to broader strategic war. U.S. nuclear forces should deter (and, if pressed, coerce) while simultaneously promoting rather than detracting from a fundamental strategic stability — the understanding that U.S. actions are not intended to deny the other side a basic retaliatory capability.10

Together, this means the United States should want a nuclear strategy, force, and posture focused on the ability and preparedness to use nuclear weapons in discriminate, tailored, and controlled ways. The logic of any such nuclear employment should focus on escalation advantage: to demonstrate Washington’s willingness to escalate to the nuclear level, and to continue escalating if grave provocations continue, but also its readiness to restrain further escalation and ultimately deescalate if the opponent is prepared to comply with reasonable demands. Accordingly, Washington should want a nuclear arsenal that can provide varying options for controlled, graduated forms of nuclear escalation in line with this basic logic that allow for different potential employment strategies (such as tit for tat or intensifying escalation), since the optimal targeting strategy is likely to vary based on the particular contingency. To be most coercively useful, such strikes should be designed to influence (or complement other efforts to influence) the sub-nuclear conflict in ways that shift the burden of escalation further onto the adversary and thus be advantageous to the United States.

Together, this means the United States should want a nuclear strategy, force, and posture focused on the ability and preparedness to use nuclear weapons in discriminate, tailored, and controlled ways.

That is, the ideal nuclear employment strategy is one that not only demonstrates political will but also, along with U.S. and allied conventional efforts, affects the sub-strategic battle in ways that make an adversary’s counter- or further escalation less attractive. For instance, the United States would benefit from having nuclear options that could heavily damage a Chinese invasion flotilla designed to assault U.S. allies in the Western Pacific and that could exercise similar effects against Russian forces attacking or directly supporting an incursion into the Baltics.11 Such capabilities would enable the United States not only to demonstrate its resolve to cross the nuclear threshold, but also markedly increase the degree of escalation the opponent would have to undertake to remedy the loss and continue the fundamentally offending action (such as the invasion of a U.S. ally in the Western Pacific or in Eastern Europe). This role would be especially important if the United States lacks plausible conventional options for exercising such an effect, especially without undertaking separate, dramatic forms of escalation (for instance by significantly expanding the scope of the battlefield or hitting new, especially sensitive classes of targets). The U.S. military fielded these types of capabilities during the Cold War but abandoned them in the post-Cold War era. This was defensible in an era of untrammeled U.S. conventional superiority; it is not in one in which the Russians and Chinese may have plausible theories of victory against U.S. allies and important partners.

From a targeting perspective, this puts a premium on being able to strike at differing sorts of targets depending on the stage of escalation — to be able to strike effectively at what the opponent values but also to communicate at stages of escalation short of general war a meaningful degree of restraint.

Accordingly, nuclear weapons that could significantly damage or impair such targets, but with lessened collateral damage, would be particularly attractive. Such weapons that would be especially useful in this context might include those with a lower yield, those that could be employed in ways that would create less radioactivity, and those that would travel on trajectories and from platforms that would be less likely to generate an opponent’s fear that they were part of or precursor to a general or attempted disarming attack. Furthermore, this nuclear strategy puts a high value on an exquisite, responsive, resilient, and supremely capable nuclear command, control, and communications (NC3) architecture.12

At the same time, limited use would not substitute for the ability to conduct large-scale and general nuclear strikes. Rather, the effectiveness of discriminate options would in fact depend upon their connection to that possibility. Consequently, the United States would need to retain the capacity to destroy Russia and China’s most valued targets, including their industrial bases and national leadership in their protected redoubts. While attacks on leadership should, as a general principle, be withheld until the very last stages of escalation, it is crucial for the United States to be able to destroy an opponent no matter where he goes, especially at the end of a chain of deliberate escalation when an adversary has had the opportunity to hide and defend himself. Accordingly, the United States needs capabilities to assuredly — and, ideally, promptly and with reduced collateral damage — destroy even targets in hardened and deeply buried facilities. It is vital to underline that this is critical for retaliatory strikes — and thus for stability — and far less, actually, for more aggressive nuclear strategies, which can aspire to decapitating enemy leaderships before they have a chance to seek safety or concealment.

Notably, this nuclear strategy does not emphasize or rely on the ability to attack an enemy’s strategic nuclear forces or command-and-control. While it does not exclude the potential value of having options to degrade an opponent’s strategic arsenal or command-and-control ability (for instance, to make an opponent’s counter-escalation options less attractive), it generally counsels restraint regarding pursuit of strategic counterforce capabilities, let alone their employment, particularly in light of the countering responses such pursuit is likely to engender.

Communication with the adversary before and during a conflict is crucial to the effectiveness of such a strategy, since ultimately it is predicated on persuading — indeed, coercing — an adversary to agree to end a war on terms acceptable (and ideally favorable) to the United States without triggering escalation to a level of war beyond what anyone would want. Accordingly, Russia and China need to understand the logic of U.S. nuclear strategy. It is not about denying their retaliatory capability nor confined to large-scale options. Rather, it is about demonstrating to them in the most painful terms that the United States has the resolve and the ability to impose progressively greater — and ultimately the greatest — damage and risk on them if they transgress core American and allied interests, and that it has the capabilities to make such a strategy plausibly implementable on bases that will play to American, rather than their, advantages.

U.S. declaratory policy should reflect this. Ambiguity about the precise conditions under which the United States would employ nuclear weapons and how it would do so are advisable for familiar reasons, but greater clarity about and emphasis on the options the United States possesses and will possess to pursue the strategy laid out here would be helpful. This may involve less changes in the wording of formal statements than shifts in how the United States exercises its forces, for instance, by building in contingencies involving deliberate escalation, and allowing the circulation of reports of such exercises.


The basic point is to make clear to Moscow and Beijing that Washington is prepared to respond to dramatic escalation on their part with plausibly implementable strategies of its own. This is especially important because the stakes over cognizable contingencies today are lower than they were during the Cold War (primarily because neither China nor Russia poses the kind of totalistic threat that many viewed the Soviet Union as representing). More apocalyptic strategies are more credible when defeat itself would seem apocalyptic, as it did to many during the Cold War. When limited defeat over peripheral interests would not appear to constitute such a catastrophe, more credible strategies are needed. The 2018 Nuclear Posture Review represented an important and commendable starting point in this direction, especially with its decision to develop a low-yield warhead for U.S. submarine-launched ballistic missiles, but the U.S. military will need to go further.

The basic point is to make clear to Moscow and Beijing that Washington is prepared to respond to dramatic escalation on their part with plausibly implementable strategies of its own.

The premise for all this is that limited nuclear war is possible. Crossing the nuclear threshold would be staggeringly dangerous, as things always might get completely out of control, leading to an apocalyptic exchange. But this is not the same as saying that such escalation to total war would necessarily happen. This is fundamentally for two reasons: because combatants under the nuclear shadow would always have the strongest possible incentive to avoid triggering the apocalypse, since doing so would almost certainly result in their own destruction, but, at the same time, advantage at the nuclear level (the highest imaginable) would be dominating. Thus the side willing and able to escalate to the nuclear level and come out ahead would have a commanding edge. Accordingly, even as the United States should seek to minimize the degree to which it relies on nuclear weapons in its defense strategy — for both strategic and moral reasons — its defense strategy must nonetheless reckon with the reality that limited nuclear war is possible and, unless anticipated and provided for, could well be an attractive and even rational course of action for opportunistic or motivated opponents.

In closing, it is worth emphasizing what the logic of this strategy would be. The United States is wisely committed to sustaining its grand strategy of alliances in key regions of the world, a strategy that is most conducive to preserving an enduringly favorable balance of power and thus international order for Americans. This is a fundamentally conservative approach, one that seeks to defend what is established rather than transform the world or upend regional orders.

This requires a defense strategy and posture that will deter a rising and increasingly assertive China and an alienated and more capable Russia. That, in turn, requires that Beijing and Moscow believe the United States might realistically put its strategy into effect despite the attendant risks and the relatively lower stakes compared with those at issue in the Cold War. In a situation of substantial mutual vulnerability over stakes that are important but not truly central to the United States, the best strategy to serve U.S. political ends is one focused on advantageously managing escalation in a way that seeks to keep or shift the burden of dramatic escalation onto Moscow or Beijing. This is highly suited to a strategy focused on defense rather than expansion or transformation, and thus is the best way to achieve the goal Hans Morgenthau set out for a wise foreign policy, that “the task of armed diplomacy [should be] to convince the nations concerned that their legitimate interests have nothing to fear from a restrictive and rational foreign policy and that their illegitimate interests have nothing to gain in the face of armed might rationally employed.”13


Elbridge Colby is director of the Defense Program at the Center for a New American Security. He was deputy assistant secretary of defense for strategy and force development from 2017 to 2018.  


Image: Ministry of Digital Development, Communications and Mass Media of the Russian Federation


1 For China, see, "'China Seeks Hegemony': America's Pacific Commander Offers a Military Warning," Wall Street Journal, Feb. 24, 2016,; Ely Ratner, "Rising to the China Challenge: Prepared Statement Before the House Committee on Armed Services," Feb. 15, 2018,; for Russia, see, A. Wess Mitchell, "Remarks at the Atlantic Council," Oct. 18, 2018,; Christopher S. Chivvis, "Russia's Determination to Revise the Post-Cold War Order," RAND blog, Sept. 30, 2016,

[2] National Security Strategy of the United States, December 2017,; “Summary of the National Defense Strategy of the United States of America: Sharpening the American Military’s Competitive Edge,” January 2018,

3 As classically laid out in Thomas C. Schelling’s Arms and Influence (New Haven, CT: Yale University Press, 1966) and The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960).

4 See RAND scorecard report on China, for instance: Eric Heginbotham et al., The U.S.-China Military Scorecard: Forces, Geography, and the Evolving Balance of Power, 1996–2017, (Santa Monica, CA: RAND Corporation, 2015),

5 Karl Mueller, David A. Shlapak, Michael W. Johnson, and David Ochmanek, “In Defense of a Wargame: Bolstering Deterrence on NATO’s Eastern Flank,” War on the Rocks, June 14, 2016, ; Nuclear Posture Review (Defense Department, February 2018),

6 “Annual Report to Congress: Military and Security Developments Involving the People’s Republic of China 2018,” Department of Defense,; Robert O. Work, “So, This Is What It Feels Like to Be Offset,” Speech at Center for a New American Security, June 21, 2018,

7 Joseph L. Votel, Charles T. Cleveland, Charles T. Connett, and Will Irwin, “Unconventional Warfare in the Gray Zone,” Joint Force Quarterly, no. 80 (January 2016),; Hal Brands, “Paradoxes of the Gray Zone,” Foreign Policy Research Institute, Feb. 5, 2016,

8 Jonathan F. Solomon, “Demystifying Conventional Deterrence: Great-Power Conflict and East Asian Peace,” Strategic Studies Quarterly 7, no. 4 (Winter 2013): 117–57,; Elbridge Colby and Jonathan Solomon, “Facing Russia: Conventional Defence and Deterrence in Europe,” Survival 57, no. 6 (2015): 21–50,; Elbridge Colby and Jonathan F. Solomon, “Avoiding Becoming a Paper Tiger: Presence in a Warfighting Defense Strategy,” Joint Force Quarterly, no. 82 (July 2016),

9 Elbridge Colby, “Nuclear Weapons Aren’t Just for the Worst Case Scenario,” Foreign Policy, Aug. 4, 2016,

10 For more on views on strategic stability, see Elbridge A. Colby and Michael S. Gerson, eds., Strategic Stability: Contending Interpretations (Carlisle Barracks, PA: U.S. Army War College Press, 2013),

11 For a recent treatment of the problem of limited nuclear war and potential scenarios involving it, see John K. Warden, “Limited Nuclear War: The 21st Century Challenge for the United States,” Livermore Papers on Global Security no. 4 (July 2018),

12 “Nuclear Matters Handbook 2016,” Office of the Deputy Assistant Secretary of Defense for Nuclear Matters, ch. 6, accessed Nov. 26, 2018,

13 Hans J. Morgenthau, “Another ‘Great Debate’: The National Interest of the United States,” American Political Science Review 46, no. 4 (December 1952): 978,