home       biography       publications       speeches       organisations       images       @contact

Nuclear Deterrence in Asia and the Pacific

Presentation to Inaugural Conference of Asia and the Pacific Policy Society (APPS)/Asia and the Pacific Policy Studies Journal by Professor the Hon Gareth Evans AC QC, Australian National University, Canberra, 7 September 2012

Sometimes the biggest policy questions of all get less attention than they deserve. So it is with nuclear disarmament, an issue that in earlier decades mobilized hundreds of thousands of activists all over the world, and on which every political leader and senior policymaker had to have some kind of an opinion, but which now barely resonates at all with policymakers or publics.

Part of the reason seems to be complacency: the perception that in the post Cold War world nuclear stockpiles are not the threat they may once have been. Another appears to be an ingrained fatalism: the perception that nuclear weapons can’t be uninvented, are always going to be with us, and that there is little point in playing Don Quixote. Perhaps most importantly – and in many ways more disconcertingly – there is the perception on which I will focus in this talk: that disarmament is actually undesirable, because nuclear deterrence works.

Having wrestled with these issues in a variety of incarnations over the years,[1] I believe that is critical that all these perceptions be challenged, and nowhere more so than in  Asia and the Pacific, where every nuclear-armed state except the UK and France is a player, and where, unlike anywhere else in the world, stockpiles of nuclear weapons – in China, India and Pakistan, quite apart from what may be happening at a much lower level in North Korea – are believed to be actually growing.

A credible case has to be made – in a contemporary 21st Century context, and focusing particularly on the realities of our own region – not only that nuclear disarmament is achievable, albeit only over a protracted period, but also that it is highly desirable. And that means above all contesting the tenacious belief, by policymakers and those who influence them in nuclear-armed states and their allies, that nuclear deterrence is of real value to their national security, and that its benefits outweigh any possible costs.


Belief in the deterrent role of nuclear weapons is not the only reason why states acquire and retain nuclear weapons, and are reluctant to pay more than lip-service to the objective of eliminating them.  In particular, for some states – most obviously France, UK and India, and these days probably Russia as well – considerations of status and prestige have been and remain absolutely crucial. 

There is also a lingering concern in some quarters that states may want to possess or acquire nuclear weapons as instruments of aggression: but I am not alone in remaining unpersuaded that any nuclear-armed, or putatively nuclear-armed, state in this region, or anywhere else, will be willing to break the international normative taboo against the aggressive use, or threat of use, of such weapons which unquestionably now exists.[2] In security terms, it is the deterrence rationale, not any warmaking one, that is overwhelmingly the driver.

If movement toward disarmament is ever to gain real momentum, the strong and widespread belief in the deterrent utility of nuclear weapons in neutralizing external threats to national security, needs to be seriously challenged, in the three different contexts in which it arises.

The first context is symmetrical nuclear deterrence, where states of roughly comparable size and resources acquire and retain nuclear arsenals that, while not necessarily themselves of roughly equal size, are in each case big enough and secure enough to credibly retaliate after a first strike by their adversary,[3] and which are designed to inhibit that adversary accordingly from contemplating any such strike – as with the U.S. and Russia, U.S. and China, Russia and China, and China and India.

The second context, an extension of the first, is extended nuclear deterrence, where a nuclear armed state operating in a symmetrical deterrence environment extends the retaliatory protection of its own nuclear arsenal to allied states living in potentially dangerous neighbourhoods, with at least part of the rationale being to dissuade those allies, who might otherwise be capable of doing so, from acquiring nuclear weapons of their own – as with the U.S. in relation to Japan, the ROK and Australia.

The third context is asymmetrical nuclear deterrence, where a state of very unequally small size and resources as compared to one or more notional adversary, acquires or retains nuclear weapons with the object of raising the pain threshold so high that would-be regime changers, territory-acquirers or punishers would think again – as with Pakistan in relation to India, and North Korea in relation to South Korea and the U.S.


The arguments against nuclear deterrence need to be given much more systematic articulation and exposure than has been common in either the academic or policy-focused literature of recent years.[4]  They are of two main kinds – those emphasizing the fragility of any nuclear deterrence calculation, and those making clear the highly dubious utility in practice of relying on it. Acceptance of both sets of arguments may not be a sufficient condition for changing policy, but they are a necessary starting point.

Fragility. Nuclear deterrence has always been a fragile basis for maintaining stable peace, for three main reasons.

First, nuclear deterrence depends on rational actors on both sides, each making rational judgments about the risk factors involved. Political actors and circumstances can change, and it cannot be assumed that complete rationality will always prevail in the stress of a real-time crisis. As Hedley Bull has said, ‘mutual nuclear deterrence …does not make nuclear war impossible, but simply renders it irrational’. And as he also wryly put it, a rational strategic man is one ‘who on further acquaintance reveals himself as a university professor of unusual intellectual subtlety’.[5]

Second, there is an endemic risk not only of human error or misjudgement under stress, but of miscommunication – the risks here now compounded by the sophistication of cyber weapons – and of basic system error, with harmless events being read as threatening. Much archival evidence of the Cold War years – when command and control systems on both sides were thought to be highly sophisticated, and were more so than are some between potential nuclear adversaries today – has now revealed how close to calamity the world regularly came, much more so than was understood at the time. It’s not a matter of good policy or good management that the world has avoided a nuclear weapons catastrophe in the last 67 years: it is sheer dumb luck.

Third, there is always a risk that new technical developments will make old calculations redundant, in particular that missile defence systems will be developed that will be extremely destabilizing – by destroying the adversary’s capacity for effective weapons delivery unless he fires first. While it is arguable that the prospect of completely effective (and thus destabilizing, if unilaterally deployed) missile defence is much more scientifically remote than claimed by both supporters and critics, the basic point about technological uncertainty remains.

Dubious Utility. Nuclear deterrence has also been a much more dubious basis for maintaining stable peace than is commonly assumed. All the main arguments in its favour have, on closer examination, nothing like the force they usually seen to possess.

First, the argument that nuclear weapons have deterred, and will continue to deter, war between the major powers. While nuclear weapons on the other side have always constituted a formidable argument for caution, it has been argued persuasively that knowledge of the existence of supremely destructive weapons (as with chemical and biological weapons before 1939) hasn't stopped war in the past; and that the experience or prospect of massive damage to cities and killing of civilians has not caused leaders in the past to back down (including after Hiroshima and Nagasaki, where the historical evidence is now that the Soviet declaration of war the same week, not the nuclear attacks, was the key factor in driving Japan to sue for peace).There is no evidence whatsoever that during the Cold War years either the Soviet Union or the U.S. was determined to go to war at any particular time, and only deterred by the existence of the other’s nuclear weapons.

Second, the argument that nuclear weapons will deter any large-scale conventional attacks. There is a long list of examples where non-nuclear powers have either directly attacked nuclear powers or not been deterred by the prospect of their intervention: e.g. the Korea, Vietnam, Yom Kippur, Falklands, two Afghanistan and first Gulf wars. Whether the calculation in each case was that the nuclear taboo, or some other factor, would inhibit a nuclear response, the point is simply that nuclear weapons did nothing to assist peace and stability.

There are also cases where the presence on both sides of nuclear weapons, rather than operating as a constraining factor, has been seen as giving one side the opportunity to launch small military actions without serious fear of nuclear reprisal (because of the extraordinarily high stakes involved in such a response): as with Pakistan in Kargil in 1999, and DPRK in the sinking of the Cheonan and shelling of Yeonpyeong Island in 2010. It may be that rather than – as the old conservative line would have it – ‘the absence of nuclear weapons would make the world safe for conventional wars’, the presence of nuclear weapons has made the world safe for such wars.

Third, the argument that nuclear weapons will deter any chemical or biological weapons attack. This is claimed by some nuclear-armed states and their allies but lacks plausibility. Given that these weapons have nothing like the destructive potential of nuclear weapons – and never will with chemical weapons, although the future risk factor is higher with biological weapons – it is difficult to paint a plausible attack scenario in which nuclear, as distinct from conventional, retaliation would be a proportional, necessary and therefore credible response.[6]

Fourth, the weakest argument of all, that nuclear weapons will deter terrorist attacks.  Nuclear weapons are manifestly neither strategically, tactically nor politically useful for this purpose. Terrorists don't usually have territory, industry, a population or a regular army which could be targeted with nuclear weapons. And to conduct nuclear strikes on another state, even one demonstrably complicit in a terrorist attack, would raise huge legal, moral, political and strategic issues:  if not contemplated in Afghanistan after 9/11, it is difficult to imagine such strikes ever would be.


I see the task of those of us interested in public policy as being not just analysis but influencing policymakers, and I hope this remains a central objective not only of the Crawford School but of this new Asia and Pacific Policy Society and the new journal advancing its aims.  What that means for me in the present context is refining, developing and communicating these arguments against reliance on nuclear deterrence in a way that has real resonance for those working on these issues on the ground. Changing mindsets on deterrence may not be a sufficient condition for moving toward nuclear disarmament, but there can’t be much doubt that it’s a necessary one.

If the attention of even the most mind-changed policymakers is going to be captured and sustained, there is also a need to focus on achievable next policy steps over the years ahead, bearing in mind that even on the most optimistic view of what is practically achievable getting to a nuclear weapon free world is going to be a very long haul indeed. I am inclined to think the most useful way forward would be to focus on what I call the ‘four Ds’ – Delegitimizing, Doctrinal Change, De-Alerting, and Downsizing – but I’ll save for another occasion spelling out what all that means. 

What I can’t spare you is a brief final reminder of why all this matters. Nuclear weapons are the most indiscriminately inhumane weapons of destruction ever invented. Some 23,000 of them still exist, all but a few hundred of them possessed by Asia Pacific states, with a combined destructive capability of 150,000 Hiroshima sized bombs, and with some 2,000 of them maintained on dangerously high launch-on-warning alert. 

As the Canberra Commission in 1996 put it most succinctly, in words that have been echoed since by many others:  So long as any state retains nuclear weapons, others will want them. So long as any nuclear weapons remain, they are bound one day to be used, if not by deliberate design, by accident or miscalculation. And any such use would be catastrophic for life on this planet as we know it.

And as I have put it on multiple occasions, there is only one other global policy issue remotely comparable in terms of its impact on planetary survivability, and that is climate change. But nuclear weapons can kill us a lot faster than CO2.


[1] Earlier as Foreign Minister (1988-96) and President of the International Crisis Group (2000-09), and more recently as co-chair of the Australia-Japan sponsored International Commission on Nuclear Non-Proliferation and Disarmament (ICNND), Convenor of the Asia Pacific Leadership Network on Nuclear Non-Proliferation (APLN), and Chair of the Advisory Board of the Centre for Nuclear Non-Proliferation and Disarmament (CNND) based here at the ANU Crawford School of Public Policy.

[2] Notwithstanding the fears that continue in some parts of the world, for example of Iran in Israel, I doubt that any state these days believes, if any ever did, that nuclear weapons would enable it to wage and win an aggressive war, or even that the threat of using nuclear weapons would enable it to achieve concessions that would otherwise be unachievable. Apart from the normative taboo, there are formidable practical obstacles involved in the use (and by extension threatened use) of these weapons at both the tactical and strategic level, not least the damage they can cause to one’s own side, and any territory being fought over: cf. ICNND Report Eliminating Nuclear Threats: A Practical Agenda for Global Policymakers (2009), para 6.2. See generally Nina Tannenwald, The Nuclear Taboo (Cambridge University Press, 2007).

[3] What is seen as ‘big enough and secure enough’ will range all the way from ‘minimum deterrence’ capability, implying a retaliatory strength seen as great enough to knock out just enough of the adversary’s civilian or military assets to deter aggression, to the kind of overkill-capability guaranteeing the ‘mutually assured destruction’ of each other’s whole states (in fact many times over) that characterized the nuclear arms race between the U.S. and Soviet Union at the height of the Cold War.

[4] There is less writing on the theory and practice of nuclear deterrence in a 21st century context than one might expect. Ward Wilson is one exception, with his seminal article ‘The Myth of Nuclear Deterrence’ Nonproliferation Review, Vol. 15, No. 3, November 2008 and forthcoming Five Myths About Nuclear Deterrence (Houghton Mifflin, 2013)The section on ‘Rethinking Deterrence’ (paras 6.9-39) in the ICNND Report on which the present paper builds,  made an effort to bring together the main relevant arguments.

[5] The first quote is from The Anarchical Society (Macmillan, 2nd ed 1995) p.234, the second from The Control of the Arms Race (Institute of Strategic Studies, 1961), p.48

[6] On the question of whether Iraq refrained from using chemical weapons in 1991 because of fear of U.S. nuclear retaliation see ICNND Report, p.238, note to para 6.18.