In September, U.S. Secretary of Defense Ash Carter called for NATO to better integrate conventional and nuclear deterrence. “Across the Atlantic, we’re refreshing NATO’s nuclear playbook to better integrate conventional and nuclear deterrence, to ensure we plan and train like we’d fight, and to deter Russia from thinking it can benefit from nuclear use in a conflict with NATO—from trying to ‘escalate to de-escalate,’ as some there call it.”1
Earlier this year, Assistant Secretary of Defense Robert Scher stated that the Department of Defense is “working to ensure an appropriate level of integration between nuclear and conventional planning and operations.”2
At first glance, these statements may seem discordant with the long-standing U.S. view that nuclear weapons are distinct and apart from other military capabilities. This fundamental distinction is reflected in almost every aspect of how nuclear weapons are treated. Only the president can authorize the employment of nuclear weapons, and the United States maintains a unique declaratory policy explaining and limiting the conditions under which their use would be considered. Nuclear weapons require special operational considerations and safeguards. Personnel with access to nuclear weapons, facilities, and materials require additional screening and monitoring. The notion of conventional and nuclear weapons integration is often portrayed as threatening to weaken or break down the special status and “profound caution” afforded to nuclear weapons.
Yet, the statements by Carter and Scher are consistent with U.S. nuclear deterrence policy and the underlying philosophy from which it stems. Much of the integration debate boils down to differing characterizations of the ends and ways of conventional and nuclear weapons integration, a question of objectives, and how to achieve them. Ensuring “an appropriate level of integration” requires a mix of maintaining and improving key aspects of integration. Doing so serves U.S. security interests, in particular, providing effective nuclear deterrence, without increasing reliance on nuclear weapons, blurring the distinction between non-nuclear and nuclear conflict, or lowering the threshold for nuclear use.
Indeed, better integration will reduce the likelihood of an adversary’s nuclear use while maintaining the U.S. threshold at its appropriately high level. Deeper integration between conventional and nuclear planning and operations is essential to ensure U.S. nuclear weapons can continue to effectively fulfill their fundamental deterrence role in the 21st century.
Fulfilling an Enduring Role
The primary role of U.S. nuclear weapons has not changed since the 2010 Nuclear Posture Review, that is, to deter nuclear attacks on the United States and its allies and partners. Deterrence is strengthened by the arsenal’s capacity to perform its secondary role of “achieving U.S. and allied objectives if deterrence fails.” Neither of these roles is new, but the security environment in which U.S. nuclear posture and strategy must support these roles is changing. So too are the nuclear deterrence challenges for which the United States must prepare.
The 2014 Quadrennial Defense Review (QDR) report illuminated this changing context, stating that U.S. nuclear forces communicate to “potential nuclear-armed adversaries that they cannot escalate their way out of failed conventional aggression.”3 More than a massive surprise nuclear attack in peacetime or a suicide attack on the U.S. homeland, the QDR highlights the danger of a calibrated and limited attack amid a conventional conflict gone awry.
Developments in Russia, North Korea, and China demonstrate why this is a salient deterrence challenge. These countries may see tacit or explicit nuclear threats as a potent means of demonstrating to U.S. leaders that U.S. stakes are materially lower than their own, thereby weakening U.S. commitment to come to the aid of allies should they find themselves in a regional conflict.4 Numerous analysts have observed that Russia sees limited nuclear attack as a potential means of de-escalating a conventional conflict by demonstrating a favorable asymmetry of stakes or at least views threatening such a course of action as useful for deterring U.S. engagement at the outset. The working assumption would be that following an initial limited nuclear attack, the side with more skin in the game would be more willing to continue the fight and accept the attendant risk of further nuclear escalation. At the other end of the capability spectrum, North Korea may see the threat of limited nuclear escalation early in conflict as an effective means of deterrence and wartime coercion in the face of vastly superior U.S. and South Korean conventional forces. Finally, China has a no-first-use declaratory policy, but debates within China over what constitutes first use and whether the declaratory policy would hold in a conflict suggest some consideration for threatening or using nuclear weapons for purposes other than responding to nuclear attack.
There is a strategic logic to these considerations of threatening nuclear use for purposes other than deterring nuclear attack. Notions of accepting and even enhancing escalation risk and of utilizing nuclear weapons to achieve a favorable outcome in conventional war have precedent in the theory of military strategy and the actual strategies of nuclear-weapon states, including the United States during the Cold War.5
Because there is escalation risk inherent in any conflict between nuclear-armed states, it would be irresponsible to extend security commitments to U.S. allies and pledge to deter conventional aggression without taking into account how potential foes may deliberately or haphazardly bring nuclear weapons into play. Thus, when dealing with a nuclear-armed adversary, there is intrinsic and unavoidable linkage between the conventional and nuclear realms. Ignoring that fact invites peril.
Managing escalation in confrontations with nuclear-armed adversaries is an essential element of U.S. national security strategy. Escalation management seeks to protect the vital interests of the United States and its allies while convincing an adversary to refrain from using the full military means at its disposal. Deeper integration of nuclear and conventional planning and operations serves three ends of escalation management. Preparing to achieve these ends weakens the coercive nuclear strategies adversaries may develop when contemplating aggression and therefore ultimately strengthens the United States’ ability to deter a conflict from starting in the first place.
The first objective of integration is to strengthen one’s ability to deter adversaries from choosing nuclear escalation in a conventional conflict. Because nuclear weapons enable a country to rapidly inflict massive levels of damage, a military confrontation with a nuclear-armed state is fundamentally different from one with a non-nuclear adversary. Whatever other political objectives brought the United States into such a war, deterring first use of nuclear weapons would automatically become a central U.S. objective.
Second, integration aims to strengthen one’s ability to achieve U.S. and allied objectives if deterrence fails. Presumably, the United States would enter any conventional conflict with a set of war aims tied to political objectives. Those aims and objectives are unlikely to disappear after a limited nuclear attack, although they might change somewhat in substance or priority. Integration facilitates efforts to keep conventional operations and nuclear posture aligned with the political objectives they are designed to support.
Third, integration increases the likelihood of successfully restoring deterrence following an adversary’s nuclear weapons use. If an adversary resorts to using nuclear weapons in a conventional conflict, the possibility that it would use them again will seem very real. Just like deterring first use of nuclear weapons would be a central U.S. objective in any conflict with a nuclear-armed adversary, deterring further nuclear use and escalation would automatically become a central objective once an adversary crossed the nuclear threshold. Scher touched on this as well, saying that “integration means being prepared to restore deterrence following adversary nuclear use, so that failure to deter first use does not translate into failure to deter subsequent nuclear use.”6
There are three principal ways to improve integration consistent with the special status afforded to U.S. nuclear weapons. The first two represent areas where improvement is needed, and the third is principally a matter of ensuring that current capabilities remain viable. Together, these aspects of integration strengthen U.S. escalation management strategy by helping the United States avoid miscalculation leading to nuclear war. If deterrence fails, they help ensure that the president’s options for responding to a nuclear attack are not limited to ceding victory to the aggressor or ordering a massive nuclear counterattack. Integration enables the additional options of continuing the conventional war after adversary nuclear weapons use without responding in kind or responding in kind while continuing conventional military operations.
Planning conventional campaigns to shape adversary escalation calculus. Deterring nuclear escalation within a conventional conflict is an important 21st century challenge. The United States must prepare to operate under the nuclear shadow while navigating through the fog of conventional war. The core principles of nuclear deterrence remain the same after the fighting starts: willingness to respond forcefully and purposefully to nuclear weapons use and willingness to show certain forms of restraint as long as the adversary does not use nuclear weapons. Yet, effectively communicating resolve and restraint—the ying and the yang of the deterrence message—amid the confusion and emotion of war may require additional measures.
The threat of response must effectively convey that the United States and its allies will not allow an adversary to escalate its way to victory, split alliances through coercive threats or nuclear attack, or achieve a favorable military situation by using nuclear weapons. At the same time, U.S. officials must sustain and communicate the promise of restraint that is inherent in every deterrence threat, the assurance that choosing to remain below the nuclear threshold will spare the adversary the threatened cost of crossing it.
Harmonizing this deterrence strategy with U.S. conventional operations is a key point of integration. As Scher explained, “[I]ntegration means conventional operations must be planned and executed with deliberate thought as to how they shape the risk that the adversary will choose nuclear escalation.”7 The United States may need to forgo certain objectives, such as regime change, that would likely lead adversary leadership to see nuclear weapons use as its only viable option for survival. In order for the adversary to understand and believe that this restraint is contingent on it not using nuclear weapons, the United States would also need to avoid military operations the adversary is likely to perceive as a precursor to regime change or disarming strategic attacks. This would likely require withholding attacks on adversary nuclear forces, nuclear command and control, political leadership, and assets or capabilities critical to an adversary’s basic ability to defend its homeland.
As the United States, Russia, China, and others expand their strategic postures and operational concepts to include conventional, space, cyber, and nuclear forces, integration requires looking across domains and functional capabilities to fully analyze escalation risks. Will a particular cyber- or space operation impact an adversary’s nuclear operations? How will adversary leadership interpret the intent of the operation? If an operation is intended to strip away adversary intelligence, surveillance, and reconnaissance capabilities, how will it impact the adversary’s ability to gauge U.S. and allied limited aims?
The twin objectives of effectively waging the conventional campaign and seamlessly executing a nuclear deterrence strategy will likely engender tension and require difficult trade-offs. For example, the United States may be at a disadvantage in executing in a conventional conflict if it does not launch conventional strikes against adversary air defenses or conventional missile systems. If these targets are located in a nuclear-armed adversary’s homeland, however, U.S. officials may be concerned that adversary leadership will perceive such actions as indicative of a drive for regime change. Integration cannot eliminate these tensions and trade-offs, but it can help illuminate critical decision points. This will help senior decision-makers weigh the benefits and escalation risks of certain courses of action. Ultimately, whether certain conventional military operations should be ruled out or curtailed in order to reduce the risk of nuclear escalation is a presidential decision. The purpose of this aspect of integration is to enable informed decisions about U.S. strategy in confrontations with nuclear-armed adversaries and to ensure U.S. military and diplomatic means are poised to execute that strategy as precisely as possible.
Strengthening conventional resiliency to nuclear operations. An adversary may see nuclear escalation as an efficient means to shift the conventional military balance in a conflict, even if only for a short period of time. Strengthening the resiliency of conventional operations to adversary nuclear attack is a second way to strengthen integration.
Conventional resiliency includes the ability to communicate, operate, and resupply in a nuclear environment. There are a variety of means for enhancing resiliency, including hardening, redundancy, and dispersing forces and points of debarkation to reduce vulnerable single points of failure.8 Yet, enhancing resiliency is also a matter of intelligence and imagination. How might an adversary employ its nuclear forces to disrupt U.S. conventional operations? What are the vulnerabilities an adversary may target? Exploring these questions will be essential as the United States enters a period of technological and operational innovation to maintain conventional deterrence against Russia and China.9
This aspect of integration contributes to managing escalation for two reasons. First, it preserves presidential flexibility in the face of limited nuclear use. Wherever possible, the president should have the option of continuing the conventional fight even after an adversary employs nuclear weapons. Furthermore, this should not be a binary strategy where conventional and nuclear options for responding to a nuclear attack are mutually exclusive. Denying presidential flexibility would essentially offer the adversary the ability to dictate the means of the conflict by choosing nuclear escalation. This would more likely favor the side that perceives itself as conventionally weaker and therefore more reliant on nuclear weapons.
Second, conventional resiliency reduces the potential benefits of attacking U.S. forces with nuclear weapons. If a limited nuclear attack is unlikely to result in a decisive operational-military advantage, then using nuclear weapons carries high risk but scant rewards. In other words, conventional resiliency contributes to deterrence.
For both of these reasons, ignoring conventional resiliency invites adversaries to elevate the role of nuclear weapons in their strategies.
Providing integrated response options that are limited and credible. Possessing credible options for responding to first use of nuclear weapons reinforces all three ends of escalation management (deterring nuclear escalation and, if deterrence fails, restoring deterrence and achieving other U.S. and allied objectives). Potential adversaries may conclude they can calibrate a nuclear attack to coerce the United States into capitulating without causing sufficient destruction to provoke a large nuclear response. The ability to respond to an attack purposefully and proportionately helps convince adversaries that no such sweet spot exists. Of course, what constitutes a purposeful and proportionate response would depend on the context. As a general rule, the U.S. response would need to be integrated into the conventional campaign to avoid disrupting U.S. conventional operations. In order to deter rather than spur another nuclear attack, the response would need to be consistent with U.S. efforts to communicate its resolve and its limited war aims to the adversary. Finally, it would also need to be integrated into the broader political strategy for orchestrating an end to the conflict.
This ability underpins the strategic message that the United States will defend the core interests of its allies even in the face of nuclear threats. Relying solely on large-scale response options may indeed be credible for deterring attacks on the U.S. homeland, but as the sole means for reacting to a limited attack overseas, it runs the risk of appearing as a hollow bluff to allies and adversaries alike. Limited options are thus an important part of extending deterrence and assuring U.S. allies.
This is not a call for returning to nuclear artillery or using nuclear weapons for tactical military effects that could be achieved with conventional forces. Rather, the United States should retain the diversity and flexibility of its current arsenal, particularly its nuclear-capable bomber and fighter aircraft. These aircraft are key to effectively deterring and responding to limited nuclear attack because they can be used to demonstrably signal deterrence messages (they are the only component of U.S. nuclear forces that is visible and recallable), they can be forward deployed in crisis and conflict and well as in peacetime, and the weapons they can carry contribute to the range of yields in the U.S. stockpile. Under the current stockpile reduction plan, these aircraft will carry a single type of gravity bomb (the life-extended B61), and the bomber force will also carry a single type of nuclear-armed cruise missile (the air-launched cruise missile, to be replaced with the modernized long-range standoff weapon).
Owing to tremendous reductions in warhead numbers and types over the past three decades as a result of negotiated and unilateral actions, the U.S. arsenal and suite of delivery platforms have reached a minimum acceptable level of diversity and flexibility. Although some numerical reductions may still be possible, warhead and delivery platform types should not be further reduced in the near term. On the contrary, those remaining capabilities should be sustained and, where necessary to remain viable, modernized to maintain the existing range of credible and proportionate response options.
In concert with U.S. land- and submarine-based ballistic missiles, this suite of capabilities is minimally sufficient for enabling integrated, limited options for achieving U.S. objectives after a limited attack when the president judges that non-nuclear responses alone are insufficient. For example, after a limited nuclear attack on U.S. forces fighting abroad, the president may judge that the United States needs to demonstrate its willingness to respond with nuclear weapons. A conventional response, even if capable of destroying the same target on a comparable timescale, would not have the same psychological impact as a response in kind and risks inviting a follow-on nuclear strike or fracturing an alliance. A larger nuclear response could be disproportionately destructive, triggering physical and operational effects that provoke rather than deter further escalation.
Under these conditions, a limited nuclear response might succeed in restoring nuclear deterrence and sustaining the alliance. Success would not be guaranteed, but the risks of alternative options would likewise be severe.
Some contend that the objectives of U.S. escalation management strategy, including deterring an adversary from escalating across the nuclear threshold and restoring deterrence if ever it fails, would be better served by reducing nuclear integration rather than by maintaining or increasing it. These critiques typically reduce integration to just its third element—limited response options—and advance one or more of three basic arguments.
First, some believe efforts to ensure nuclear and conventional integration lower the nuclear threshold by making it easier for the United States to use nuclear weapons first in a conflict. These claims are often tied to the supposed pursuit of new nuclear weapons with lower yields that make them more “usable” than those in the existing arsenal. Neither this general claim nor its supporting elements are consistent with the scope of the U.S. nuclear modernization plans or the defense strategy it supports. Low-yield weapons have been a part of the U.S. stockpile for half a century, and Pentagon officials have stated unequivocally that current plans, including the life extension of the B61 gravity bomb, do not entail expanding the range of yields already available.10 More generally, despite the deliberate ambiguity inherent in U.S. declaratory policy, the notion of U.S. first use for tactical advantage or for de-escalating a conventional conflict is far removed from U.S. nuclear strategy, which focuses on credible options for responding to and therefore deterring nuclear attack.
Second, some would claim that forgoing the ability to respond in a limited way would strengthen deterrence because it would imply the threat of massive nuclear retaliation in response to even a limited attack. Eliminating limited options would thus decrease the likelihood of adversary first use. Although automatic large-scale retaliation would indeed negate any rational gains an adversary may hope to achieve through a limited attack, the threat to do so only works if the adversary believes it. The United States cannot responsibly count on all adversaries concluding that the threat of massive retaliation is always credible. It is difficult to imagine that, in the immediate aftermath of a limited nuclear attack against a U.S. ally, even critics of limited response options would advise the president to order a massive strike on the grounds that credibility demands it or that total escalation is inevitable. Removing limited options would weaken deterrence if adversaries believe available U.S. nuclear responses are far less likely to produce an acceptable outcome for the United States and its allies, let alone a desirable one. Similarly, sole reliance on large-scale nuclear response options would do a poor job of dissuading allies from seeking independent deterrent capabilities.
Given the costs of capitulating to nuclear coercion and the risks of a strategy based on threatening massive response, what would the United States gain by removing the option of a limited nuclear response? Some contend that a conventional response to limited nuclear use is the better course under any circumstances. They believe that limited options are undesirable because they make it more likely a president will unnecessarily choose a nuclear response and because pursuing them drives requirements for types of nuclear weapons that do not increase U.S. security. A purely conventional response might indeed be the best way to limit further nuclear escalation and achieve U.S. and ally war aims in some cases, but it is unreasonable to assume a priori that this will always be the case.
Contrary to the objectives of escalation management strategy, solely continuing the conventional fight might encourage further nuclear attacks aimed at finding the U.S. and allied pain threshold or measuring the relative stakes and resolve of the two sides. This is especially likely if the adversary’s goal is to stop the conventional campaign and its initial nuclear attack fails to achieve this goal but does not elicit the type of U.S. response it most fears. Furthermore, a strategy of continuing the conventional campaign toward victory after adversary limited nuclear use would likely provide the enemy with ample time and incentive to employ additional nuclear attacks. Ultimately and in anticipation of or in response to further nuclear attacks, holding to the conventional-only response might create pressure for negotiating a cessation to hostilities at all costs, implying U.S. capitulation and an adversary’s successful implementation of its nuclear coercion strategy.
Moreover, credible options for deterring subsequent nuclear strikes provide an essential underpinning of a conventional-only response to nuclear attack. In order to restore deterrence, the United States would need to convince the adversary that any further nuclear use would result in costs that outweigh potential gains. For example, an adversary may believe that a limited nuclear attack or even a demonstration shot will compel the United States to capitulate. If that strategy fails and the United States continues fighting, adversary leadership might resort to a nuclear strike on U.S. military forces in the theater to raise the stakes and blunt the ongoing campaign. The fact that deterrence already failed once would no doubt raise questions about U.S. defense strategy, but the likelihood of a U.S. nuclear response in this case might be perceived as higher than the chance the United States would retaliate with nuclear weapons in response to a first attack that inflicted little or no damage.
The threat of a large-scale response might succeed in deterring follow-on attacks, but it might not be perceived as credible, particularly if the adversary has a survivable arsenal. A large-scale response may also be incompatible with the U.S. political objectives associated with the conventional fight. Thus, we find it difficult to imagine a U.S. president sustaining conventional operations after an initial nuclear attack if massive retaliation is the only nuclear option for responding to a second limited attack.
The better course is neither to prejudge presidential decisions nor surrender the option most likely to be credible and aligned with political objectives. Some fear that calls for greater integration imply a dangerous level of confidence in U.S. escalation-control strategy. Yet, effective deterrence requires an approach to escalation risk that avoids absolutism of either extreme. Confidence in one’s ability to deliberately start a limited nuclear war between major nuclear powers and control subsequent escalation would be the ultimate miscalculation, but inherent uncertainty about one’s ability to control escalation should not translate into certainty that any nuclear use would automatically lead to uncontrolled escalation up to global annihilation.
The point about uncertainty is that no one can know for certain what the eventual outcome would be. Virtually everyone would want the president at least to try to limit escalation following an adversary attack. Consequently, it makes no sense to voluntarily relinquish the kind of credible response options below the level of massive retaliation that every president has required since the Soviet Union first acquired the ability to inflict unacceptable damage on the United States.
Nuclear weapons are unique in their ability to inflict and deter violence and should never be treated as more powerful analogues to conventional munitions. Ensuring and strengthening integration of nuclear and conventional planning and operations is consistent with this long-standing principle. It is also critical to maintaining an appropriately balanced approach to escalation management and meeting the most salient of contemporary deterrence challenges.
1. U.S. Department of Defense, “Remarks by Secretary Carter to Troops at Minot Air Force Base, North Dakota,” September 26, 2016, http://www.defense.gov/News/Transcripts/Transcript-View/Article/956079/remarks-by-secretary-carter-to-troops-at-minot-air-force-base-north-dakota.
2. Robert Scher, Statement before the Senate Armed Services Subcommittee on Strategic Forces, February 9, 2016, p. 3, http://www.armed-services.senate.gov/imo/media/doc/Scher_02-09-16.pdf (hereinafter Scher statement).
3. Department of Defense, “Quadrennial Defense Review,” p. 14, March 2014, http://archive.defense.gov/pubs/2014_Quadrennial_Defense_Review.pdf.
4. Nikolai N. Sokov, “Why Russia Calls a Limited Nuclear Strike De-Escalation,” Bulletin of the Atomic Scientists, March 13, 2014, http://thebulletin.org/why-russia-calls-limited-nuclear-strike-de-escalation; Keir A. Lieber and Daryl G. Press, “Coercive Nuclear Campaigns in the 21st Century: Understanding Adversary Incentives and Options for Nuclear Escalation,” PASCC Report, No. 2013-001 (March 2013); U.S. Office of the Secretary of Defense, “Annual Report to Congress: Military and Security Developments Involving the People’s Republic of China 2016,” April 2016, http://www.defense.gov/Portals/1/Documents/pubs/2016%20China%20Military%20Power%20Report.pdf. See Brad Roberts, The Case for U.S. Nuclear Weapons in the 21st Century (Stanford, CA: Stanford University Press, 2016).
5. David S. Yost, “The History of NATO Theater Nuclear Force Policy: Key Findings From the Sandia Conference,” Journal of Strategic Studies, Vol. 15, No. 2 (June 1992): 228-261.
6. Scher statement, p. 3.
8. Joint Defense Science Board/Threat Reduction Advisory Committee Task Force, “The Nuclear Weapons Effects Enterprise,” U.S. Office of the Under Secretary of Defense for Acquisition, Technology, and Logistics, June 2010, http://www.acq.osd.mil/dsb/reports/ADA523661.pdf.
9. For more information, see Bob Work, speech on the Third Offset Strategy at the Reagan Defense Forum, November 7, 2015, http://www.defense.gov/News/Speeches/Speech-View/Article/628246/reagan-defense-forum-the-third-offset-strategy.
10. Scher statement.
Vincent A. Manzo is a fellow in the Defense and National Security Group at the Center for Strategic and International Studies. Aaron R. Miles is a fellow at the Center for Global Security Research at Lawrence Livermore National Laboratory. The views are those of the authors.