The False Allure of Escalation Dominance


“Anyone who thinks they can control escalation through the use of nuclear weapons is literally playing with fire. Escalation is escalation, and nuclear use would be the ultimate escalation.”

So warned then-Deputy Secretary of Defense Robert Work in his 2015 congressional testimony. Coming at a time of seemingly more prevalent nuclear saber-rattling, Work’s statement may appear to indicate a laudably cautious policy and a useful reminder to America’s rivals about the need for restraint in a crisis.

Yet the notion that nuclear escalation is a binary and prohibitively risky strategic choice belies decades of theorizing and planning for the possibility of limited nuclear war. Indeed, the prospect of limited nuclear escalation was central to NATO’s deterrence policy for much of the Cold War. Beginning in the 1960s, as Soviet nuclear forces approached parity with U.S. capabilities, policymakers increasingly sought to moderate the risks of mutually assured destruction with what came to be known as “limited nuclear options”: weapons and warfighting strategies that might include initiating nuclear use, but were designed to terminate conflict short of Armageddon.

One of the earliest and most influential concepts associated with these efforts was Herman Kahn’s “escalation ladder,” which defined 44 “rungs” on a metaphorical ladder of escalating conflict. The rungs ranged from “sub-crisis maneuvering” to civilization-destroying nuclear exchanges. In between were two dozen distinct levels of escalation beyond the threshold of nuclear use, including such “limited” attacks as non-lethal demonstrative detonations, tactical strikes on military forces, and small-scale attacks on civilians.

In this context, Kahn also introduced the term “escalation dominance,” which became shorthand for one school of thought in deterrence and nuclear strategy. The idea posits the ability of a state to maintain such a markedly superior position over a rival, across a range of escalation rungs, that its rival will always see further escalation as a losing bet. Such dominance, the thinking goes, serves as the most effective possible deterrent to conflict, as well as the most reliable means for managing escalation if deterrence fails. As Colin Gray and Keith Payne put it in their famous “Victory is Possible” article in 1980, “an adequate U.S. deterrent posture is one that denies the Soviet Union any plausible hope of success at any level of strategic conflict.” Beyond U.S.-Soviet competition, analysts have continued to apply the concept to such contemporary strategic challenges as the India-Pakistan rivalry, a potential Taiwan Strait crisis, and even Middle Eastern terrorism.

Is escalation dominance still relevant to U.S. strategy today? A debate on this question may soon be revived. In the next few months, the Trump administration will publish the results of its Nuclear Posture Review, the first comprehensive review of U.S. nuclear strategy and capabilities since 2010. Among the topics most worth watching is whether and how the review addresses U.S. strategy for managing escalation. Perhaps even more than the Obama administration’s team in 2010, the current Nuclear Posture Review authors must confront a growing risk of escalation from limited regional conflicts to nuclear war. Tensions with North Korea may pose the most obvious of such risks at the moment, but unfortunately, the problem is broader and more deeply rooted. Two related factors account for this growing risk.

First, the erosion of U.S. conventional superiority — especially with the growing sophistication of “anti-access / area denial” capabilities — means adversaries may be increasingly tempted to engage in quick, limited, territorial aggression — a fait accompli — against a U.S. ally. For China, this could be over contested islands in the South or East China Seas or over Taiwan. For Russia, it could be anywhere in its “near abroad,” even against the Baltic NATO allies. On the Korean Peninsula, U.S. conventional superiority remains intact, but major gains in North Korea’s nuclear and missile capabilities could serve as a shield for a number of different limited provocations against South Korea.

However, even where U.S. power projection advantages over regional challengers may be strained, they are still sufficiently robust to potentially foil or reverse an attack. So the result of this evolution in conventional military balances is a set of increasingly plausible scenarios in which a nuclear-armed power launches a limited attack on its neighbor, only to find itself on the verge of conventional defeat at the hands of the United States.

The second risk driver is the presence, in precisely these limited-war scenarios, of asymmetric interests between the United States and its potential adversaries. In conflicts over, say, Taiwan, South Korea, or Lithuania, U.S. rivals could plausibly calculate that their resolve is decisively superior to that of the United States. While these conditions are not new, their danger is heightened when paired with the conventional balances described above. In such cases, nuclear threats or even limited nuclear use could become an aggressor’s last-ditch war-winning strategy. “Escalate to de-escalate,” as some Russians may (or may not) put it. Or, as Brad Roberts has put it, U.S. rivals have “nuclear theories of victory.”

In light of these dynamics, there is a natural temptation for U.S. policymakers to seek solutions in a strategy like escalation dominance. Interest in the concept has waxed and waned over the past 50 years, peaking in the later part of the Cold War. Even then, many considered this level of ambition to be more dangerous than stabilizing in light of the approximate U.S.-Soviet parity in capabilities that prevailed then, and the concept was never codified explicitly in U.S. declaratory policy. But after the Cold War, U.S. planners effectively became, in Elbridge Colby’s words, “accustomed to escalation dominance,” thanks as much to the course of historic events as to deliberate strategy. Analysts often saw American escalation dominance as a key to regional stability. Today the concept continues to generate some interest among think tanks and other analysts focused on nuclear strategy and regional security issues. Escalation dominance may also exert some indirect influence on policymakers’ strategic thinking. Many military and civilian leaders are not steeped in details of nuclear strategy debates, but may find that the concept’s philosophy and intellectual pedigree resonates with their intuition about the need for dominance.

Indeed, escalation dominance is superficially appealing. Its deterrent logic is easy to grasp. What could dissuade a regional challenger more effectively than comprehensive superiority? And it comports well with the strategic habits of mind ingrained in a generation of U.S. policy makers by unrivaled post-Cold War military superiority.

Nevertheless, the concept has always suffered from serious flaws and is particularly poorly suited to the regional deterrence challenges the United States faces in the 21st century. At least five problems are cause for concern.

Asymmetric stakes. As already noted, one of the key sources of escalation risk is the asymmetry of interests between prospective combatants. The most plausible scenarios of escalation involve core, vital interests of the challengers juxtaposed with American extended deterrence commitments to allies and partners. For instance, the United States seeks to deter a Chinese attack on Taiwan. But under extreme circumstances, Chinese leaders may well see the survival of their regime riding on a military victory, while U.S. stakes in protecting Taiwan lie in more abstract goals of maintaining stability, order, and deterrent credibility. Will the United States really engage in nuclear war over Taiwan? Or, in a NATO-Russia conflict, risk trading Virginia Beach for Vilnius?

Such potential imbalance of interests is a long-standing problem of extended deterrence. Thomas Schelling famously observed that escalation may take the form of a “competition in risk taking,” and therefore may be governed at least as much by “balance of resolve” as by balance of capabilities. This poses a challenge for any escalation management strategy, but is especially problematic for escalation dominance, which relies heavily on superiority in capabilities. While theoretically plausible, establishing “dominant” resolve as well as dominant capabilities is a difficult standard to meet in a conflict where a capable, nuclear-armed rival has already gambled great stakes.

Conventional balance. For escalation dominance to produce the desired deterrent effect, both parties must recognize one side’s superiority at multiple levels of escalation, below and above the nuclear threshold. Superiority itself is of no use if it goes unrecognized or doubted. On this point, escalation dominance faces another formidable obstacle. As alluded to above, the longstanding conventional capability gap between the United States and both China and Russia has shrunk in recent years. The match-ups are becoming too closely balanced to confidently predict that one side will prevail, especially in those fait accompli scenarios in which an adversary seeks a quick victory enabled by local tactical advantages. Even analysts who believe that relative regional shortfalls in U.S. conventional strength have been exaggerated could readily agree that America’s ability to dictate the pace and intensity of a conventional war has diminished.

Information and decision-making challenges. Crisis decision-making is subject to a host of extra-rational factors and information limitations, which makes it difficult for actors to precisely evaluate their rivals’ escalation thresholds. In a crisis involving nuclear weapons, factors like time pressures, risk tolerance, incomplete or conflicting intelligence, and psychological stress are a few of the unpredictable elements that complicate fine-tuned chess moves of escalation. This is true regardless of one’s escalation management strategy, but its pathologies may be magnified by a strategy that depends on establishing and communicating superiority at every turn. The party asserting dominance may be more apt to underestimate its adversary’s resolve, while the ostensibly “dominated” party may become more risk-tolerant in the face of a tempting but fleeting opportunity for successful escalation of its own.

Similarly, misperception of adversary behavior, intent, and communication is a common feature of international affairs and military history, including in cases of crisis escalation. Even if U.S. decision makers are confident of their own information and analysis, it is not possible to reliably discern adversary values and interpretations of thresholds. In a seminal 2008 study of escalation, RAND Corporation analysts concluded that, relative to the Cold War, “predicting how [U.S. opponents] will perceive U.S. actions is not dramatically easier and, in some cases, can be even more challenging.”

Influence of new technologies. Rapidly improving and proliferating capabilities in long-range precision strike, and cyber and space operations have complicated the concept of an escalation “ladder.” Kahn’s original ladder with 44 rungs spanning conventional and nuclear war was already quite complex. Today, the menu of non-nuclear options available to strategic competitors to signal or attack each other has expanded dramatically. Whether the proper metaphor for 21st-century escalation is a ladder or a vortex or something else entirely, there is little doubt that contingency planning for escalation is harder than ever. Where in the hierarchy of escalation does a disabling but reversible Chinese attack on U.S. military satellites belong? Is a Russian cyber attack on the U.S. electrical grid more or less escalatory than missile strikes on European bases? This kind of complexity also compounds the challenges of misperception. If a shared framework among potential adversaries for understanding escalation thresholds was elusive in the Cold War, it is only farther from reach today.

Peacetime provocation. Quite apart from the dynamics of crisis decision-making, pursuit of escalation dominance as a declaratory policy is, itself, escalatory. It could exacerbate unreasonable fears of U.S. aggression and prompt otherwise unnecessary arms races. This point is not simply a matter of taking the dovish side of the eternal security dilemma. There are certainly limits to the importance of declaratory policy, and U.S. challengers clearly have many motivations beyond reacting to U.S. provocation. Still, official discussion of managing escalation through dominance or comprehensive superiority supports the prevailing narrative of opponents of American power, and thereby may help empower those factions in the Chinese and Russian governments most dedicated to frustrating U.S. interests. Moreover, U.S. advocacy of escalation dominance may complicate alliance politics, undermine assurance, and impede cohesion within NATO.

In combination, these factors make U.S. pursuit of escalation dominance a risky strategy.

So what is the alternative? Richard Betts once argued that just because strategy is not always possible does not mean it is unnecessary. Is escalation dominance like that — the worst framework except for all the others? No, but neither can the United States settle for the status quo, allowing the risks of limited nuclear war to creep upward under the venerable but wrong-headed rationale that “there is no such thing as limited nuclear war.”

Unfortunately, there are no easy answers to these escalation management challenges. But policymakers — including those crafting the current Nuclear Posture Review — can still pursue escalation advantages without pursuing dominance. Rejecting escalation dominance need not be a prescription for abandoning extended deterrence, preemptively conceding defeat in certain scenarios, or curtailing modernization of U.S. strategic forces.

To the contrary, the United States must still convince challengers that it retains the capabilities to hold their interests at risk using reliable, discriminate levels of force. This means continuing to seek robust and flexible options across a wide range of capabilities, which should include maintaining nuclear forces with a diverse set of platforms, yields, and operational characteristics. Non-nuclear options for managing escalation are also highly relevant here, including precision strike, cyber and counter-space tools, as well as defenses against equivalent potential adversary capabilities. And to ensure coherent integration of these capabilities, the United States can still do much to improve its planning, exercising, and command and control for limited war scenarios with nuclear rivals.

But policymakers should be equally clear that the purpose of building discriminate escalation options is not dominance. The purpose is to extend the concept of strategic stability beyond its classic conception of avoiding general war and into the realm of managing escalation in the midst of limited war. There is no single strategy for implementing this: each potential adversary requires a tailored approach. And framing this approach as a set of advantages rather than outright dominance draws, perhaps, a fine distinction. But overall, to make U.S. regional deterrence credible and sustainable, it is vital to resist the urge to convince all rivals that they will always and everywhere be strategically inferior to the United States. This was probably never feasible, but certainly is not today. And if some strategists find this judgment too dour, they can look for some measure of comfort in Schelling, who recognized that uncertainty and shared risk have long been at the heart of strategic deterrence and escalation management.

Finally, Russia and China should take heed of this same advice. All but the first of the problems noted above apply to them as well, and they should ensure that their own maturing philosophies on escalation shun the false allure of dominance.


Michael Fitzsimmons is Visiting Research Professor at the Strategic Studies Institute, U.S. Army War College. The views expressed are those of the author and do not necessarily reflect the official policy or position of the Department of the Army, Department of Defense, or the U.S. Government.

Image: U.S. Air Force