CIAO DATE: 04/03

No First Use

Lawrence Freedman

Pugwash Conferences on Science and World Affairs

November 15-17, 2002

The question of a no-first use (NFU) declaration originally arose as a means of challenging NATO's reliance on the threat of nuclear first use for purposes of deterrence. The proposal would require an authoritative shift away from such an apparently reckless default setting. My own view in cold war days, and for that matter now, is that the objective made sense but that the method was flawed. NATO countries. and indeed any other nuclear powers, should avoid situations whereby their only pre-planned choice in war would be to move to nuclear use. It is best to be able to contain any conflict at the lowest possible level of escalation. However wars have dynamics all of their own and it would be naïve to assume that, whatever peacetime promises had been made, that at some point countries, anxious to secure a victory or stave off defeat or just simply in a confused and desperate condition, would not resort to nuclear use. If nuclear powers are on the verge of war I would prefer there to be an element of uncertainty about where a conflict might lead, rather than expressions of confidence, based on no-first use declarations, that the conflict has reliable restraints in place. To take a contemporary example, I would not expect those advocating a military campaign to eliminate Iraq's mass destruction capability to argue for action on the grounds that Iraqi promises never to use this capability first had removed one major risk.

The objective behind the original pressure for NFU was realised at the end of the cold war, although only briefly were nuclear weapons moved firmly to the last resort. When NATO leaders met in London in July 1990 the new goal was to seek 'the lowest and most stable level of nuclear forces needed to secure the prevention of war.' The new role for nuclear weapons was also clarified:

These will continue to fulfill an essential role in the overall strategy of the Alliance to prevent war by ensuring that there are no circumstances in which nuclear retaliation in response to military action might be discounted. However, in the transformed Europe, they will be able to adopt a new NATO strategy making nuclear forces truly weapons of last resort.

The communiqué declared that the allies would 'modify the size and adapt the tasks of their nuclear deterrent forces.' The French government judged that NATO had gone too far, and it explicitly disassociated itself from the military parts of the London Declaration. A year later, when drafting NATO's Strategic Concept of November 1991, Britain and France asked for the words 'weapons of last resort' to he deleted. The document stated that:

Nuclear weapons make a unique contribution in rendering the risks of any aggression incalculable and unacceptable. Thus, they remain essential to preserve peace. 'The fundamental purpose of the nuclear forces of the Allies is political: to preserve peace and prevent coercion and any kind of war. They will continue to fulfill an essential role by ensuring uncertainty in the mind of any aggressor about the nature of the Allies' response to military aggression. They demonstrate that aggression of any kind is not a rational option.

At any rate the 'last resort' is not an objective test, as it depends not on a moment of utter desperation but the last point at which it supposed that action can be taken before that moment arrives. The advantage of the new situation was that this had become an even more hypothetical problem than before. It was nonetheless easier to build an alliance consensus by acknowledging some role for nuclear deterrence, while accepting an orderly reduction in force levels. A residual nuclear arsenal had some value 'just in case'. The continued existence of nuclear weapons provided a reminder of the dangers of total war. Even without an authoritative specification of the chain of events which might end in catastrophe the thought that 'it just might', could introduce an immediate element of caution into any developing conflict. NATO's position could be taken as emphasising a fact of strategic life rather than setting guidance for force planning.

The case for NATO asserting the right to initiate nuclear war could no longer be based on the presumption of conventional inferiority vis-à-vis the Warsaw Pact. With the abrupt shift in the conventional balance in favour of NATO, one of the past drivers of nuclear strategy - the need to develop credible systems, doctrines and tactics with which to escalate - ceased to be a problem. Another argument also fell away. Nuclear munitions had once been judged essential if a range of critical military tasks were to be accomplished. Now not only was there no longer any need to worry about the quantitative advantage of the Warsaw Pact, but also extraordinarily precise and lethal conventional weapons could devastate large force concentrations, and shatter hardened targets. Nuclear weapons were required neither to compensate for weaknesses in conventional forces, nor to intimidate non-nuclear powers armed with only conventional forces.

That left threatened nuclear use by another as a proper subject for deterrence. That was the view taken with regard to western capabilities but of course with so many states non-nuclear the opportunities for their intimidation by a nuclear state would always be there. This always, for example, seems to me to be the best answer to the 'what if' question of how would the allies have acted during the Gulf War if Iraq was already known to have a nuclear capability. Certainly they would have been more circumspect if it came to liberating Kuwait, but the point is that Saudi Arabia and Kuwait would already have been in appeasing mode, or else have required explicit security guarantees from the US, so the crisis would have arisen in a quite different form. There is good reason to suppose that the moderation of Arab governments aspirations to eliminate the state of Israel during the 1970s was in part a reflection of not only failure in conventional war but a recognition that Israel had a 'last resort' capability - and would probably be prepared to use it. The 'nuclear weapons can only deter other nuclear weapons' proposition thus puts an enormous premium on alliance (in testing conditions) or else is an argument for proliferation. Thus as radical Arab states lost confidence in Soviet security guarantees they attempted to acquire their own nuclear - or at least chemical - arsenals

The proposition also did not help with regard to force structures. Some readily accessible nuclear capacity would be needed. If, in a euphoric response to happier political circumstances, established force structures were dismantled then, should the security situation deteriorate, a sudden demand for nuclear deterrence might re-emerge. There would then be a need for a 'reconstitution' policy to reinforce nuclear guarantees in an emergency. Yet once the weapons had been removed would it be possible to give them a revived role without aggravating the crisis that has occasioned the re-appraisal? Bringing weapons out of store, putting others back on alert, coyness about deployment and targeting plans would appear provocative. Again the alliance problem would be a severe complication because of the need to reassert guarantees through a physical presence.

Of course the main purpose behind the proposition that the only plausible purpose for one nuclear arsenal was to deter the use of another was to support the case for a move to a nuclear-free world. It was argued in reports, such as that of the Canberra Commission, that somehow, it would be possible to move to a nuclear-free world through a general agreement to eliminate all arsenals simultaneously, as the rationale for the existence of each was wholly bound up with the existence of others. This implied a rather unlikely degree of orchestration.

As strong a line of thought among arms controllers was that a more appropriate route than that of abolition would be marginalisation. With little to deter, the weapons could be assigned an even lower profile, kept off alert status and made as difficult as possible to use. They still accepted that they could come back into play and could never quite be eliminated. The focus of this approach was on usability, both in operational and diplomatic terms. If politicians were able to resist brandishing them at times of crisis then they would gradually lose their legitimacy and move further to the margins of international affairs.

Marginalisation was a deceptively easy option for the West. Now that they had conventional superiority over all-comers, NATO countries had every reason to drain nuclear weapons of any residual legitimacy. At the same time, their potential opponents, realising that they could not win a conventional battle, had incentives to keep the nuclear option. In this sense the West's conventional superiority provided a potential boost to the spread of nuclear weapons, or at least of other weapons of mass destruction. The more proliferation of this sort, the harder it was going to be to really push western nuclear arsenals to the margins.

For example, just as NATO decided that it did not need first use threats the Russians began to regret their embrace of the 'no-first-use' principle in 1982. The difficulties of sustaining this position were noted when the new military doctrine was first drafted in 1992 and by November 1993 it had been discarded. As NATO never embraced the principle it was in no position to complain - nor did it do so. In October 1999 the draft military doctrine addressed nuclear weapons in rather traditional terms, reserving the right to use them 'in situations critical to the national security of the Russian Federation and its allies.' Those Russians questioning the renuclearisation of security policy could only note that it had been of little help in arguments with NATO countries over Iraq and Balkans, and that Russia's geographical position meant that it had an even greater stake than the United States in preventing the proliferation of weapons of mass destruction.

While the case for marginalization gained through comparison with abolition, in practice moves in this direction were half-hearted. Instead the drive was compromised by the reluctance to leave anything to chance. In October 1993 under Clinton's first Secretary of Defense, Les Aspin, a Nuclear Posture Review was initiated. Its conclusions, after a year of study, was that Russia remained the main concern, that the number of launch platforms could be reduced but the triad of ICBMs, SLBMs and bombers should be retained, and that there was a case for extending deterrence to chemical and biological weapons. This latter issue was one reason why there was reluctance to support proposals for a no-first use declaration. The conclusion was that there really was no alternative but to keep the enemy guessing: any attempt to define with precision the circumstances under which a nuclear counter-strike would be launched would generate great controversy and send confusing signals.

When NATO later considered whether it should make an explicit link between chemical or biological attacks and nuclear first use it came to the same conclusion. Until the scale and intensity of any attack was understood it was difficult to he sure of the appropriate response. In most cases sufficient retribution could be exacted by conventional means, but it probably did no harm if those contemplating such attacks took account of the possibility that they just might lead to nuclear retaliation. PDD-60 in November 1997 did allow for the possible use of nuclear weapons in case of a chemical or biological weapons attack, and Secretary of Defense William Cohen repeated the same message on the 17 March 1998:

We have made very clear to Iraq and to the world that if America or US forces are attacked by nuclear, chemical or biological weapons, we have the ability and will to deliver a response that is overwhelming and devastating.

The same themes were picked up by the Bush Administration, which conducted its own Nuclear Policy Review. The shifts were in the integration of non-nuclear and nuclear systems, an enhanced role for active defences, reflecting the administration's determination to end sole reliance on offensive options, and the development of a system that could respond quickly to changes in the strategic environment. It described an essentially residual role for nuclear forces, but one which it was nonetheless prepared to expend substantial resources. While 'nuclear forces, alone' were 'unsuited to most of the contingencies for which the United States prepares' they might be needed for targets able to withstand non-nuclear attack, (for example, deep underground bunkers or bio-weapon facilities)'. It was for this reason that new nuclear earth-penetrating weapons were being considered.

In the Foreword Secretary of Defense Donald Rumsfeld promised that the US would 'no longer plan, size or sustain its forces as though Russia presented merely a smaller version of the threat posed by the former Soviet Union'. Instead he thought it more likely that 'Terrorists or rogue states armed with weapons of mass destruction will likely test America's security commitments to its allies and friends.' The body of the report explored the contingencies that might prompt consideration of nuclear use. It began with 'well-recognized current dangers' including 'an Iraqi attack on Israel or its neighbors, a North Korean attack on South Korea, or a military confrontation over the status of Taiwan' before moving on to potential, plausible contingencies such as 'the emergence of a new, hostile military coalition against the United States or its allies in which one or more members possesses WMD and the means of delivery', and then on to 'sudden and unpredicted security challenges, perhaps resulting from 'a sudden regime change by which an existing nuclear arsenal comes into the hands of a new, hostile leadership group, or an opponent's surprise unveiling of WMD capabilities.' North Korea, Iraq, Iran, Syria, and Libya were listed as being among the countries that could be involved in these contingencies. They all shared 'longstanding hostility toward the United States and its security partners' and they all 'sponsor or harbor terrorists, and all have active WMD and missile programs.'

China, which had loomed large in the initial Administration descriptions of immediate threats, and fitted the classical concerns about a rising strong power, (compared with the main list which consisted of regimes that had all fallen on the wrong side of history), and its 'developing strategic objectives and its ongoing modernization of its nuclear and non nuclear forces' was duly noted. While the review acknowledged that U.S. relations with Russia might significantly worsen in the future, and observed that it still possessed 'the most formidable nuclear forces, aside from the United States, and substantial, if less impressive, conventional capabilities' it struck a conciliatory note to the old adversary

There now are, however, no ideological sources of conflict with Moscow, as there were during the Cold War. The United States seeks a more cooperative relationship with Russia and a move away from the balance-of-terror policy framework, which by definition is an expression of mutual distrust and hostility. As a. result, a [nuclear strike] contingency involving Russia, while plausible, is not expected.

This focus was reflected in the development of policies on missile defence. At the start of the nuclear age it had been almost a moral imperative to develop some system that would make it possible to defend against an incoming bomber or missile attack. By the 1960s this imperative had been undermined by the thought that the main result of such an endeavour would be to persuade the opponent to improve his offensive capabilities, thereby ensuring an even worse outcome than if the condition of mutual assured destruction had just been accepted as a fact of life. This conclusion, elevated to a doctrine, inflamed more conservative strategists and producing a continuing campaign for a missile defence plan. One could imagine circumstances in which leftist strategists would have been equally inflamed. When the conservatives came closest to success, with President Reagan's strategic defence initiative, the main effect was to undermine confidence in deterrence - with disarmament rather than defensive hardware left as the most practical alternative.

During the 1990s proposals came forward for missile defences geared to otherwise weak but potentially-nuclear powers, in particular Iran, Iraq and North Korea. The proposals betrayed a lack of confidence in Western deterrence of these states, who might expect complete devastation should they dare to mount some attack on the United States. This assumed that these states had regimes that were both desperate and reckless - hence the description as 'rogues'. These states were also in practice most likely to be responding to regional concerns rather than some heroic desire to take on the United States directly. The attack on 11 September certainly rekindled interest in homeland defence, but it also demonstrated how useless a missile defence would have been against terrorists capable of using the lowest (but oldest) military technology of a knife to turn a civilian airliner into a guided missile.

In 1990, the Gulf crisis came to be bound up with not only liberating Kuwait from Iraqi occupation but also frustrating Iraq's drive to acquire nuclear, along with chemical and biological, weapons. The cumulative evidence of this drive, and embarrassing disclosures about the culpability of Western countries in abetting it, that led to the rapid deterioration in relations with Iraq during the first months of 1990 and encouraged Western leaders in their efforts to deal decisively with Saddam Hussein. As Saddam had shown himself ready to use chemical weapons, against both the Iranians and Kurds, and had also mounted missile attacks against Iranian cities, it was always likely that threat would be part of Iraqi strategy.

In terms of deterrence theory the Gulf offered a significant case study. It had nothing to say about deterrence in conditions of parity but did offer indications about how to deter unusually reckless states with access to serious means of destruction. The question posed prior to hostilities was how to stop Iraq using chemical weapons either on the battlefield or against Israel and Saudi Arabia. One possible answer was that nuclear threats might be sufficient for this purpose. The British took the view that past negative security guarantees, that is promises made during the 1978 UN Special Session on Disarmament not to use nuclear weapons against non-nuclear states, ruled this out. They assumed that a combination of defensive measures, including protective suits for troops, and overall conventional superiority, meant that the allies could respond as they wished to further outrages without having to perpetrate outrages of their own. Just before the start of hostilities this was made explicit. The French took a similar view.

In private the Americans had no intention of resorting to nuclear use, and this has been confirmed in a number of memoirs. Regardless Secretary of Defense Richard Cheney remained ambiguous in public, on the grounds that it was best to keep Saddam guessing. Saddam needed to be aware, Cheney said,

that the President will have available the full spectrum of capabilities. And were Saddam Hussein foolish enough to use weapons of mass destruction, the U.S. response would be absolutely overwhelming and it would be devastating. He has to take that into consideration, it seems to me, before he embarks upon a course of using those kinds of capabilities.

In addition it was made clear to Iraqi commanders in the field that they would be considered personally responsible for the consequences of chemical use, the most specific deterrent threat, made by Secretary of State James Baker to Tariq Aziz on 9 January 1991, was that if chemical weapons were used then the United States would ensure that Saddam's regime was toppled.

There were a number of indications that the Iraqis understood the threats of retaliation to include nuclear weapons. This was of course not only a question of an American, but also an Israeli response. There may have been an element of ex-post facto rationalisation here. After all it suited Iraq to present its failure to use its chemical arsenal as a result of high strategy, exalting its position as a country that had to be deterred by the most powerful forces of the most powerful state, rather than because its local commanders were disoriented and frightened or because its means of delivery were unsuitable and in disarray. Certainly Israeli studies of the mechanics of using nuclear weapons to deter chemical attacks indicated a number of problems. In addition to the specifically Israeli one of acknowledging a hitherto covert nuclear status, there was also the issue of what to do about poorly executed chemical attacks that failed to make any impact.

Iraq was not completely deterred during Desert Storm. Scud missiles, albeit with conventional warheads, were launched against Saudi Arabia and Israel, oil wells were set on fire and oil pipelines were opened into the sea. The use of oil as an environmental weapon was unpleasant but, in the end, manageable. The Scud attacks were in themselves limited in their physical impact but psychologically they did considerable damage and required a variety of extraordinary exertions from the coalition. One response was to deploy Patriots for the purposes of missile defence. As with the Scuds they were supposed to stop, these also had a psychological effect, in this case calming, disproportionate to their physical achievements. The net result of this episode was to draw attention to the potential influence of small, and not necessarily very destructive, attacks against civilian populations and to pose the issues of active defence and/or deterrence through punishment in a new light.

Prior to the war the stress on Iraqi nuclear capabilities had been seen in a rather cynical light, a convenient rationale for a prospective military operation that was notably failing to generate popular enthusiasm. As already noted, post-war evidence of just how far the Iraqis had progressed raised some disturbing 'what if' questions. Would Iraq would have been in a position to move to regional pre-eminence through a combination of crude threats reinforced by occasional take-overs while the West would have become inclined to stay clear through fear of becoming involved in nuclear engagements? The fear of Iraqi chemical weapons reinforced opponents of military action in the run up to the Gulf War: imagine the clamour if American and British troops had been presented as prospective targets for nuclear weapons or if Riyadh and Tel Aviv had been promised incineration. Kuwait would have been left to its fate. As with all counter-factuals one difficulty was to judge whether this feature of the situation would have made itself felt before the question of American attitudes became an issue - for example with Israeli pre-emption or Saudi appeasement.

Even more worrisome than the possibility of a 'rogue' state was a terrorist group that was bent on inflicting mass casualties in support of an extreme political agenda, religious conviction of just some eccentric vendetta against the civilised world. Hints of what might be possible emerged out of some of the more notorious terrorist incidents of the 1990s: the first attack on the World Trade Center in 1993, the sarin gas attack on the Tokyo subway in 1995 and the bombing of US embassies in East Africa in 1998. Yet up to 11 September 2001 the worst that terrorists seemed able to accomplish with conventional explosives alone had been casualties in the low hundreds. Attacks of a more severe sort could be found in popular fiction and policy analysis. This 'superterrorism', relying on chemical and biological, or even nuclear, weapons, would cause casualties in the low thousands and upwards. Such events were not assigned a high probability. These did not appear to be the weapons of choice for terrorists and that they would have problems handling them.

The al-Qaeda group, led by Osama bin Laden, was not quite a non-state actor, because of the position it had gained as effectively part of the Taliban regime in Afghanistan. It had political demands, but of a nature that put them beyond accommodation. Moreover, by turning commercial airliners into guided missiles, it found a way of causing thousands of casualties through careful planning and training rather than technical prowess and access to noxious materials. It therefore challenged the categories within which security risks had been identified and understood, and pointed to the dangers arising out of weak states and fractured societies, as much as radical regimes which attempted to gain some artificial standing through attempts to gain construct nuclear devices. Yet it was this latter possibility that the second Bush Administration chose to focus upon in the aftermath of 11 September, so that the second stage of the 'war on terrorism', which had led at first into Afghanistan to deal with the Taliban and al-Qaeda, was widely expected to have as its second stage an attempt to change the Iraqi regime, and the two other members of the triumvirate (Iran and North Korea) named by Bush the 'axis of evil.' Iraq in particular was seen as a likely source of unconventional weapons to terrorist groups and that experience now demonstrated the wisdom of dealing with threats before they materialised rather than after. This push was greeted with a notable lack of enthusiasm among the US allies, not so much because of objections in principle as to doubts as to whether the US would be able to follow through on its objective and the likely reaction in the Islamic world.

It is by no means clear that Iraq is on the edge of a nuclear capability although absolutely clear that it would like to be. The same can be said with North Korea and Iran. Should controls over the Russian nuclear stockpile prove to be unreliable a surprising state - or even non-state - may come to wield nuclear weapons. Fortunately these contingencies are few but they each pose sharply a question that is also posed by many other lesser contingencies: the risks the major powers are prepared to accept in providing guarantees and support to the strategically disadvantaged. Where there is alliance (as with South Korea) the issue is in check. Where there is not, as in much of western and central Asia, further proliferation could be profoundly destabilising. In this context the no-first use issue seems to me to be largely irrelevant. The people who are most likely to use nuclear weapons first are the least likely to be interested in making the declarations, and the victims are most likely to be those without the capacity to use weapons second.