NATO’s Nuclear Deterrence: More Important, Yet More Contested

Print Friendly, PDF & Email
Image courtesy of NATO/Flickr. (CC BY-NC-ND 2.0)

This article was originally published by the NATO Defense College in February 2019.

A more competitive international environment, and in particular Russia’s assertive policies, have sparked renewed interest in the concept of nuclear deterrence as part of NATO’s approach to security. This has manifested itself in devoting greater attention to Russia’s nuclear policy and posture as well as to NATO’s own nuclear arrangements, and in a stronger emphasis on nuclear deterrence in public statements. However, this renaissance of nuclear deterrence takes place against the backdrop of new developments that seek to challenge the military rationale and moral legitimacy of that very concept. Put differently, just as nuclear deterrence is again becoming more important, it is also becoming more contested.

Does nuclear deterrence deter?

Nuclear weapons confront mankind with a fundamental paradox: their destructive power makes their use potentially suicidal, but it is precisely the fear of the catastrophic consequences that moderates international relations. Accordingly, nuclear deterrence has been a centrepiece of NATO’s approach to security almost since its creation. Initially viewed as a means to compensate for NATO’s conventional weakness vis-à-vis the Warsaw Pact’s conventional superiority, nuclear weapons were increasingly understood as a political means of war prevention. Hence, while the end of the Cold War saw a dramatic reduction of the number of nuclear weapons, and Western interest in nuclear issues waned, nuclear deterrence remained a major pillar of Western security. For example, NATO Allies consider nuclear deterrence a “core element” of their overall strategy. While they believe the circumstances in which any use of nuclear weapons might have to be contemplated as “extremely remote”, Allies also state that nuclear weapons are “unique”, and that any use of them against NATO “would fundamentally alter the nature of a conflict”.1 In sum, “[as] long as nuclear weapons exist, NATO will remain a nuclear alliance”.2

Deterrence is the threat of using force to discourage an opponent from taking an unwelcome action. This can be achieved through the threat of retaliation (deterrence by punishment) or by denying the opponent’s war aims (deterrence by denial). A major problem of that concept – nuclear or otherwise – is the impossibility to prove that it works. For example, it is impossible to state with certainty that peace is the result of successful deterrence. By contrast, the failure of deterrence (i.e., the outbreak of a conflict) is easy to recognise. Some analysts therefore conclude that nuclear deterrence is overrated, and that the absence of great power war since the dawn of the nuclear age is either the result of other important factors, like recognised spheres of influence, or sheer luck.3 Such arguments have gained considerable traction in the context of calls for the abolition of nuclear weapons: if nuclear weapons do not deter – i.e., if they do not contribute to security, they can safely be eliminated.

However, the argument that the lack of hard empirical evidence invalidates the concept of deterrence is in itself problematic. For one, there are historical cases, albeit few, that support the conclusion that an opponent was actually deterred from taking a certain course of action. Above all, however, the assumption that the absence of nuclear war for over 70 years has something to do with the risks associated with the use of nuclear weapons remains simply more plausible than any alternative explanation.4 Many critics of nuclear deterrence – inadvertently – admit as much: by warning about the danger of nuclear weapons they underscore the point made by deterrence advocates: nuclear weapons make certain types of aggression so risky that nations will try to avoid them. Former US Secretary of Defense James Schlesinger described this existential and permanent notion of deterrence by stating that nuclear weapons were being “used” every day.5

Are there limits to nuclear deterrence?

There are many limits to nuclear deterrence.6 The most obvious is the lack of credibility of nuclear threats as a response to smaller, non-nuclear, non-existential forms of aggression. Deterring these threats requires different means, for example conventional forces or offensive cyber capabilities. Put differently, nuclear deterrence is an essential tool for inducing restraint in international relations, yet in a crisis it will only work at the “high end”. Hence, the common criticism that nuclear deterrence is worthless because it does not work in “smaller” scenarios is beside the point.

One of the most common mistakes in thinking about deterrence is to focus only on nuclear hardware and ignore the interests of the parties. Yet a large nuclear arsenal does not automatically translate into a credible deterrent. Since a state will only take nuclear risks in defence of existential interests, an opponent may still resort to force if he concludes that the issue at stake is simply not existential to the defender. The Russia-Ukraine crisis offers a strong reminder in this regard. Russia correctly calculated that a limited conventional and hybrid attack on a non-NATO country was not likely to trigger a NATO military response, let alone a nuclear one. Thus, if critics argue that this crisis had demonstrated the irrelevance of nuclear deterrence, they are erecting yet another smokescreen: there simply was no deterrence relationship, neither nuclear nor otherwise. The crisis was determined by asymmetric interests, not by military balances.

Another limit derives from moral considerations. Irrespective of whether a “nuclear taboo” or a “tradition of non-use” actually exists, history shows that after the atom bombs were dropped on Hiroshima and Nagasaki in 1945 no nuclear power has used such weapons in a conflict with a non-nuclear adversary, even when the risk of nuclear retaliation was low. Famously, in 1951 General MacArthur was relieved from his post when he suggested using nuclear weapons in the Korean War. This logic has been codified in so-called “negative security assurances” – i.e., statements by nuclear weapon states that they will not use nuclear weapons against non-nuclear weapon states that are not allied with a nuclear weapon state.

Yet another limit to nuclear deterrence is human psychology. A stable deterrence regime requires all actors to adhere to a “rational” cost-benefit calculus. Thus, nuclear deterrence cannot work against actors that are “irrational” to begin with, for example, suicidal fanatics. However, the most likely scenario in which rationality could disappear is, ironically, defensive: since humans tend to fear suffering losses more than they value gains, the fear of losing something valuable will make leaders take far greater risks than the opportunity of changing the political or military status quo in their favour. Hence, as much as one would want to have the upper hand in a crisis, one should still avoid pushing a nuclear adversary into a corner.

Does extended deterrence really work?

Even in the nuclear domain credible deterrence depends on the interests that one seeks to protect. If a nation’s existence is at stake, the use of nuclear weapons in self-defence is credible. Accordingly, deterrence between nuclear weapon states is considered to be fairly “stable”. By contrast, extending one’s national nuclear deterrence to allies is much more complicated. Allies of nuclear powers constantly need to be reassured that the protector considers their security a truly vital interest. Or, as British Defence Secretary Denis Healey famously put it in the 1960s, one only needed five percent credibility to deter the Russians, but 95 percent to reassure the Europeans. Despite this “Healey Theorem”, however, extended nuclear deterrence has become a central pillar of the international order. Today, over 30 countries in NATO and the Asia-Pacific region are under the US “nuclear umbrella”.

Critics of this concept of “extended deterrence” often assert that the United States would never risk its own destruction in order to protect its allies in Europe or Asia. However, the credibility dilemmas of extended deterrence appear to be more theoretical than real: by explicitly extending its nuclear (and conventional) deterrence to other countries, Washington sends a powerful signal that it views the security of these allies to be a fundamental national security interest. Moreover, the “nuclear umbrella” also relieves allies from the need to develop nuclear weapons of their own. While critics sometimes question the causality between extended deterrence and non-proliferation, history shows otherwise: several US allies who felt abandoned by Washington had started nuclear programmes, only to halt them once relations with the US had recovered.7 Hence, irrespective of the credibility problems of extended deterrence, if the US were to end this policy, the result might be a major new wave of nuclear proliferation.

Does nuclear sharing still make sense?

Among the favourite targets of anti-nuclear activists are NATO’s so-called “nuclear sharing” arrangements. As part of these arrangements, several European allies host dual-capable aircraft (DCA) and US nuclear gravity bombs, while the launch authority remains with the US. This concept was developed in the 1960s, largely as a means to reassure European allies about the US nuclear commitment, thus paving the way for their joining of the Nuclear Non-Proliferation Treaty (NPT) as non-nuclear weapon states. After the end of the Cold War, the military value of these arrangements has sometimes been questioned, yet their political benefits continue to outweigh any military limitations. Above all, the DCA option allows the non-nuclear allies to participate in the planning and implementation of the nuclear mission, either by hosting aircraft and nuclear weapons or by suppressing hostile air defences. These arrangements underscore the fact that, by sharing nuclear risks and burdens, the allies consider their solidarity to extend to the nuclear domain.

Contrary to some critics’ assertions, nuclear sharing neither violates the NPT nor is it an obstacle to non-proliferation. On the contrary, developments in East Asia vindicate the rationale of NATO’s arrangements. In South Korea, where US nuclear weapons were withdrawn in the early 1990s, and to a lesser extent in Japan, there is now a lively discussion on whether to ask for some US nuclear weapons to be re-deployed to the region. While the bilateral nature of US security relations with South Korea and Japan cannot be compared with the arrangements in NATO, the Asian debate suggests that Europe is better off maintaining the status quo. This is all the more true as the military value of the DCA option is now being restored by a new generation of (stealthy) aircraft. In short, the logic of Allies managing the nuclear reality together remains compelling – it is the essence of NATO’s self-characterisation as a “nuclear alliance”.

Do we need nuclear warfighting capabilities?

One of the many contradictions of nuclear deterrence is that it relies on the fear of escalation, yet should deterrence fail and a conflict break out, one will want to prevent such escalation from occurring. This is why some nuclear weapon states seek to acquire weapons that allow for limited nuclear options. Critics perceive this as a dangerous “lowering of the nuclear threshold” by making nuclear weapons more “usable”. However, weapons that are manifestly “unusable” offer little deterrence value. This does not imply that possessing a nuclear arsenal that includes smaller yield weapons would make its owner more “trigger-happy”, as some critics insinuate. It simply means that nuclear weapon states, for the sake of maintaining their credibility in the eyes of adversaries as well as of allies, will seek options that do not result in obvious “self-deterrence”.

It is impossible to know how much a specific new nuclear capability will contribute to the credibility of deterrence. However, if an opponent’s nuclear arsenal grows, including by fielding systems that are in violation of arms control agreements, and if that opponent has a nuclear strategy that indicates the limited use of nuclear weapons in order to achieve war termination on his own terms, the defender will at some point have to evaluate whether his own nuclear arsenal is still sending a credible deterrent message. Not reacting could mislead the opponent into believing that he has gained a useable advantage. However, such calculations must be balanced with the need not to unsettle the public. As NATO’s history has shown, the deployment of certain weapons systems intended for restoring a credible deterrent can backfire, resulting in massive public protests. Since a deterrence message always has two addressees – one’s opponent and one’s own population – the challenge for NATO is to impress the former without frightening the latter.

What does the Nuclear Ban Treaty mean for NATO?

The Treaty on the Prohibition of Nuclear Weapons (TPNW, also called Nuclear Ban Treaty), which includes a comprehensive set of prohibitions on participating in any nuclear weapon activities, is now becoming a political reality. The Treaty, which may enter into force in 2019, seeks to stigmatise nuclear weapons by declaring them illegal. This, its proponents argue, will make it much harder for the nuclear weapon states to cling to these weapons as a means of deterrence. The TPNW’s supporters admit that it will not in itself lead to nuclear abolition, yet it could create a dynamic that would pave the way for the elimination of nuclear weapons. In essence, the Treaty is meant to be a short-cut to nuclear abolition, since its supporters argue that progress in the traditional framework of the NPT has been much too slow.

No nuclear weapon states (and none of their allies) have signed the TPNW, hence they cannot be bound by it. However, the Treaty could deepen the rift between nuclear and non-nuclear weapon states, as well as complicate NATO’s relations with some partner countries. Stigmatising nuclear weapons as illegal could also pull the rug from under the NPT, which remains the only near-universal framework for managing nuclear possession and non-possession. What is more, despite the TPNW’s universal vocation, some of its wording appears to specifically target NATO and its nuclear policies. Leading supporters of the Treaty have confirmed this prejudice against Western nuclear arrangements.8 Moreover, since the TPNW will only have an impact in countries with strong civil societies, authoritarian regimes such as North Korea or Iran or “managed democracies” such as Russia are not likely to be affected. Hence, the TPNW remains a problematic endeavour, all the more so since its very premise – the dangers and/or irrelevance of nuclear deterrence – is questionable. If abolishing nuclear deterrence would make major (conventional) war more likely, it can hardly be regarded as the morally superior alternative.9

The future of nuclear deterrence

The return of nuclear deterrence is inevitable. In light of what appears like an increasing militarisation of international relations, an essentially Western monologue about the promises of global nuclear abolition looks increasingly out of touch with reality. NATO’s reaffirmation of nuclear deterrence does not suggest that Allies would have to renege on their commitment to create the conditions for a nuclear-free world. For the foreseeable future, however, these conditions will not exist. The same holds true for initiatives to stigmatise nuclear weapons. In the current security environment, they amount to little more than doubtful attempts to delegitimise the defence policies of Western democracies.

That said, in defending nuclear deterrence, one must also take great care not to trivialise its risks, such as, for example, cyber-attacks on nuclear infrastructure. A commitment to nuclear deterrence thus needs to entail a commitment to nuclear safety and security. One must also realise that with more countries obtaining nuclear weapons, deterrence will become far more difficult to manage than the Cold War bipolar nuclear standoff. Hence, nuclear deterrence must not be regarded as an eternal condition, but as a “time-buying strategy”: it should provide the time needed to overcome the political antagonisms that make nuclear deterrence necessary in the first place.


1 Warsaw Summit Declaration, 2016, para. 54.

2 NATO Strategic Concept, 2010, para. 17.

3 For different views on the importance of nuclear deterrence see the issue “Do nuclear weapons matter?”, Foreign Affairs, Vol. 96, No. 6, Nov.-Dec. 2018.

4 See B. Tertrais, “In defense of deterrence: the relevance, morality and cost-effectiveness of nuclear weapons”, IFRI Proliferation Paper No. 39, IFRI, Paris, Fall 2011.

5 See M. Kirkpatrick, “Why we don’t want a nuclear-free world”, interview with James R. Schlesinger, The Wall Street Journal, 13 July 2009.

6 See M. Rühle, “Deterrence: what it can (and cannot) do”, NATO Review, April 2015.

7 See R. K. C. Hersman and R. Peters, “Nuclear u-turns: learning from South Korean and Taiwanese rollback”, Nonproliferation Review, Vol. 13, No. 3, November 2006, pp. 539-553.

8 See the contributions in S. Shetty and D. Raynova (eds.), “Breakthrough or breakpoint? Global perspectives on the Nuclear Ban Treaty”, Global Security Special Report, European Leadership Network, December 2017.

9 See B. Roberts, “Ban the bomb? Or bomb the ban? Next steps on the Ban Treaty”, Global Security Policy Brief, European Leadership Network, March 2018.

About the Author

Michael Rühle is Head of the, Energy Security Section, Emerging Security Challenges Division, NATO.

For more information on issues and events that shape our world, please visit the CSS website.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.