Policy Paper

Living With a Nuclear-Arming North Korea: Deterrence Decisions in a Deteriorating Threat Environment

As North Korea’s nuclear program advances, the US must pursue both diplomacy and an updated allied deterrence strategy
Part of the Project
38 North
By Brad Roberts

As the latest chapter in US diplomacy aimed at denuclearizing North Korea comes to another unhappy end, it is time to take stock. As the diplomatic strategy adjusts to new realities, it is also time to revisit deterrence strategy. Having been put on the back burner these last few years, that strategy is in need of attention. Some difficult choices lie ahead about how to further adapt and strengthen that strategy and the associated regional deterrence architecture. The most contentious issues are likely to be choices about nuclear policy and posture. This special report reviews the key decisions now facing policymakers in the United States, South Korea and Japan when it comes to a nuclear-arming North Korea.

On this page
Share on twitter
Share on facebook
Share on linkedin
Share on email

Executive Summary

As the latest chapter in US diplomacy aimed at denuclearizing North Korea comes to another unhappy end, it is time to take stock. As the diplomatic strategy adjusts to new realities, it is also time to revisit deterrence strategy. Having been put on the back burner these last few years, that strategy is in need of attention. Some difficult choices lie ahead about how to further adapt and strengthen that strategy and the associated regional deterrence architecture. The most contentious issues are likely to be choices about nuclear policy and posture. This report reaches the following key judgments and recommendations:

  • North Korea’s progression from the nuclear threshold to a heavily armed nuclear state with intercontinental reach is doing serious damage to the foundations of strategic stability in Northeast Asia by eroding confidence in the ability and will of the United States to fulfill its security guarantees to South Korea and Japan.
  • While working for denuclearization, the United States, South Korea, and Japan have also been working to adapt and strengthen deterrence so as to stay ahead of the emerging threat. This project was put on the back burner in recent years as a sign of good faith. It is time to bring it back to the front burner.
  • The US nuclear umbrella, a legacy of decisions made in very different security environments in 1991 and 2010, is no longer fit for purpose as presently composed. It must be modified to enable improved signaling of collective resolve to stand up to North Korea’s nuclear bullying.

A more “NATO-like” approach could serve the US-Republic of Korea (ROK) and US-Japan alliances well. But this requires coming to terms with confusion about how NATO practices extended nuclear deterrence. A more NATO-like approach would entail changes not just to the posture of US nuclear forces but also to the organization and practice of consultations as well as new approaches to strategy and policy development.


Alas, the latest chapter in US diplomacy aimed at persuading Pyongyang to abandon its nuclear ambitions has been no more successful than prior chapters. North Korea is now armed with nuclear weapons. It is also arming itself further, meaning it has some flexibility in responding to further developments in the policies and postures of the United States and its allies. Barring war or regime change, North Korea’s nuclear weapons appear to be here to stay. How should the United States and its two Northeast Asian allies, South Korea and Japan, respond?

No American president is likely ever to formally recognize North Korea as a nuclear-armed state or to give up on attempting to roll back its nuclear capabilities. But the nuclear problem now requires something much more than a continued US commitment to a diplomatic breakthrough. It requires also a renewal of deterrence, including especially US extended deterrence.

The existing deterrence architecture is no longer fit for purpose. Without some strengthening and adaptation of that architecture, the nuclear danger presented by North Korea will only grow as it arms further. For the United States and its allies to be successful in this endeavor requires coming to terms with broad resistance in the United States and elsewhere to any steps that might be seen as increasing the role and salience of US nuclear weapons.

This special report reviews the key decisions now facing policymakers in the United States, South Korea and Japan. What are the strategic consequences of North Korea’s improving nuclear capabilities? What are the basic policy response options? How robust is the existing deterrence posture? What can and should be done to strengthen and adapt deterrence to new circumstances? What adjustments are needed to the nuclear deterrence the United States extends to South Korea and Japan? What lessons can be learned from NATO?1The views expressed here are those of the author and should not be attributed to any institution with which he is or has been affiliated. The author is grateful to Mike Albertson, Paul Choi, Markus Garlauskus and Shane Smith for comments on an earlier draft of this essay and to participants in roundtables sponsored by the National Defense University and the Henry L. Stimson Center where the paper was discussed. This report also builds on the recent 38 North article by Shane Smith on this topic. See: “Renewing US Extended Deterrence Commitments Against North Korea,” 38 North, May 13, 2020,

What’s at Stake?

A skeptic might argue that a country with thousands of nuclear weapons can afford to ignore the threats of a country with dozens. After all, any major war between North Korea and the United States would result in the North’s destruction and the demise of the Kim dynasty, so why not discount Pyongyang’s threats? There are not many examples of this kind of complacency and wishful thinking—except when it comes to discussion of US extended nuclear deterrence policy, where the resistance in the United States to doing anything new and different is substantial.

North Korea’s progress in arming itself with nuclear weapons and long-range delivery systems is affecting US and allied security interests in various perverse ways. The Democratic People’s Republic of Korea’s (DPRK)’s’s nuclear forces are now or will soon be of sufficient scale to present a credible threat to the existence of South Korea and Japan. Such destruction has been well beyond the reach of North Korea’s general-purpose military forces for seven decades.2A massively destructive nuclear strike by North Korea on South Korea and/or Japan would also result in the deaths of many foreigners. There are approximately 230,000 Americans resident in South Korea and 90,000 in Japan. See Victor Cha, “Giving North Korea a ‘bloody nose’ carries a huge risk to Americans,” Washington Post, January 30, 2018,

Pyongyang evidently hopes to exploit this vulnerability to try to persuade Seoul and Tokyo to stand by or restrain Washington rather than support the United States in a future DPRK-US confrontation.3For a discussion of the role of nuclear weapons in North Korean military strategy, see Brad Roberts, “The First New Strategic Problem,” chapter 2 in The Case for U.S. Nuclear Weapons in the 21st Century (Stanford, Calif: Stanford University Press, 2015), especially pp. 58-74. The expectation of future attempts at nuclear-backed coercion has raised anxieties among leaders in South Korea and Japan. This has created significant new demands on US efforts to assure its allies that they remain safe under its nuclear umbrella (a critical requirement for the maintenance of alliance cohesion, especially in crisis and war). This comes at a time of deep political acrimony in the bilateral political relationships of both countries with each other and with the Trump administration, further fueled by the doubts cast by the president on his willingness to defend allies in times of crisis.4Uri Friedman, “America’s Alliance System Will Face One of its Biggest Tests Yet,” Atlantic, May 23, 2019,

North Korea’s nuclear progress is also bringing it to a point where it may soon have the capability and capacity to overwhelm the missile defense protection of the United States homeland. The military operator of those defenses, Northern Command, has predicted that this may happen within the next five years.5Jason Sherman, “NORTHCOM: US to assume “increased risk” against North Korean ICBMs in 2025,” Inside Defense, July 26, 2020, Simply put, North Korea may deploy long-range missiles more rapidly than the United States can deploy missile interceptors.

The threat to the United States may not be existential, in the way it is for US allies, but it would be severe nonetheless. American vulnerability to even a limited attack would raise questions about the willingness of the United States to defend its allies in times of crisis. Indeed, the prospect of such vulnerability has already done so, as experts and officials have debated the potential “de-coupling” effects of such vulnerability on US relations with its allies.

Since the 1999 Missile Defense Act, it has been the policy of the United States to “stay ahead” of the threats to the homeland from rogue states. As North Korea’s forces grow and improve, this will prove increasingly challenging (and also increasingly damaging to strategic stability with Russia and China).6See Brad Roberts, “Missile Defense: Past, Present, and Future” in Roberts, ed., Fit for Purpose? The U.S. Strategic Posture in 2030 and Beyond (Livermore, CA: Center for Global Security Research, 2020).

North Korea’s nuclear progress also impacts the perceived effectiveness and thus credibility of a strategy for the conventional defense of the Republic of Korea that depends on the projection of power from US ports and bases elsewhere in the region, from Hawaii, and from the continental United States. Here it is the kinetic impact that matters, as opposed to the psychological impact associated with the manipulation of vulnerability and risk, as discussed above.

The emerging regional missile force may allow North Korea to have a credible anti-access, area-denial strategy. And its limited nuclear force could seem operationally attractive to a regime whose survival might depend on crippling US operations at relatively few vulnerable nodes. As the National Defense Strategy Commission concluded in its 2018 review of US defense strategy and posture, the United States could well lose a regional war against a nuclear-armed adversary, in part, because it has failed to account for the particular operational challenges that such enemies might be able to impose on US and allied military operations.7Eric Edelman and Gary Roughead, Providing for the Common Defense: The Assessment and Recommendations of the National Defense Strategy Commission (Washington, DC: United States Institute of Peace, 2018), This uncertainty over US effectiveness at the conventional level of war comes at a politically sensitive point in the US-ROK defense relationship, as reflected in difficult discussions about burden-sharing and OPCON transfer (that is, transfer of operational control of allied forces from the United States to the ROK).

North Korea’s nuclear progress also deepens the central dilemma facing US alliances in Northeast Asia: how to strengthen deterrence vis-à-vis North Korea without damaging the strategic relationship with China. A main theme of US deterrence strategy is that it is tailored to the particular requirements of different potential adversaries. In Northeast Asia, that tailoring is complicated. The two US allies do not have identical views of the regional security environment, with both focused on North Korea, but Japan much more focused than South Korea on China. Some of the possible adjustments of the US/allied regional deterrence architecture that may be necessary to deal with a nuclear-arming North Korea, as discussed further below, could generate reactions by China that would have damaging consequences for the security of South Korea and/or Japan and for US forces deployed there.

In assessing the consequences of a nuclear-arming North Korea for the security interests of the United States and its allies, a fundamental uncertainty exists: what impact will the North’s nuclear progress have on its future behavior. Will the DPRK become more cooperative in its relations with others because its possession of a nuclear deterrent allows it to shake off past insecurity? Or will it become more aggressive because its leader(s) become more risk-tolerant, believing the new deterrent protects them from reprisal?

The messages from Pyongyang are mixed. In July 2020, Kim stressed the importance of nuclear weapons as a guarantee of the country’s survival in a speech to an assembly of military veterans:

War is an armed clash which can be unleashed only against a weak one. None can now make little of us…Thanks to our reliable and effective self-defence nuclear deterrence, the word war would no longer exist on this land, and the security and future of our state will be guaranteed for ever.8“Supreme Leader Kim Jong Un Makes Speech at Sixth National Conference of War Veterans,” KCNA, July 28, 2020,

But recently, Kim has also spoken about “shocking” and “offensive” action to chart “a new path with the United States and South Korea.”9Choe Sang-Hun, “What Kim Jong-un’s Latest Threats Say About His Trump Strategy,” New York Times, January 1, 2020, An unnamed senior government official has also promised that Pyongyang will be “more zealous for our important projects aimed to repay the US with actual horror and unrest for the sufferings it has inflicted upon our people.”10Hyonhee Shin and Heekyong Yang, “North Korea says Pompeo undercuts its interest in restarting talks,” Reuters, March 30, 2020, This may just be bluff and bluster. Or it may foreshadow a return to the lethal conventional provocations of the past, whether on the scale of a decade ago or something much more lethal.

In sum, a nuclear-armed and -arming North Korea is already having a perverse effect on US and allied security interests. This impact can be expected to become more damaging over the coming decade if and as the balance of strategic power and influence continues to shift in ways favoring Pyongyang.

What Are the Basic Policy Options?

Motivated as they are to prevent this damage, leaders in Washington, Seoul and Tokyo must now decide how to adjust strategy. Notionally, their list of options is long:

  1. Conduct preventive strikes intended to cripple North Korea’s emerging nuclear forces and infrastructure;
  2. Conduct a preventive war intended to remove the regime before the nuclear threat has fully matured;
  3. Return to “strategic patience” with the hope of out-waiting Pyongyang at the negotiating table while sanctions take their toll;
  4. US withdrawal from its alliance with the ROK in order to leave the problem to Seoul, Beijing and Tokyo;
  5. Nuclearization by South Korea and perhaps also Japan;
  6. Recognize the DPRK as a nuclear weapon state in order to pursue some form of arms control;
  7. Refocus diplomatic efforts away from denuclearization and onto containment while ramping up military preparedness; or
  8. Pursue a dual-track approach combining diplomatic efforts to engage the DPRK with steps to adapt and strengthen deterrence with the hope of creating the stalemate necessary for a future political breakthrough.

Practically, as opposed to notionally, the list of options is much shorter. Options one and two would likely have catastrophic consequences for South Korea, Japan and perhaps also the United States. Option three has already been shown to fail. Options four, five and six would have substantial negative impacts on the US global role and on the nonproliferation regime. This leaves options seven and eight. Option seven should be unacceptable to all, as the US and its allies should never abandon the effort to find a political solution. This leaves option eight.

The dual-track approach has, in fact, been the mainline of US policy over multiple administrations. The Clinton administration pursued the Agreed Framework while adding missile defenses to the US strategic toolkit to deal with the North Korean intercontinental ballistic missile (ICBM) threat. The George W. Bush administration exerted significant political and economic pressure on North Korea in support of a diplomatic solution while pursuing its “new triad” of strategic forces aimed at negating the emerging nuclear deterrents of “rogue states.” The Obama administration extended an “open hand” to Pyongyang while working with Seoul and Tokyo to strengthen deterrence through a “comprehensive approach” encompassing both nuclear and non-nuclear means. It also worked with its allies to update strategic messaging and to physically demonstrate the US capability and intent to extend nuclear deterrence on their behalf.11See: “Quadrennial Defense Review Report,” United States Department of Defense, February 2010,; “Nuclear Posture Review Report,” United States Department of Defense, February 2010,; and “Ballistic Missile Defense Review Report,” United States Department of Defense, February 2019, See also Brad Roberts, Extended Deterrence and Strategic Stability in Northeast Asia (Tokyo: National Institute for Defense Studies, 2013).

This dual-track effort has been on partial hold during the Trump administration. The effort to strengthen and adapt deterrence was slowed by the suspension of certain activities, such as large-scale US-ROK joint military exercises. It also became less coherent, as South Korea, Japan and the United States began to move in different directions.

It is time to re-accelerate and re-integrate efforts to improve the deterrence architecture, while also adjusting diplomatic strategies. In short, it’s time to move on to option eight. Reshifting strategic focus in this way invites a series of questions about the appropriate goals of the two tracks and how to ensure they do not work at cross purposes.

How Robust Is the Current Deterrence Architecture?

How much acceleration and integration is necessary? How close are the United States and its allies to having the deterrence architecture they need? Looking back over three decades, the United States and its allies have spent a lot of time, energy and money strengthening the regional deterrence architecture and adapting it to new purposes. General-purpose military forces have been modernized. Missile defenses have been introduced and improved. Strike capabilities have improved. Operations in cyberspace have become more resilient. Deterrence strategy has been crafted. Moreover, US military planning is now squarely focused on modern regional conflict against nuclear-armed adversaries. In 2018, then Chairman of the Joint Chiefs of Staff, Joseph Dunford characterized such conflict as “multi-dimensional, multi-domain, and trans-regional” in character.12Paula Thornhill and Mara Karlin, “The Chairman the Pentagon Needs,” War on the Rocks, January 5, 2018,

Progress should not, however, be conflated with success. How should we understand success? A robust deterrence architecture is one that can be expected to have a decisive effect on a potential adversary’s calculus of the benefits, costs, and risks of attacking the United States and/or its allies in times of crisis, of escalating in times of war, or of conducting lethal and other major provocations in circumstances short of war. It should be tailored to the particular capabilities, values and mindset of specific adversaries and to the particular characteristics of the wars that adversary has prepared. It should also make the best use of available deterrence means.

Does the existing deterrence architecture in Northeast Asia meet these criteria? Without interviewing North Korean leaders about their personal “deterrence calculus,” it is nearly impossible to know. But one very important data point comes from the bipartisan US National Defense Strategy Commission. In November 2018, as noted above, it concluded that the United States could well lose such a war of the kind defined by Dunford. Its findings describe a Department that does not understand the particular dynamics of regional wars under the all-domain escalation shadow or the strategies of our adversaries to deter and defeat us and our allies. The Commission also concluded that the US and its allies have failed to both understand the political and operational dilemmas adversaries intend to create for us with their escalatory acts and to develop the needed operational concepts to impose such dilemmas on their adversaries.13Eric Edelman and Gary Roughead, Providing for the Common Defense: The Assessment and Recommendations of the National Defense Strategy Commission (Washington, DC: United States Institute of Peace, 2018),

In addition, the credibility of US promises to defend its allies from attack and to respond as necessary, perhaps even with nuclear weapons if the vital interests of those allies are put at risk, has eroded in recent years. North Korea’s growing capability to make Americans pay a price for defending an ally is one key factor. Another is the political gap between the United States and its allies that has widened in the Trump period. That credibility has been further undermined by the simple fact that the United States proved to be unwilling to use military force to prevent the emergence of this new threat, despite many times threatening to do so.

What Can and Should Be Done to Strengthen and Adapt Deterrence?

The first step on the new pathway should be to address the problems identified by the National Defense Strategy Commission. That is, the US-ROK alliance should put its intellectual house in order. It must develop a core set of ideas about the nature of the war North Korea might bring them, the operational dilemmas it hopes to force upon the Combined Forces Command, and the decision points its hopes to force upon allied leaders and publics. They must then develop strategies for forcing Pyongyang to bear the burden and risks of escalation or to choose to de-escalate. The alliance needs a theory of victory in both peacetime and war—that is, a collection of ideas about how its actions will result in preferred restraint by Pyongyang.14Brad Roberts, On Theories of Victory, Red and Blue, Livermore Paper No. 7 (Livermore, CA: Center for Global Security Research, 2020),

This is a task for the US-Japan alliance as well, given the important role it would play in a regional war initiated and expanded by North Korea. These ideas can take shape within the separate bilateral alliances but at some point, they must converge. The political obstacles to improved trilateral military cooperation are, of course, substantial but they will disappear in a crisis once North Korea has fired the first shot. Policymakers in both Seoul and Tokyo should do what is necessary now to reduce the risks of ever having to face such a crisis. Trilateral operational defense planning is unlikely but improved coordination to ensure complementarity ought to be possible.

This effort to strengthen and adapt deterrence must account for the diversifying strategic deterrence toolkit, which now includes not just nuclear weapons but also missile defenses, non-nuclear missile strike systems, as well as capabilities in cyberspace and outer space. Effective multi-domain deterrence requires that the whole be more than the sum of the parts—that is, that the integration of capabilities produces positive synergistic effects. Two forms of integration stand out as salient and also as especially complicated in a tripolar context.

The first is offense/defense integration. A mix is essential because neither tool alone is reliable for deterrence. Integration requires that the tools can be flexibly applied as needed to achieve desired effects on Pyongyang’s calculus of the benefits, costs and risks of different courses of action. In the absence of a trilateral approach, one state may be more vulnerable than another. These differences could drive problematic political demands—for new preemptive strike capabilities, for example. Asymmetric vulnerabilities would also be eagerly exploited by North Korea in times of crisis and war.

The other is conventional/nuclear integration. This does not imply nuclear war-fighting or the integration of tactical nuclear weapons at the tactical or operational levels of war. Rather, it implies the ability to display in peacetime, crisis and war the nuclear capabilities of the United States, as well as the resolve of the two alliances to stand up to DPRK nuclear coercion and nuclear attack. It also implies the ability to sustain combat operations even if North Korea crosses the nuclear threshold. This brings us to the most important steps that should be taken to strengthen and adapt the regional deterrence architecture: adjustments to its nuclear component.

What Adjustments Are Needed to Extended Nuclear Deterrence?

To answer this question requires having a clear understanding of the existing approach. The US nuclear umbrella in Northeast Asia is the result of two decisions. In 1991, the George H.W. Bush administration decided to remove all nuclear weapons from East Asia as part of a broad set of Presidential Nuclear Initiatives implemented in parallel with the Soviet Union, then Russia. Accordingly, the United States withdrew from Asia and destroyed all of its forward-deployed tactical nuclear weapons, ceased deploying nuclear weapons aboard naval surface combatants, and put into storage nuclear-armed Tomahawk missiles while maintaining the ability to redeploy them aboard attack submarines in times of crisis.

The second decision was made in 2010, when the Tomahawks reached the end of their service lives. As they were retired, the Obama administration decided to make a fleet of nuclear-capable F-15 and F-16 fighter-bombers armed with nuclear gravity bombs “globally available” in times of crisis to support its extended deterrence commitments to allies.15See the Nuclear Posture Reviews of the Obama and Trump administrations. These decisions to maintain non-strategic nuclear capabilities explicitly for the nuclear umbrella to allies in Europe and Asia were both made with a clear understanding that the triad of strategic nuclear forces also plays a central role in underwriting US commitments to its allies. Accordingly, the United States regularly exercises those forces as, for example, with regional deployments of nuclear-capable B-52 bombers.

Obviously the security environment in Northeast Asia has evolved in dramatic and troubling ways since 1991 and 2010. For a long time, creeping North Korean nuclearization cast no significant doubt on the efficacy of the US approach. This author was one of many officials in multiple administrations making the case that North Korea’s progress in developing a nuclear arsenal had not fundamentally altered the strategic balance of power in the region. Our argument hinged on the fact that the United States can destroy any target in North Korea by nuclear means within a very short period of time. With retaliation assured, deterrence should be robust.

This argument fails to account adequately for the fact that it is American will and not its capability that is in question. North Korea has tested ICBMs three times and conducted a large underground nuclear test that it claims is a thermonuclear ICBM warhead. If a US president ever considered employing nuclear weapons on behalf of an ally whose vital interests are in jeopardy, his or her willingness to do so would be burdened by the vulnerability of the US homeland to nuclear retaliation.

It is not just American strategic intentions that are in question. The new North Korean nuclear threat is aimed in part at breaking the will of both the ROK and Japan to stand with the United States in a mounting crisis and limited war. Displays of US nuclear capability and resolve are not a substitute for displays of their resolve to stand up to nuclear coercion and to defend their interests even under nuclear attack.

In short, North Korea’s progress in assembling a nuclear force capable of threatening South Korea, Japan and the United States puts a new premium on displays of collective nuclear resolve by the US-ROK and US-Japan alliances. For this purpose, the legacy US nuclear umbrella in Northeast Asia falls short. Adjustments are needed to strengthen the credibility of alliance threats to respond to North Korean nuclear attacks despite new nuclear vulnerabilities.

It is hardly a surprise, therefore, that there has been a rising chorus of South Korean voices calling for something substantially different from the 1990s-vintage version of the US nuclear umbrella—a minority advocate for the development of an independent nuclear deterrent.16Toby Dalton, Byun Sunggee, Lee Sang, “South Korea Debates Nuclear Options,” Carnegie On Line, April 27, 2016,; and “South Korea opposition party calls for nuclear armament,” Financial Times, August 14, 2017, Others seek greater insight for the ROK government into the nuclear operational plans of the US military.17Chun In-bum, “U.S. Extended Deterrence and the Korean Peninsula’s Changing Threat Environment,” Council on Foreign Relations, April 29, 2019, Private conversations with experts in Japan reveal similar thinking there.

Still, others argue for something more “NATO-like,” sometimes with the supplementary view that US allies in Europe are America’s “first-tier allies” who “got a better deal on nuclear deterrence than your second-class Asian allies.”18This view is frequently expressed in Track 1.5 dialogues. In my experience, there is a lot of confusion about the actual practice of extended nuclear deterrence in the NATO context. But there is also something useful to be gained in examining that practice.

What Lessons Can Be Learned From NATO?

What could it mean to become more “NATO-like” in the practice of extended nuclear deterrence in Northeast Asia? Although an approach built by a multilateral alliance in a different context cannot readily be imported into Northeast Asia, NATO’s approach and experience offer a number of useful lessons.

The first lesson is about the need to understand that effective deterrence requires much more than nuclear weapons. NATO’s nuclear deterrent has six main components.19See: Office of the Deputy Assistant Secretary of Defense for Nuclear Matters,“Nuclear Matters Handbook 2020,”; North Atlantic Treaty Organization, “Active Engagement, Modern Defence: 2010 NATO Strategic Concept,” November 19, 2010,; North Atlantic Treaty Organization, “The Deterrence and Defense Posture Review 2012,” May 20, 2012,; “Nuclear Posture Review Report,” United States Department of Defense, February 2010,; “Nuclear Posture Review Report,” United States Department of Defense, February 2018,; and Jessica Cox, “Old and new challenges show need to bolster NATO’s nuclear deterrent,” King’s College London, April 10, 2019,

  1. Nuclear Forces: These include: 1) the strategic nuclear forces of its three nuclear-armed members, which NATO describes as the “supreme guarantee of the security of the allies”; and 2) the “sharing arrangements” whereby some allies host US nuclear weapons and operate dual-capable aircraft (DCA) certified to deliver nuclear weapons in wartime under the command of the US president. The sharing arrangements serve two deterrent functions: the presence of US nuclear weapons in Europe automatically implicates the United States in any response to an attack on them, thus reinforcing the “transatlantic link” at the core of the North Atlantic Treaty; and the sharing of nuclear roles among numerous allies reinforces the collective defense message that an attack on one will be treated as an attack on all. The Alliance has expressed repeatedly an enduring commitment to an “appropriate mix” of nuclear and non-nuclear capabilities for deterrence.
  2. Operational Planning: Planning by the three nuclear-armed allies for employment of their independent nuclear forces in support of alliance commitments is a national matter. Alliance operational planning focuses on potential utilization of the DCA in flexible deterrent operations. As the Alliance does not formally identify any state as an enemy, it maintains no plans for the employment of nuclear weapons against any country; instead, it exercises a so-called adaptive planning process so that it can act quickly and competently in times of war to plan nuclear operations. Guidance for these planning efforts is provided by defense ministers.
  3. Training and Exercises: This includes training and exercises for crisis deployment and for nuclear strike operations. It also includes training and exercises of the conventional air cover that would be provided to nuclear strike operations. Some allies not participating in the DCA mission participate in such operations, called Support of Nuclear Operations With Conventional Air Tactics (SNOWCAT).
  4. Declaratory Policy: These statements convey the intentions of the Alliance about when and why it might employ nuclear weapons—and when not. The Alliance as such issues such policy via periodic updates to its Strategic Concept and regular communiques when heads of state and government gather, usually every two years. Such statements are supplemented with leadership statements from the national capitals of the three nuclear-armed members.
  5. Nuclear Policy Development and Oversight: Defense ministers come together regularly to discuss nuclear issues. When they do so, they convene as the Nuclear Planning Group (NPG); a separate entity is required to account for the fact that the minister from France does not participate, as France did not re-join the nuclear coordination mechanism when it rejoined NATO in 2009. The NPG is supported by the High-Level Group, which consists of representatives from national capitals and is chaired by the United States. It is also supported by the NATO Nuclear Planning Group Staff Group, which consists of representatives in Brussels and is chaired by a member of the NATO international staff. Over the last decade, the Alliance’s senior body, the North Atlantic Council (NAC), has played a growing role in formulating the Alliance’s nuclear strategy.
  6. A Consultation Mechanism for the Possible Employment of Nuclear Weapons: US nuclear weapons may be employed only on the order of the president of the United States. He or she takes a recommendation from the Supreme Allied Commander Europe (SACEUR), who takes a recommendation from the North Atlantic Council. The consultative process serves various interests. The European allies want assurance that the United States will not employ nuclear weapons when not absolutely necessary—and that it will do so when absolutely necessary. The United States wants assurance that its allies will share in the responsibility for the decision to employ nuclear weapons.

Toward a More NATO-Like Posture in Northeast Asia

It would not be practical to simply lift the NATO nuclear model and apply it willy nilly to Northeast Asia, given differences in alliance structures, geography, strategic culture, historical experiences and domestic politics. But it certainly possible for the US-ROK and US-Japan alliances to become more NATO-like with regard to extended nuclear deterrence while also respecting the unique circumstances and requirements of the region.

As a point of departure in this direction, recognize that some of the components noted above already have analogs in Northeast Asia. The central question is how to tailor them further to new challenges. For example, defense ministers currently meet to discuss a wide range of issues, including nuclear deterrence, but not in a specialized forum for that particular purpose. The creation of such a dedicated ministerial-level mechanism could have both deterrence and assurance value. As another example, existing training and exercise programs could do more to account for a possible nuclear dimension to an ongoing conventional conflict.

The most difficult questions relate to how to modify the legacy nuclear umbrella to new requirements. Notionally, the NATO “sharing arrangement” could be replicated in South Korea with US nuclear weapons permanently deployed there along with dual-capable fighter-bombers that would be flown by pilots from both countries.20Amy Woolf and Emma Chanlett-Avery, Redeploying U.S. Nuclear Weapons to South Korea: Background and Implications in Brief, Congressional Research Service, September 14, 2017, As this seems politically impractical even in the new strategic circumstance, an alternative would be to deploy DCA but not weapons. Weapons could remain stored in the United States or could be stored on Guam. Or preparations could be made to store them on Guam in times of crisis. If sharing arrangements were to be established that include roles for South Korean and/or Japanese pilots, an operational planning process for nuclear employment would have to be created.

Alternatively, or in addition, pilots from South Korea and/or Japan could be trained to provide air cover or advance reconnaissance for nuclear strike missions. South Korean conventional missile forces could exercise suppression of North Korean air defenses to pave the way for a strike. ROK special operations forces could even train using protective gear for combat operations in a North Korean environment contaminated by radiation in the aftermath of a strike.

In addition, there is no NPG-like consultative mechanism in the US-ROK and US-Japan alliances. Although there are many existing mechanisms to enable military and political consultations between the United States and its allies in times of crisis and war, none of these is tailored explicitly to enable the needed nuclear consultations. Moreover, there is no mechanism tailored to address the needed trilateral consultations in times of nuclear crisis and war. This is perhaps the biggest gap in the existing deterrence architecture and promises to become even more glaring as the threat develops.

These seemingly dry questions of bureaucratic structure mask political issues of major consequence. Imagine a situation in which a threat has suddenly emerged to the sovereignty and integrity of South Korea and/or Japan and the US president is deliberating whether to employ nuclear weapons to try to protect the ally. In such a situation, the leaders of that ally will want a seat at the table to understand the consequences of different choices and to ensure their views are heard. While they count on the United States to do the right thing and to know what that is in the circumstance, they expect to be consulted and listened to—as they should. At NATO, the actual table sits in conference room one, along with a formal consultative process, which is shaped by guidance agreed by the allies. South Korea and Japan should not need to ask whether they will have a seat at the table, where the table is, and how consultations will proceed.

Creating a consultative process now would be reassuring to South Korea and Japan and would help to address the rising concern about the political health of these alliances. It would also help to reduce the likelihood that it would ever be needed, by signaling alliance resolve to act when threatened. It would also reassure the American president that he or she will receive timely input from those whose interests are most at risk.

Other Valuable Lessons From NATO

If the first lesson that can be learned from NATO is that there’s more to a credible nuclear deterrent than nuclear weapons, what other lessons should be learned?

A second lesson is that there is value in committing to maintain an “appropriate mix” of nuclear and non-nuclear capabilities for deterrence (now also including missile defenses). This commitment serves the alliance well by compelling regular leadership focus and continuing debate in the expert community about what adaptations might be possible or necessary in a changing security environment. NATO periodically reviews this mix to assess whether it is “fit for purpose.” In 2012, it conducted a full Deterrence and Defense Posture Review and the resulting report conveyed to allied publics the roles of the different instruments of deterrence and the commitment to continue to adapt the deterrent in response to changing circumstances.

The US-ROK and US-Japan alliances would be well served by adopting a similar formulation and conducting similar reviews. One of the primary benefits would be great clarity and consensus about the particular roles of nuclear weapons in that deterrent mix.

A third lesson is that deterrence strategy alone should not define an alliance’s nuclear strategy. As British military historian Michael Howard argued in 1983, “People expect their governments to provide them with adequate protection, but they also expect them to seek peace and ensure it, and if they are not seen to be doing so, consensus over defense will crumble away.”21Michael Howard, “Reassurance and Deterrence: Western Defense in the 1980s,” Foreign Affairs, Winter 1982-83. In Europe, those twin expectations took shape in the 1960s as nuclear dangers mounted in Europe and deterrence seemed to require an endless arms race. A pivotal moment for the alliance arrived in 1968, when its leadership recognized the separate but related military and political roles of the alliance. Informed by the so-called Harmel Report, the allies concluded that their security required that NATO pursue both defense and dialogue, both deterrence and détente.

The US-ROK and US-Japan alliances face the same challenge today. They cannot meet the new nuclear threat with just defense and deterrence; they need dialogue and détente as well. The Harmel Report would be useful reading in both Seoul and Tokyo today. There is a strong case to be made that further adaptation and strengthening of the regional deterrence architecture discussed above can help to create the conditions for a future political opening. By preserving a robust deterrence architecture against North Korean capabilities, such measures frustrate Kim’s ambitions to draw on those new capabilities in renewed attempts to re-make the regional security order. Prolonged stalemate may improve his willingness to cooperate or even conciliate.

A fourth lesson is that the long-term viability of strategy is the result of continuing efforts to ensure leadership buy-in and public support. NATO’s nuclear strategy is regularly reviewed by heads of state or government. It is communicated to the public through summit communiques and periodic updates to the alliance’s Strategic Concept. The result is a strategy that has been regularly updated but lasted for decades.

The US-ROK and US-Japan alliances could both do more to enhance the long-term viability of their strategies to manage nuclear dangers. Public messaging would be greatly enhanced with the addition of a trilateral component.

A final lesson is that political unity is a strategic asset in its own right. Political divisions within NATO are a target for Russia and others and a potential opening for opportunistic aggression. Alliance unity is usually hard won through patient deliberation and consensus building. Collective resolve is the essential requirement of a credible nuclear deterrence strategy.

As nuclear policy is especially prone to deep disputes, NATO has sought to manage it in ways that sustain and, where possible, deepen consensus.22S. Paul Choi has made an excellent case for such restraint and unity. See Choi, “Deterring North Korea: The Need for Collective Resolve and Alliance Transformation,” 38 North, July 23, 2020, In the late 1950s, NATO’s original nuclear strategy—massive retaliation—began to lose its credibility as US supremacy eroded with the Soviet build-up of long-range missiles. Beginning late in the Eisenhower presidency, more than a dozen proposals were floated by various allies. NATO deliberated for about eight years in order to arrive at the elements of the nuclear posture sketched out above. As deliberated on the needed arrangements for nuclear-needed forces, they considered separately the question of how to improve high-level consultations. Defense ministers took a year to define what became the Nuclear Planning Group.

East Asian observers of this experience in consensus-building should understand that agreement about structures and processes did not encompass agreements about nuclear employment or warfighting concepts. In Rich Kugler’s analysis, the new NATO approach successfully shifted questions of nuclear employment from something “automatic” to “a conditional matter of judgment, not automatic.”23Richard L. Kugler, The Great Strategy Debate: NATO’s Evolution in the 1960s, RAND Note N-3252-FF/RC (Santa Monica, CA: RAND, 1991), p. 45. This was not fully satisfying to those allies who were motivated primarily by a desire to gain control over US choices, especially after the Cuban missile crisis. Nor was it fully satisfying to those in the United States motivated primarily by a desire to ensure control over escalation at a time when other allies were developing counter-city capabilities.

This fifth lesson is especially significant for South Korea and Japan. If unity is a strategic asset, more should be done to narrow the differences between the two. They are not allies but they would have to act in concert in times of crisis and war. This way of thinking also urges caution on those who might argue for bold steps to re-make the regional nuclear order.

Applying These Lessons

A more NATO-like nuclear umbrella makes good sense in Northeast Asia today. But this does not mean simply replicating NATO’s sharing arrangements with the ROK or Japan. Rather, it means putting in place some deliberative processes to determine whether and how best to adapt their strategies and force postures to new requirements.

The US-ROK and US-Japan alliances should begin by updating and elaborating alliance nuclear strategy, broadly defined, and defining an “appropriate mix” of nuclear and non-nuclear means of deterrence. At the same time, defense ministers should review existing military and civilian consultative mechanisms in times of crisis and war, as well as the capabilities and procedures developed in the last decade to counter the North Korean nuclear missile threat. They should then assess their utility for purposes of deterrence and assurance, and, if necessary, recommend improvements.

This should be a careful, thorough, consensus-building process so they should expect to take a year or more to complete it. The process should look not only at format and composition but also the technical systems necessary to consult quickly and securely. The ultimate goal should be to exercise these mechanisms and capabilities in peacetime to prepare for crisis and war.

Additionally, the 2021 review of US defense strategy should fully address the potential for North Korean escalation that have heretofore been left to separate reviews of individual capabilities (e.g., the Nuclear Posture Review).24Brad Roberts, “It’s Time to Jettison Nuclear Posture Reviews,” Bulletin of the Atomic Scientists, January 2020, It should also fully integrate allies in the process of addressing escalation, de-escalation and war termination. Without such innovation, the next National Defense Strategy is unlikely to get very far in addressing the problems identified by the NDS Commission in 2018. If the United States is unable to put its intellectual house in order for modern strategic conflict, its alliances are likely also to fall short in the attempt.

This review process must take due account of the potential reactions of China and Russia to steps the US-ROK and US-Japan alliances might take to adapt and strengthen the regional deterrence architecture in light of North Korea’s continued nuclearization. Both are certain to react strongly, as their leaders believe that America’s hidden strategic ambition is to negate their strategic deterrents. The US and its allies must be able to make a clear and credible case that the adaptations to strategy and posture are necessary and appropriate to the new circumstances and can credibly be defended as not damaging to strategic stability with either major power.


The choice to live with a nuclear-armed and -arming North Korea is inherently risky. But those risks appear to be more acceptable than the risks of the plausible alternative choices. An optimist could argue that a new stable balance of power may yet emerge, if the government in Pyongyang can be persuaded to accept the political status quo, and if both South Korea and Japan can be persuaded to have full confidence in US extended deterrence. This could open the door to accommodation at some future time. A pessimist could argue that the North has a different ambition—to secure a unified peninsula under its leadership—and that the assurance of South Korea and Japan is a wasting asset. This could open the door to crisis and war. A realist must argue that we cannot know. It makes sense to return to the project of adapting and strengthening deterrence, but in a manner that reinforces rather than erodes a diplomatic strategy of engagement.

The United States and its allies in Northeast Asia have been following a business-as-usual approach to adapting and strengthening extended deterrence that has outlived its utility. The rapidly evolving North Korean nuclear threat requires new ways of thinking and the United States and its allies have some important homework to do. They should address the concerns raised by the National Defense Strategy Commission in 2018 by developing a common understanding of the North Korean way of war and of how Pyongyang plans to leverage the threat of nuclear use to achieve strategic victories. With this common operational picture, the three countries—working collaboratively—should develop their own theory of victory and align capability development with that theory.

Finally, the United States should respond positively to rising calls in South Korea and Japan for more “NATO-like” extended nuclear deterrence in Northeast Asia. But they should resist simply importing the NATO model and instead adapt deterrence to the particular strategic context of contemporary Northeast Asia, while building on progress already made.

Change is always difficult and there is a lot of inertia behind maintaining the status quo. But the United States, South Korea and Japan will face increasing danger if they are unable or unwilling to stay ahead of—or at least keep pace with—the nuclear threat from North of the 38th parallel.

About the Author

Dr. Brad Roberts is director of Lawrence Livermore National Laboratory’s Center for Global Security Research. Previously he served as Deputy Assistant Secretary of Defense for Nuclear and Missile Defense Policy (2009-2013). In this role, he served as Policy Director of the Obama administration’s Nuclear Posture Review and Ballistic Missile Defense Review and had lead responsibility for their implementation. From 1995 to 2009, Dr. Roberts was a member of the research staff at the Institute for Defense Analyses in Alexandria, Virginia and an Adjunct Professor at George Washington University. His book, The Case for U.S. Nuclear Weapons in the 21st Century (Stanford University Press) was recently recognized by the American Library Association as one of the outstanding academic titles of 2016.  A member of the Council on Foreign Relations, Roberts has a bachelor’s degree in international relations from Stanford University, a MA. from the London School of Economics and Political Science, and a PhD in international relations from Erasmus University.

Choose Your Subscription Topics
* indicates required
I'm interested in...
38 North: News and Analysis on North Korea
South Asian Voices