Lowy Institute

Debate: Is extended nuclear deterrence dead?

Learn more about Interpreter debates  |  View debate archive.

Reverse order

Order: earliest first

These are confusing times in nuclear strategy.

The Obama Administration is promoting the vision of a world without nuclear weapons. At the same time, power politics and coercion are making a comeback, particularly in Asia, where repeated instances of Chinese assertiveness and the use of armed force by North Korea are unsettling US allies including South Korea, Japan and Australia. Half a world away, NATO has struggled to reconcile nuclear disarmament imperatives with concerns about Russia in its revised strategic concept. In South Asia and elsewhere, fears of nuclear terrorism are rising. And Iran's atomic ambitions could rewrite deterrence calculations across the Middle East.

All of this points to a vital question, the answer to which will be critical to international stability in the years ahead: is the age of extended nuclear deterrence (END) coming to an end? For decades, the US has made the seemingly-credible threat that it would use nuclear weapons to protect its allies against large-scale aggression — the so-called 'nuclear umbrella'. But how viable is such a strategy in a changing nuclear order and an altered strategic environment? And are there feasible, palatable alternatives?

Here at The Interpreter, we think it is time to foster a dynamic and truly global debate on this issue. To launch the exchange, we have invited contributions from four of the world's leading experts on nuclear arms control and strategy: George Perkovich of the Carnegie Endowment for International Peace, Bruno Tertrais of the Fondation pour la recherche stratégique, Shen Dingli of Fudan University in Shanghai, and the Lowy Institute's own Hugh White.

Read More

In the days and weeks ahead, their initial posts will be followed by other solicited contributions from prominent security thinkers and practitioners. We will also open up to the debate to readers and give the original contributors opportunity to reply and expand on their arguments. If you have your own views on the subject, please send them to Sam Roggeveen, editor of The Interpreter.

As the debate progresses, we will be interested in exploring not only the core analytical question of whether END is dead or alive, but also the policy options for the nations concerned, including those US allies who have sometimes entertained nuclear options of their own.

It should prove a fascinating, important and at times confronting conversation. 

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user Paul [W] Campbell.


George Perkovich is Director of the Nuclear Policy Program at the Carnegie Endowment for International Peace.

As the unipolar era ends, the pro-Western imagination somehow remembers the Cold War as halcyon days for extended deterrence. In fact, extended deterrence was always problematic. It is not much more so today.

Extended deterrence often is conflated with extended nuclear deterrence. The two are not the same. I assume we are here debating the life or demise of extended nuclear deterrence. In that case, the key question is, 'what are we expecting US nuclear security guarantees to deter?'

US allies naturally wish that American policies and weapons will deter a wide range of possible threats. In Eastern Europe today, this could include threats of Russian military intervention in territorial disputes (a la Georgia), or Russian energy blackmail, or cyber-attacks as Estonia experienced in 2007. 

In East Asia, American allies and friends worry over China's growing aggregate power and the possibility of being pushed around over economic issues, control over natural resources and disputed islands, policy toward North Korea, and the security of cyber networks. South Korea and Japan understandably also fear North Korean aggression of various sorts.

In evaluating deterrence of these threats (by whatever means) we must not ask the US to do for others what it has not been able to do for itself. The US was unable to deter Afghanistan from enabling al Qaeda to undertake the 9/11 attacks. It did not deter Saddam Hussein from invading Kuwait in 1990, nor coerce Saddam to abide by UN resolutions in the run up to the 2003 war. India's nuclear weapons did not deter Pakistan from the Kargil intervention in 1999. 

In short, nuclear-armed states have been unable to deter a number of very unwanted challenges to their security and those of their allies. What they have been able to deter, generally, is large-scale aggression that would threaten their own existence or those of their protectorates. In other words, nuclear weapons deter only those threats against which it is credible to use nuclear weapons.

Read More

In cases where the potential aggressor wields nuclear weapons, this means that the threat must be sufficiently grave for the deterrer to risk escalation to nuclear war, with the potential result of mutual suicide. It is not credible to take such mortal risks unless the alternative would also threaten one's existence or, in the case of allies, one's identity and viability as a great power. 

Therefore, extended nuclear deterrence should only be relevant against threats of such magnitude that the possible use of nuclear weapons to deter or defeat them would be proportionate. Against lesser threats, other means of deterrence must be relied upon. In East Asia today, this means, for example, that nuclear weapons should not be expected to deter potential Chinese efforts to occupy disputed islands in the South and East China seas, or to deter most probable forms of cyber disruption or political-economic bullying. Other means, including strengthened combined-allied conventional forces and operations, should be enhanced for these purposes.

The DPRK is a particularly troubling case. What scale and type of aggression by Pyongyang would credibly justify and call for US nuclear retaliation (including pre-emptively)? The US can destroy North Korea's military and economy, albeit at enormous costs not only to North Korea but also to South Korea. But the underlying challenge then (and now) is to replace the government in North Korea by one that is civilized to its own people and others. Nuclear weapons offer little to this end.

In the nuclear age, revisionist states, insurgents, and terrorists have learned to utilise asymmetric 'weapons' to fight more powerful nuclear-armed states. Nuclear weapons will remain for the foreseeable future in the background to deter existential-scale threats, but other means must occupy the foreground. Rather than wishing for nuclear weapons to do the job, or hand-wringing that they cannot, we should concentrate on developing credible strategies and capabilities to deter insidious new forms of aggression and subversion.

There is no material reason to think that US interest in moving gradually and multilaterally toward a world in which no one possesses nuclear weapons should undermine extended nuclear deterrence. The US is emphatically clear that, as long as anyone has nuclear weapons, and the nonproliferation regime is not extremely robust, the US will retain nuclear weapons and a state-of-the-art nuclear infrastructure. 

US interest in nuclear disarmament stems from the perception that a world without nuclear weapons would give it a greater advantage against others that might threaten it or its allies.  The others – particularly China, Russia and North Korea – recognize this! They see the Obama agenda as a means of strengthening the US advantage. Hence they (and Pakistan) are likely to impede nuclear disarmament. How does this weaken extended nuclear deterrence?           

It is easy to raise questions about the viability of extended nuclear deterrence in a vacuum, forgetting that nuclear deterrence comes with the real risk that the weapons will be used. Serious policy-making should focus on how to redress insecurities and deter and prevent war without the catastrophic risks of nuclear proliferation and use. The US is committed to leading in this direction. It would welcome similarly serious efforts by other influential states to address these underlying political-security challenges.


Photo courtesy of Wikipedia.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.


No country will ever run the risk of a nuclear attack for the sake of protecting one of its allies, General Charles de Gaulle once said, thus justifying the building of an independent French deterrent.

The credibility of the US 'nuclear umbrella' was the focus of considerable debate during the Cold War, and there is no reason why this debate will not continue. But extended Western deterrence is alive and well. Not only it has survived the end of the Cold War, but its scope has even been expanded. In Europe, NATO has almost doubled in membership in the past 20 years, and the new members are keen to emphasize how much the US umbrella matters to them. (France itself now declares that, given growing European integration, its deterrent force also protects its neighbors.)

In Asia, North Korean provocations and China's military modernisation have led to a strong reaffirmation of US protection, and even to the creation of a mechanism for US-South Korean nuclear consultations. In the Gulf, the three Western nuclear powers have made new security commitments since 1990, either informally (the case of the US) or through defence agreements (in the case of France and the UK). Recently, fears of an Iranian bomb have also led the US to hint at the extension of a 'defense umbrella' over the Arabian Peninsula.

During the 2010 Nuclear Posture Review (NPR) process, the US administration was particularly careful to listen to its key allies. The NPR Report includes a strong reaffirmation of its nuclear umbrella, and discards the 'no-first-use' of nuclear weapons largely for extended deterrence reasons. Washington has also made it clear that the reason why it wants to maintain parity with Russia is that its allies would be wary of a perceived US 'nuclear inferiority'.

Read More

The usefulness of extended deterrence for non-proliferation can hardly be debated. A large body of historical evidence shows that it was the key to the renunciation of nuclear weapons by many Western countries. And if Iran was to cross the nuclear threshold, a strong reaffirmation of US guarantees will be helpful to prevent countries such as Turkey and Saudi Arabia looking for alternative options.

Extended deterrence is, however, changing. Due to technological progress, force reductions, as well as a growing allergy in many countries to nuclear weapons, it increasingly relies on non-nuclear assets such as missile defense and long-range precision strikes. Conversely, due to the evolution of the strategic landscape, it is  geared towards preventing nuclear and other WMD threats more than conventional invasion.

The fact that extended deterrence will rely more on non-nuclear means will not necessarily make it less credible. As long as a state possesses nuclear weapons, its adversary has to take into account the risk, however remote, that he could ultimately face nuclear retaliation. And an emphasis on non-nuclear capabilities may in fact increase its effectiveness – relying too much on nuclear weapons can lead adversaries and allies to doubt the protector's resolve, and he himself may be 'self-deterred'.

Likewise, the fact that extended deterrence relies less on in-theatre means does not necessarily affect its credibility; deterrence is fundamentally a psychological process.

However, this also means that extended deterrence will have to be even more carefully nurtured than in the past. This is a complex exercise since it has to be credible in the eyes of three parties: the protector, the protected, and the adversary. There have been many examples of failures of extended deterrence in the past – the most egregious of which was the North Korean attack in 1950. But nowadays, the perception that Western countries are 'weak' is widespread. This can affect the value of deterrence both in the eyes of allies and of adversaries.

This calls for careful calibration of statements and declarations regarding security commitments, adapting them to various regional contexts and domestic sensitivities. This also calls for caution when decisions to alter a military posture are made for financial, political or ideological reasons – both allies and adversaries are watching.

A topical example is Europe, where many consider that nuclear weapons could be replaced with missile defense, and argue that what matters is a physical US military presence on the continent. But other believe that the withdrawal of US nuclear weapons could make the NATO deterrent less credible in the eyes of Russia or Iran, or encourage a Turkish nuclear program. (The withdrawal of US nuclear weapons from North-East Asia in 1992, largely conceived as a goodwill gesture towards Pyongyang, did nothing to prevent North Korea from going nuclear.)

Finally, Western countries should remember that the way they manage their commitments in one part of the globe can have a much broader impact. End-game decisions regarding operations in Iraq and Afghanistan, for instance, will be watched very carefully by both friends and foes.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user ♥Unlimited.


Shen Dingli is Professor and Executive Dean of the Institute of International Studies, Fudan University.

The answer to the question 'Is the age of extended deterrence over?' cannot be a clear one. It is neither a firm 'yes' nor categorical 'nay'.

To start with, nuclear deterrence has both succeeded and failed. On the successful part, the US kept the secret of Soviet participation in the Korean War from the public to contain the likelihood of crisis escalation that may have led to direct conflict with a nuclear Soviet Union. Also during the Korea War, China was deterred by the US nuclear arsenal, which partly explained China's acceptance of the armistice.

On the unsuccessful part, America, despite its possession of nuclear weapons, did not deter North Korea's attack in 1950 or that of the Viet Cong in the 1960s-70s. China has even proclaimed a no-first-use policy, refusing to deter the US from selling weapons to Taiwan, though Beijing deems Taiwan a core national interest.

America's extended nuclear deterrence has also had a mixed record. The credibility of US extended deterrence for its NATO allies has not been challenged, and it assured nuclear nonproliferation among them. Nevertheless, America's virtual ally, Israel, has gone nuclear despite the US security commitment. During the 1970s, South Korea and Taiwan clandestinely launched their nuclear weapons programs for fear of US withdrawal from East Asia.

Read More

Recent trends seem to indicate the acceleration of the irrelevance of extended nuclear deterrence. If Pyongyang might not have been completely responsible for sinking the Cheonan, at least it didn't mind US extended deterrence for South Korea when it launched its artillery barrage against Yeonpyeong. The US didn't dispatch USS George Washington to the West Sea/Yellow Sea in the wake of Cheonan sinking. Even if the US did so after the artillery exchange in November 2010, it still didn't retaliate militarily against North Korea.

America mistakenly entered Iraq in 2003 under the pretext of Saddam's WMD build-up. However, Iran's persistent nuclear quest, regardless of various IAEA and UNSC warnings, has ridiculed any international pressures, including US deterrence. Israel is increasingly under pressure to launch a pre-emptive strike against Tehran so as not to burden US extended nuclear deterrence.

Beyond the seeming decline of extended nuclear deterrence, the aforementioned trend is also indicative of America's caution in wielding the nuclear option. What is at stake is not deterrence or extended deterrence, but security assurance per se. The US has actually not loosened its security commitment to its allies, but in order to maintain the credibility of its deterrent, America is now less willing to coerce its non-nuclear rivals while it has increasingly more non-nuclear tools in its policy kit. This was manifested in the Obama Administration's Nuclear Posture Review last year.

Both North Korea and Iran have understood this, to their benefit. Iran understands that its violation of various UNSC resolutions would at most incur a non-nuclear US response. Indeed, UN Secretary General Ban Ki-moon is mediating a nuclear swap arrangement to lessen international concern, which could legitimise Iran's uranium enrichment. North Korea has also ventured into crises short of inviting US military strikes, and it has presented its own conventional deterrence plus a looming nuclear deterrent. Deterrence is no longer a US gadget.

Mutual nuclear deterrence among major powers is indeed less relevant nowadays, in a globalising age. At a co-dependent time, there are more incentives for inter-state compromise and reconciliation, while contingencies which require a nuclear showdown are implausible — hence the decline of nuclear deterrence among major powers, and extended nuclear deterrence backed by various non-nuclear options.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user US Army Korea - IMCOM.


It is best to start thinking about how extended nuclear deterrence (END) might work in future by looking back at its Cold War origins. It is, of course, an American concept. America deliberately promoted expectations that it would respond with nuclear weapons to a Soviet attack on its allies in Europe and Asia. This policy had two aims: to deter the Soviets from attacking US allies, and to keep the allies loyal by reducing their incentive to build their own nuclear forces.

For END to work, America needed both the evident nuclear capability to strike the Soviets hard enough, and the evident willingness to use that capability if its allies were attacked. The capability bit was never in doubt, but it proved harder to persuade both Moscow and America's allies that it would actually use its nuclear forces to defend others. 

The proportionality of a nuclear response to a conventional attack was one problem, but the real question concerned costs to the US once the Soviets had the ability to strike back. Would Washington risk nuclear retaliation against the US itself to defend an ally an ocean away?

Much of US nuclear strategy in the Cold War was devoted to persuading both friends and foes that it would. Ultimately, the US succeeded because it convinced others that it saw the loss of a European or Asian ally as posing a direct threat to the US, because they feared that would lead to Soviet domination of Eurasia, which would make Moscow strong enough to overmatch and dictate to the US.

What does this tell us about END over the next few decades? The capability element seems to me pretty clear. The US can easily maintain nuclear forces able to devastate any adversary, and – speeches in Prague notwithstanding — I think there is little doubt that it will do so. But America's ability credibly to threaten nuclear attack to defend other countries is much less assured. It depends on whether Washington can persuade others – adversaries and allies alike – that it would be willing to go ahead and launch a nuclear attack if its bluff was called. 

Read More

There is no reason to doubt American willingness to use nuclear forces against powers that cannot retaliate directly against the US, at least in relation to nuclear attacks when proportionality is not an issue. It is, for example, entirely credible that Washington would launch a nuclear attack against Pyongyang or Tehran if either of them used nuclear forces against a US ally, because neither have (for the time being) an assured capacity for nuclear retaliation against the US. 

But these are not the tough cases. The real question about the future of END is whether the US can credibly threaten nuclear attack against an adversary that could retaliate against the US – which for present purposes means Russia or China. 

Take Europe first. If Moscow launches a conventional invasion of a NATO ally – say Latvia or Estonia – could America credibly threaten nuclear attack to force Russia to withdraw? Could it persuade Moscow that the independence of the Baltic States is so important to America that it would accept nuclear attack on the US to preserve it? That seems to me very doubtful, because Russia today does not have the potential to dominate Eurasia the way it did in the Cold War. The stakes for America are simply not as high – and not high enough.

What of China? As General Xiong once asked, is Taiwan's independence worth Los Angeles? More broadly, is anything on the western side of the Pacific important enough to Washington to convince Beijing that America would accept nuclear attack on its homeland to defend it? Even Tokyo?

This is the core question for the future of END: what is so important to Americans today that they are willing to suffer nuclear attack to defend it? And can the US persuade friends and allies that it is willing to make this sacrifice? Many would suggest the maintenance of a US-led world order is important enough, but I doubt that, especially when the non-military challenges to that order are so strong anyway.

I think the costs of nuclear attack are so high that nothing except the defence of America's own territory and independence would justify them. In the Cold War, when it seemed one power could dominate Eurasia, America's own security appeared to be at risk from a Eurasian hegemon. Today, with at least two and maybe four major Eurasian powers, that risk seems very remote.

So against any power capable of delivering nuclear weapons onto American soil, END is an anachronism. Neither America not its allies yet accept this. The sooner they do the better for everyone.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user -Alina-.


Dr Stephan Frühling and Dr Benjamin Schreer are Lecturer and Senior Lecturer in the Strategic and Defence Studies Centre, ANU.

Previous entries in The Interpreter's debate on the future of extended nuclear deterrence (END) make a valuable contribution to the discussion about this important aspect of Asian security order. Both Hugh White and Shen Dingli see US willingness and ability to provide assurance to its allies as being in decline. 

Hugh's deliberate mirror-imaging follows an established approach to the theory on deterrence. There is a long tradition of analysts using it to argue for their policy prescriptions, not least Nobel-prize winning Thomas Schelling. 

But as Bruno Tertrais points out, deterrence is a psychological phenomenon. Therefore, judgements about its viability must ultimately rest on empirical evidence of real-life attitudes and policies. And to us, the evidence points to a more positive view of the viability of US END in East Asia.

We must be careful not to ascribe to END a greater remit than it really has. George Perkovich remarks that END is only suited to dealing with the possibility of a major, existential threat. Skirmishes on the Korean Peninsula are nothing new, and no argument that END, or even extended deterrence in general, is failing. 

Read More

It is also simply not true that NATO today is threatening nuclear retaliation for a Russian attack in the Baltics. Ever since the adoption of Flexible Response in 1968, NATO has deliberately developed the ability to respond conventionally against a major but limited Russian conventional attack. The 'threat that leaves something to chance' is still there, but NATO does not rely on it. 

And END has never been, and cannot be, a fail-proof panacea even against major war. Nevertheless, even a small reduction in the likelihood of major war would make it a valuable part of global and Asian security order.

In hindsight, the Cold War can easily seem more straightforward than it really was. Care is needed when using it as a quarry for arguments about future policies in Asia. The sources of US commitment to the defence of Europe, for example, were much more varied than is suggested by a purely realist perspective: a good case can be made that, from the late 1950s, outright conquest of Western Europe would have been a poisoned chalice for the Soviet Union. The willingness of the US to defend Europe even at the risk of nuclear attack on America rested not only on national interest, but also on common values, shared history and personal relationships. And extended deterrence was much stronger, and more credible, for it.

As long as the US remains a nuclear power, END will be an inseparable, even if implicit, part of its alliance relationships. And we do not see any US policy, or that of a major allies or potential adversary, indicating that the strength of the decades-old US alliance relationships in East Asia is fundamentally weakened. 

That is not to say that extended deterrence relationships do not evolve. But at a time of changing great power relativities, an increased attention to END in official circles is a sign of vitality, rather than of decline. As long as there was no prospect of a major threat, there was also no need for practical and visible nuclear cooperation. To us, the increasingly formalised dialogue about nuclear matters between the US and Japan and South Korea thus is a good sign of the continuing relevance of END in Asia. 

The debate about extended deterrence is as old as nuclear weapons themselves. As in the past, its intellectual attraction arises in large part from the tension between arguments about how the world should be in theory, and how it is in practice. At its core, END remains a leap of faith. But as long as all those decision-makers who matter are prepared to make that leap, END remains viable. And we see no empirical reason to doubt that this will not, for the foreseeable future, be the case for American END in East Asia.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.


Peter Layton writes:

Hugh White's realist analysis readdresses many worries of the Cold war, but maybe deterrence is more in the mind than in the hard materialism of realism.

While Hugh notes that the US 'succeeded' in convincing others of extended nuclear deterrence (END), it really did not, in the sense that the Soviets saw no reason to call its bluff. END may be dead, but in this second nuclear age I suggest that again no one will wish to call another nuclear power's bluff in this matter. END lives because no one truly knows if it is really dead — or is willing to risk their life for it.

Of course, before END was tested, as postulated in the MAD games of Albert Wholsetter and Herman Kahn, it is likely that in the initial state of a nuclear showdown, the nuclear powers involved would attempt to make their homelands a sanctuary. Nuclear weapons would be used only on the allied nations of the adversary (or at sea or in space) in an attempt to end the conflict before the homeland was attacked. This is what really worried the Europeans, that the USSR and the US would fight the second nuclear war on European soil rather than their own. END might indeed be dead, but from an allied viewpoint the debate was irrelevant because they could well end up nuclear targets anyway.

Read More

This prospect worried De Gaulle and, like Hugh, he publicly doubted END and so France developed — admittedly for many other reasons as well — a nuclear force and a nuclear technology base. But other European nations were both more sanguine and the receipts of American generosity. The Eisenhower Administration embraced nuclear sharing and several European countries became able to deliver loaned US nuclear weapons, albeit with a dual key.

This was a powerful rebuttal to those who doubted END, and sowed more seeds of uncertainty. If Bonn was vaporized would the Americans allow the West Germans to head east? And not just that; could America in the confusion stop them even if it wanted to? Would you risk Kiev on it?

At the end, END lives because nuclear weapons are really, really scary and so even little risks loom very large in decision-makers' minds, encouraging them towards caution. But the Cold War was very dangerous; the American military wanted to invade Cuba in 1962, not realising that nuclear weapons were already in Cuba awaiting them, while the Soviets got confused over Able Archer 83 and considered a pre-emptive attack in Reagan's first term.

Today though in the nuclear stakes, its hard to get past Pakistan, a Chinese and American ally that keeps building bombs while it gratefully accepts foreign aid. The New York Times reported last week that Pakistan is now the fifth largest nuclear power, surpassing the UK and will soon pass France, the fourth. Numbers are not yields, and this measure may not be as important as might be thought, but Pakistan's nuclear ambitions will keep many thinking and worrying about where this all END(s).

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.


Shen Dingli and Hugh White make valuable points in their contributions to the debate on extended deterrence opened by The Interpreter. However, they also construct strawmen, which limits the value of their arguments.

Shen Dingli mentions the 1950 invasion of South Korea; but at that time, there was no US extended deterrent to South Korea. The bilateral treaty was signed in 1953, after the war, in order to prevent a resumption of hostilities. His other example is the Vietcong war against Saigon; but likewise, at that time South Vietnam was not covered by an explicit defense commitment which promised retaliation against the North in case of aggression. And the Vietnam war was very different from the Korean one: it was more a slow-motion escalation than a full-scale state attack.

Finally, the case of Iran's 'persistent nuclear quest' is irrelevant to extended deterrence. It is true that 'Iran's violations would at most incur a non-nuclear US response', but so what? That has nothing to do with extended deterrence. For sure, one can argue that Washington and the other four permanent members of the UN Security Council have so far failed to deter Iran from continuing its nuclear march, and that Tehran's nuclearisation would have far-reaching consequences for US extended deterrence in the region. But that is a completely different debate. 

Shen Dingli's contention that extended deterrence did not prevent the bombing of Yeonpyeong is much more relevant and interesting.

Read More

It could indeed be argued that Pyongyang was testing the limits of US protection to Seoul. However, another intriguing possibility should be raised: that North Korea was very careful to strike a 'disputed' area, one which was – at least from its point of view – not part of the territory covered by the US umbrella. In so doing, it may have acted like Egypt and Syria did in 1973: facing a nuclear-armed adversary, they were very careful not to attack Israel on its 1948 borders.

Hugh White raises an extremely important question: he wonders if Washington would threaten nuclear war for the sake of the Baltic states, since the risk of 'losing Europe' would be limited. In other words, the stakes would be much lower than during the Cold War. But, as in most discussions about deterrence, one has to consider the problem the other way round: would Moscow take the risk of invading the Baltic states, given that it would face three nuclear-armed adversaries and risk a military escalation that could, ultimately, end up in a nuclear exchange?

My answer is 'no'. It may be precisely because Georgia was not under NATO or US protection that Russia considered it could take a chance in 2008. (Incidentally, here again we see an intervention on what was arguably a disputed area: the Abkhazian and South Ossetian territories. Western reaction could very well be what stopped Moscow from going further.) 

White also quotes with appreciation the reported 1996 statement by a Chinese official, according to which the US cares more about Los Angeles than it does about Taiwan. However, during a crisis in the Taiwan straits, the 'deterrence dialogue' would not stop there. The question is: what would Washington tell Beijing if China threatened to escalate the conflict to the US homeland? My guess: the US president would say 'try me'.

In other words, the credibility of the US extended deterrence to Taiwan depends not on the relative values of Taiwan and Los Angeles for Washington, but on the respective beliefs of Chinese and American officials about the sequence of a US-China nuclear war. A classic deterrence issue.

Shen Dingli makes a different but related point when he states that globalisation and interdependence have made nuclear deterrence among major powers less relevant than it used to be. That may be true, and there is something like a 'financial balance of terror' between Washington and Beijing. But I would caution against definitive conclusions: the First World War erupted at a time of (then) unprecedented economic interdependence between the major European powers of the time. This did not trump political passions and the strength of ideologies.

UPDATE: I described in an earlier post the 1950 North Korean attack as a 'failure of extended deterrence'. I should clarify what I meant. There was no formal extended deterrence to South Korea at that time. But by stating that the Korean peninsula was not part of the defensive perimeter of the US, Washington signaled explicitly to North Korea and its allies the absence of such a deterrent and gave a green light to the invasion.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user The US Army.


Jeffrey Lewis is Director of the East Asia Nonproliferation Program at the James Martin Center for Nonproliferation Studies, Monterey Institute for International Studies, and editor of Arms Control Wonk.

The place to start any discussion about the future of extended deterrence – which is essentially an American phenomenon – is with a heresy: there is no such thing as the 'nuclear umbrella'.

Yes, the US has security commitments. For example, the ANZUS Treaty committed the parties to 'act to meet the common danger' from an armed attack 'in the Pacific Area on any of the Parties'. And, yes, the US also has a very large arsenal of nuclear weapons that is second to none.

But there is no specific commitment to use any of those nuclear weapons in defence of any ally. The ANZUS Treaty and other US defence agreements do not commit the US to use nuclear weapons. The nuclear umbrella is, at best, an inference. It is, in certain cases, a very reasonable inference, of course. And, were I Kim Jong Il, I wouldn't push it.

One way to think about much of the history of extended deterrence is as a kind of conjuring trick or alchemy – an effort to make real this commitment that is merely implied by the dual reality of US security commitments and the existence of nuclear weapons. 

In Europe, this trick took the form of planning activities and 'nuclear sharing' arrangements in which European pilots in so-called 'dual capable' aircraft trained to drop American nuclear bombs. In other places, like Japan (until 1972) and South Korea (until 1991), the commitment was implied by US nuclear weapons stationed on their territories.

Read More

With the withdrawal of US nuclear weapons from South Korea in 1991, there are no US nuclear weapons forward deployed in the Asia Pacific. Instead, American officials have largely pointed to specific capabilities in the American arsenal that are said to be maintained for the unique task of extend deterrence, on the grounds that if we are willing to spend money, we must be serious.

So, for example, in 2001, the Bush Administration told Japan that the US was retaining the option to deploy nuclear-armed Tomahawk missiles (which were sitting in storage) on US attack submarines just to show we were serious about defending Japan.

The problem with this approach is that US conventional and nuclear capabilities continue to evolve – relying on hardware commits us to those capabilities long past their obsolescence. The US would not, under any conceivable circumstance, have redeployed the nuclear Tomahawk. All along, the Navy intended to retire the system in 2013.

In 2010, the Obama Administration had to choose between explaining to Tokyo that, perhaps, American officials hadn't been entirely truthful in 2002 and the system would be retired, or spending money the Navy didn't have to maintain the system in storage. Fortunately, the Obama Administration decided to proceed with the retirement of the nuclear Tomahawk. The Administration calculated, correctly, that consultations were much more important than the nuclear Tomahawk. 

Unfortunately, the 2010 Nuclear Posture Review also continued the short-sighted practice of pointing to one shiny piece of hardware to replace another. This time, the US asserted that its commitment to extended deterrence was demonstrated by the effort to make nuclear-capable the Joint Strike Fighter (and extend the life of the B61 nuclear gravity bomb it would carry) and maintain the capability to forward deploy US bombers, like the B2, particularly in Guam.

These are, however, irrelevant capabilities that may not survive the current budget austerity. There are no military missions for the B61s deployed in Europe – one NATO official admitted to me that 'we would never drop a B61 off the wing of an airplane' – and the Air Force does not want to spend money giving the JSF an obsolete nuclear capability (nor do our European partners seem keen to modernise their own 'dual capable' aircraft).

Nor would the US forward deploy nuclear-armed B2s, either to Guam or elsewhere (conventionally-armed bombers are another matter). The B2 can reach targets from North Korea to Iran directly from Missouri, which is what the US did in the early stages of operations against Yugoslavia, Afghanistan and Iraq.

Either this Administration or the next is going to end up explaining that the 2010 Nuclear Posture Review, in its own way, was not entirely truthful.

Perhaps the solution is to stop talking about extended deterrence as though it were a separate mission. Instead, we should talk about core deterrence – the central mission of nuclear weapons to deter nuclear attacks against the US, our forces abroad and our allies and partners. One advantage to talking about core deterrence is that it forces us to set aside the hardware and focus on the real question: how do we demonstrate that an ally is part of the core?

On the narrow subject of nuclear weapons, this means emphasising consultation and joint planning that binds countries together. But more broadly, changing our language reflects the reality that the credibility of the 'nuclear umbrella' is really just a manifestation of the credibility of our shared interests and values. Forging an alliance on this basis isn't quite as easy as pointing to shiny new piece of hardware, of course, but it is likely to endure long after this bomb or that missile has gone to rust.


The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.


Photo by Flickr user solidether.


Nobumasa Akiyama is associate professor at Hitotsubashi University and an adjunct fellow at the Japan Institute of International Affairs.

As Bruno Tertrais writes, deterrence is a psychological process. The ability to pose unbearable damage and the will to use such a capability are not sufficient to constitute the deterrent. Such capability and will must be recognised by adversaries as well as allies which are provided extended deterrence. Thus capability, will and perception are the three essential elements for maintaining the credibility of extended deterrence.

It is true that the formula for a credible extended deterrence has become much more complicated. It is especially true in East Asia. Here I would argue two factors among others.

First, the nature of security threats is changing, and thus the roles of deterrence and extended deterrence have been changing. In East Asia, it is highly unlikely that the US would use nuclear forces to retaliate against a North Korean insurgent attack or a Chinese invasion of small islands under dispute in South and East China Seas, for reasons of proportionality. US nuclear force structure seems to be moving away from the idea of massive retaliation against such small attacks.

Even if the nuclear element of extended deterrence was strengthened, the credibility of extended deterrence would not increase so long as adversaries did not consider nuclear retaliation against relatively small-scale aggression plausible. If potential adversaries perceive that the US would not retaliate with nukes, nuclear deterrence may not work. But it should NOT be considered a failure of US extended 'nuclear' deterrence. Rather, it is typical of the 'stability-instability paradox'. So for strategic stability, and increased security of US allies in the region, it is not sufficient that alliances only strengthen the nuclear part of extended deterrence.

Read More

Although the role of nuclear weapons in extended deterrence will not vanish, other elements of deterrence and extended deterrence are increasingly important to maintain credibility by assuring allies and adversaries in various crisis scenarios. Since the role of deterrence is to reduce the risk of conventional aggression as well as nuclear exchange, extended deterrence should pursue the best mix of nuclear, non-nuclear (conventional) forces and political means.

Furthermore, as Shen Dingli points out, there are various interpretations of the effectiveness of extended nuclear deterrence. It indicates that the perception element among three elements of credible extended nuclear deterrence needs to be further considered.

Second, asymmetry between the US and China in force structure and strategic objectives makes it complicated to establish strategic stability in East Asia. While US strategic forces shift towards long-range strategic nukes, the majority of Chinese nuclear forces remain short- and medium-range (although China has been developing its long-range capabilities). US arms control policy values a balance of forces as well as transparency and verification, while China puts more emphasis on confidence building measures such as no-first-use and the negative security assurance, both of which are difficult to verify.

During the Cold War, strategic stability between the US and Russia converged into a balance in the number of strategic nuclear weapons, and was consolidated through arms control treaties by the concept of mutually assured destruction. Two countries developed transparency and verification measures to maintain strategic stability. In such a relationship, the US and Russia admitted that they were both in a mutually vulnerable situation, and mutually confirmed a kind of 'pro forma' standardised strategic stability.

In a time when we are experiencing a paradigm shift regarding the role of nuclear weapons in international security and relationships among nuclear armed states, traditional nuclear deterrence logic that was built upon the history of the US-Soviet (Russian) confrontation does not necessarily guarantee the legitimacy of nuclear weapons or nuclear deterrence. Rather, the necessity of structuring a new logic of 'strategic stability' is rising. There is a need for strategic dialogue to converge different arms control philosophies into a shared vision, involving both nuclear weapons states and their allies.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user m'sieur rico.


Crispin Rovere is a PhD Candidate at the Strategic and Defence Studies Centre, ANU.

One way of predicting the future of extended nuclear deterrence (END) is by measuring its utility at present. To what extent does it shape the strategic environment or influence the behaviour of allies? To begin with, END does not dissuade allies from seeking nuclear weapons of their own.

Israel maintains them in secret. Its nuclear arsenal inspires other countries across the Middle East to seek nuclear weapons, undermines sanctions against that pursuit, and destroys any hope of implementing a Middle East Nuclear Weapons Free Zone. On a long enough time scale, it is not credible for Israel to maintain a regional nuclear monopoly.

One wonders what strategic advantage Israel's nuclear weapons confer that adequately compensate for all these obvious drawbacks, especially given Israel would, if it disarmed, remain firmly protected under America's END. Israel maintains its capability even though END is most credible in a region where none of the states subject to a US nuclear attack could retaliate against the American homeland.

In Europe, it is hard to imagine what imminent threat is posed to British or French sovereignty that justifies maintaining an independent nuclear deterrent. As Hugh White points out, Russia is not going to embark upon a war of conquest against continental Europe, and nuclear weapons would not be employed to defeat it if Russia did. Frühling & Schreer's argument that even a small reduction in the risk of a major war is a good thing only holds if a nuclear war never occurs. As unhappy as such an equation might be, three major conventional wars between non-nuclear states are preferable to a single nuclear one. 

The US uses END as its justification for maintaining nuclear forces substantially larger than required for minimum deterrence. If the US were to disarm, the argument goes, nuclear proliferation will become more likely because its allies, devoid of a nuclear umbrella, will pursue nuclear options themselves.

Read More

This policy creates cynicism towards the NPT; an increasing belief that, far from being a road map to a nuclear free world, it is a document enshrining nuclear weapons as a God-given right for some, but which no one else has any right to acquire. This status quo, along with the END argument that underpins it, is wholly unsustainable.

The NPT is further undermined by double standards toward those (such as India) existing outside the NPT but who have nevertheless become treated as NPT nuclear weapons states, minus any of the responsibilities or obligations that come with it.

The existence of END damages nuclear disarmament directly by providing cover for other nuclear armed states. Russia remains largely free of criticism so long as it maintains parity reductions with the US. Other nuclear powers claim they will participate in multilateral reductions once the two Cold War arsenals are reduced closer to minimum deterrence levels, but since this cannot happen while END exists, they get a free ride.

The US believes its revised negative security assurance reassures non-nuclear states while maintaining nuclear deterrence against perceived proliferators. This is nonsense. In no way does the revised negative security assurance have this effect, and in some important ways it causes considerable harm. Short of a country being a launch site for a nuclear attack, there is no scenario in which the US would conduct a nuclear strike against a non-nuclear weapons state; the assurance that it will not do so is simply stating the obvious.

Also, the overt exclusion of Iran and North Korea from the negative security assurance is a serious error. Everyone knows that the US is never going to launch a pre-emptive nuclear strike on either country, so the only practical effect of the revised assurance is to lend legitimacy to the Iranian and North Korean nuclear programs. Surely if there is one legitimate reason for acquiring nuclear weapons, it is somebody openly threatening to use them against you.

The nuclear taboo that constrains nuclear powers and limits the utility of nuclear weapons holds similar implications for END. Since no state actor can achieve desired foreign policy goals by nuking non-nuclear weapon states, END is unnecessary. Its removal would assist disarmament efforts and strengthen the nuclear taboo even further.

It is also the case that non-nuclear weapon states are seldom deterred or successfully blackmailed by another's nuclear arsenal. Argentina was not deterred by Britain's nuclear weapons in the Falklands, and China was not deterred by America's nuclear weapons in Korea. In the aftermath of the Second World War, the Truman Administration was perpetually dismayed by just how unfazed the Soviet Union was by America's global nuclear monopoly; the Soviet Union continued to maintain all of its political demands and even blockaded Berlin.

END is a self-serving policy with little, if any, benefit to the common strategic good. It is saddled with many drawbacks, in particular providing political cover for major nuclear powers and undermining nuclear disarmament efforts.

The argument that other countries will acquire nuclear weapons without END is unpersuasive. By far the greatest constraint on horizontal proliferation is the NPT, and its continued degradation, in part due to END, is one of the biggest threats to nuclear disarmament worldwide.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user windy_sydney


Hirofumi Tosaki is a senior research fellow at the Center for Promotion of Disarmament and Non-Proliferation, Japan Institute of International Affairs.

It is quite difficult to form a definitive answer to the question of whether the age of extended nuclear deterrence (END) is coming to an end, since END is inherently complex, and its complexity is increasing after the Cold War. But my answer is 'no' for the foreseeable future, unless the security situation, or US nuclear policies, or Northeast Asian policies, change dramatically. Take Japan as an example.

During the Cold War, what Japan ultimately expected END to do was deter threats posed by the Soviet Union to Japan's national survival as a liberal, democratic country. Currently, no such 'existential threat' exists. However, Japan has heightened its concerns about challenges to its national interest, such as issues of territory, maritime interest, regional and international order, and possible attempts by other counties possessing nuclear forces to change the status quo using their military powers.

In such a security environment, Japan, which maintains an 'exclusively defensive defense' policy and does not possess any capability to retaliate against other country's territory, expects US END to continue to play an important role of deterring a wide range of possible challenges to its national interest.

Japan has been concerned about the credibility of US END. There are possibilities of the US being deterred, of the stability-instability paradox increasing or even of Japan facing abandonment. This is because of the asymmetric scale of interests between the US and a certain adversary in the region, and that particular adversary's development of asymmetric capabilities, especially its acquisition or reinforcement of nuclear retaliatory capabilities against US forces and its homeland. However, this does not mean that END is becoming anachronistic or irrelevant.

Read More

As Bruno Tertrais suggests, it is inconceivable that an adversary, if rational, would fail to consider the possibility of US nuclear retaliation, however remote it is from the US, and even if it possesses a reliable second strike capability, when it comes to conducting even a limited military option against Japan.

Of course, we should not place exaggerated hopes on END, as US allies such as Japan have occasionally done. Even if the US maintains a robust defense commitment, including the provision of END, an adversary may conduct a low-intensity military option under deliberate consideration, or by miscalculation, misperception, overestimation of its capabilities, or even underestimation of US resolve. END is not a panacea, and we need to keep contemplating the roles and limits of END in a realistic manner.

END is indeed significant, but it is just one of various means for guaranteeing an ally's security and mitigating threats. Rather than just relying on END, it is imperative for Japan to reinforce its efforts, under close coordination and cooperation with the US, to construct stronger conventional capabilities — including addressing anti-access or area denial capabilities, developing missile defense and sharing roles, burdens and capabilities — for adequately deterring the low-intensity, limited challenges that Japan may face, as well as denying an adversary's escalation or achievement of objectives.

Such efforts will contribute to further enhancing Japan's own deterrent capability, the Japan-US alliance, and the credibility of US END.

Finally, END will become less and less imperative where military challenges or threat have been mitigated or have vanished. In achieving this goal, the importance of enduring efforts toward deepening strategic dialogues, developing transparency and confidence building measures, creating a communication mechanism in case tension heightens, and promoting arms control and non-proliferation are steadily but surely increasing in Northeast Asia.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user Ciro@Tokyo


Hyun-Wook Kim is Professor at the Institute of Foreign Affairs and National Security in Seoul.

Is extended nuclear deterrence dead? It is not easy to answer this question, but my answer is that it is still effective and cannot be ignored.

The first argument concerns providing assurance to allies. South Korea has always been very sensitive to changes in US security policy. After the end of the Korean War, Seoul was shocked by the decrease of US forces in the region, forcing the US to introduce US tactical nuclear weapons to South Korean territory. When President Nixon withdrew US forces from South Korea in 1970s, the South Korean president attempted to develop a domestic nuclear capability. And most recently, when the US Nuclear Posture Review identified a reduced role for nuclear weapons in providing extended deterrence to allies, the response of South Korean elites was very sensitive. Along with the recent North Korean military provocations, there even emerged voices that US tactical nuclear weapons should be reintroduced to South Korea.

As mentioned by Bruno Tertrais, the psychological impact that nuclear weapons possess is very significant. Everybody knows that the actual use of nuclear weapons is an uncommon thing unless it pertains to vital or existential interests of states. But the case of two Koreas belongs to this category.

Why would North Korea pursue nuclear weapons? From the perspective of Western countries, it may be as diplomatic and domestic bargaining chips. But the North Korean position is that its security is tremendously in danger from the US, and that this vital danger pushes the North to develop nuclear weapons. On the Korean peninsula, nuclear weapons are closely tied with vital interests. Living with North Korea as an imminent danger, South Korean vital interests are also under threat, which necessitates the sincere provision of US extended nuclear deterrence.

Read More

What I mean by 'sincere' is not just verbal promises but real action by the US. After the first North Korean nuclear test in 2006, more than 67% of South Koreans approved the idea of developing South Korea's own nuclear weapons. With the imminent threat of North Korea, South Korea did not feel secure with the provision of extended nuclear deterrence from the US, which is so remote from South Korea. So the problem does not lie in whether extended nuclear deterrence is needed or not, but whether extended nuclear deterrence can be provided in assured and valid ways.

The second argument for extended nuclear deterrence concerns the credibility of deterring North Korea. The probability that the North would actually use nuclear weapons is very minimal, and most of the skirmishes in the Korean peninsula are limited to the small-scale aggressions like the Cheonan ship sinking and Yeonpyong island military attack. In this vein, George Perkovich argued that the role of non-nuclear deterrence is significant. But still, in this case, the role of nuclear deterrence cannot be ignored.

Recent North Korean military provocations prove that US extended deterrence has failed. Despite ongoing US promulgations of its solid extended deterrence provision to South Korea, North Korea has continued its military provocations because the North knows that the US would not willing to retaliate: the US does not want small skirmishes to be escalated into a large-scale warfare. This implies that the US and South Korea are deterred by the North.

This stalemate necessitates the role of nuclear weapons. All the partners within and surrounding the Korean peninsula are well aware that the risk of using nuclear weapons is very minimal. But it is more about psychological effect and a game of bluff. It is such a game of bluff that made North Korea restless when the US nuclear-powered aircraft carrier George Washington participated in the US-South Korea military exercises near North Korea. Actual use of nuclear weapons is unlikely, but it still retains a psychological deterrent effect.

Cold War strategic stability was maintained by the nuclear strategies of the US and Soviet Union based upon their virtual use of nuclear weapons. Now, it rests more on psychological, if not symbolic, effect.

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user US Army Korea - IMCOM.


In the late 1960s and early '70s, Australian strategic policy underwent two transformations. Conventionally, fears about the concurrent retrenchment of British and American power led Canberra — for the first time in its history — to begin shedding its strategic dependence in favour of a more self-reliant defence policy.

In the nuclear realm, things went in the exact opposite direction. Whereas Canberra had spent parts of the 1960s in active, if sporadic, pursuit of its nuclear ambitions — first by lobbying the British to supply ready-made nuclear weapons, later by devising plans for an indigenous uranium enrichment capability — by the early 1970s, Australian had reversed course. With a change of government and the advent of détente and a global arms-control regime, nuclear plans were shelved, the NPT signed and Canberra's place under the US nuclear umbrella reoccupied and reserved indefinitely.

The legacy of this episode is an enduring tension in Australian strategic policy. On the one hand, Canberra is committed to defence self-reliance, defined by the 2009 White Paper as the ability to 'deter and defeat armed attacks on Australia...without relying on the combat or combat support forces of other countries.' Yet against the threat of nuclear weapons and ballistic missiles, the most destructive weapons of all, the operational employment of which could have a devastating effect without warning and in a single strike, Australia remains entirely dependent on the US for extended nuclear deterrence.

Is this a viable strategy? Is it prudent? Or is END an article of faith, as some of Australia's best strategists have suggested, fated to obsolescence by the ongoing transformation of the regional strategic order and the fluid nuclear landscape this entails?

Read More

There are a number of reasons to be pessimistic. To begin with, Australia's nuclear dilemma reflects a broader problem in its alliance relations with the US, which I've explored elsewhere. The dual risks of abandonment and entrapment, common to all alliances, sharpen as the distribution of world power shifts, with important implications for the credibility of America's nuclear umbrella.

The risks of entrapment arise in two forms. At a general level, Canberra's strategic dependence on Washington, including for END, circumscribes its ability to dissent from American policies that may not accord with Australian interests. Cognisant of its privileged place under the US nuclear umbrella — and of its limited range of alternatives, should END be withdrawn in the event of Australian non-compliance (a la New Zealand circa 1984-1987) — Canberra could find itself dragged into a costly war alongside the US that its fortuitous strategic geography might otherwise allow it to avoid.

More directly, the joint facilities that Australia hosts as a quid-pro-quo for US nuclear assurances — critical as they are to America's intelligence, surveillance and reconnaissance network — could present as valuable targets in major war that, by virtue of their distance from Northeast Asia, only nuclear-tipped ICBMs may be reliably assured of destroying.

This may sound rather fantastic, yet it's not a consideration that Australian defence planners can automatically discount. Moreover, there is a historical basis to such concerns. As former defence minister Kim Beazley once noted in the context of the Cold War, 'we accepted that the joint facilities were probably targets, but we accepted the risk of that for what we saw as the benefits of global stability.'

These concerns notwithstanding, it is the risks associated with abandonment that are the most acute. Credible nuclear deterrence operates on the basis of a defending state's threat — through some combination of explicit pronouncements and the deployment of nuclear capabilities — to impose intolerable costs on an attacker, thereby preventing the attacker from using nuclear weapons against the defender's ally. By contrast to conventional deterrence, where risks are generally limited, nuclear deterrence relies on the defender — in this case, the US — conveying a willingness to sacrifice large swathes of its population on behalf of Australia, which is an essentially non-vital, possibly even marginal, strategic interest. The question for Australian defence planners is: would the US government really trade San Francisco for Sydney?

Indeed, if Canberra's broader quest for self-reliance reflects enduring concerns about the potential limits of Washington's willingness or ability to furnish assistance in a strategic crisis, there is no compelling logical reason why, with the balance of power shifting away from US primacy, those concerns should be so sharply limited to conventional threats. If anything, Australian defence planners should expect Washington to act with greater caution and more reticence in a nuclear crisis, however unlikely one might be, when the costs and risks to itself will inevitably be exponentially greater.

I ventured a few thoughts on how Australia might begin redressing its nuclear dilemma in this article a few years ago. Unfortunately, however, the traditional impetus for change in Australian strategic policy is crisis, or at least perceptions of crisis, at which point it's usually too late to do anything other than 'muddle through'. This is likely to be doubly true of nuclear issues, which, given the wide-spread aversion to even discussing them in Australia, remain largely beyond the realm of acceptable political discourse. So don't expect any movement for at least a couple of decades. 

In the meantime, is END dead? Let's hope we never have to find out.


Some have argued that nuclear deterrence is not failing, as it is for deterring big events – national survival, for instance. This claim may never be proven, as no state on this planet would initiate a nuclear strike against America (North Korea and Iran would not, even if they could), not because they are deterred but because they have no need.

America's long-standing nuclear deterrence posture, including extended deterrence, has three assumptions of circumstances: if American territory, its overseas military presence or allies are attacked by a non-nuclear rival in an alliance with a nuclear power, the US would resort to first use of nuclear weapons.

But even given extended nuclear deterrence, who really believes that the US would shoot a nuclear weapon unconditionally in defence of an ally if that ally was being attacked by a non-nuclear rival that is in alliance with a nuclear weapons state? The US, after long domestic debate, has officially ended this policy and this type of extended nuclear deterrence. The Obama Administration's Nuclear Posture Review of 2010 stated clearly that the US would not use nuclear weapons against non-nuclear-weapons NPT members, if they meet the NPT requirement (Iran could be an exception).

Read More

This is a much commended, progressive nuclear policy, reducing the role of nuclear weapons in international politics. There is a dilemma: the more the US stresses deterrence and extended deterrence, the more its allies could feel assured and less likely to pursue their own nukes, but the more likely some other countries would feel the importance of nuclear weapons and their associated deterrence and therefore seek their own nuclear path. Everyone has to weigh the balance between security due to the nuclear umbrella and insecurity due to nuclear proliferation.

Obama's policy has obviously reduced the role of nuclear weapons in America's national security, so as to create a situation more conducive to nonproliferation. Obviously this has an impact on extended nuclear deterrence — the US now is clear that it would not protect its allies with nuclear weapons against non-nuclear attack, possibly with Iran as an exception. Its current nuclear doctrine has deviated from traditional extended nuclear deterrence — if an aforementioned non-nuclear state rival now brings harm to America or its ally, as long as it is in good standing with the NPT, the US has promised not to retaliate with nuclear weapons.

This may sound frustrating to some but one has to remember that every coin has two sides – America wants to bring more security to itself and its allies in an age of proliferation. It has to commit less to its nuclear umbrella and to its friends, though it could still assure its allies through other means in the future.

Whether one likes such a change or not, this is the Obama position. So how we can say nuclear extended deterrence has not declined?

The Nuclear Reactions column is supported by the Nuclear Security Project of the Nuclear Threat Initiative, as part of a wider partnership between the NSP and the Lowy Institute.

Photo by Flickr user x-ray delta one.