Ten Dillemas of Nuclear Deterrence

We will investigate here what can be called the ‘Ten Dilemmas of Nuclear Deterrence’.

The presence of nuclear weapons between rival countries cannot guarantee the prevention of nuclear exchange or war. It is always a gamble, and it is a gamble that can always fail at some time in some place. The claim that nuclear deterrence ‘works’, at least to such an extent that you can safely rely upon it, is untenable. It is untenable because the conditions that must be established for assured safety to arise from the workings of nuclear deterrence are so restrictive that these are impossible to establish in practice and in the real world in which we live. This defect becomes clear when we understand what nuclear deterrence is – it is simply a psychological state of mind! It is a psychological state of mind that must exist in the minds of one’s nuclear opponents, i.e., among those who decide on whether or not to push the nuclear button. And it is a state of mind that you (who are relying on nuclear deterrence for your country’s security) must ensure will always exist in your opponent although you can never guarantee or ensure that it will always so hold.

To put it another way: nuclear deterrence is nothing but the irrational hope that a terrible fear of the consequences of nuclear war will continuously promote wise decisions by fallible human beings operating under intense pressure in changing circumstances that neither they nor you can fully control!

If to become more secure you need the threat power of nuclear weapons then you are trying to establish your security by making the other side vulnerable and more insecure because they are afraid of what your nuclear weapons can do. Similarly, the other side must try to make itself more secure by making you more vulnerable and insecure to its own nuclear arsenal. This situation is often called the security-insecurity paradox. Your own security is supposed to rest on making the other side more fearful and insecure so that it behaves the way you want it to. Then the other side seeks to overcome its insecurity to make itself more secure by promoting greater insecurity/vulnerability in its opponent. Thus we have what can also be called the ‘vulnerability paradox’. The way out of this would seem to lie in both sides accepting mutual vulnerability and a mutual ‘balance of terror’. This then should be a form of stability in the nuclear equation.

Unfortunately, no such stability arises and instead we have a constant and powerful spur to continuous nuclear arms racing. To understand why this is so we first need to understand what are the basic conditions that have to be met if there is to be nuclear security even by the logic of nuclear deterrence thinking. According to this ‘logic’ a country must have, what is called, ‘an assured second-strike capacity’. Between two countries having nuclear weapons there is always the possibility (and the temptation) that the country that uses its nuclear weapons in a properly targeted fashion first can hope or expect to finish off all or most of the opponent’s nuclear capacity to retaliate after such a first-strike. Therefore, a country must be able to ‘absorb’ a first-strike and have enough left over to devastate the opponent so that the opponent is not tempted to strike first. Developing more nuclear weapons in this way may seem to the side doing it a reasonable thing to do in order to protect itself. But to the other side this means your opponent is making itself stronger in the name of developing second-strike capacities but can also make an even more massive first-strike (regardless of whether or not it declares a policy of No First Use), and therefore must be a spur to the threatened side also developing more nuclear weapons in the name of strengthening and ensuring the survival of its own second-strike capacity.

A third dilemma can be called the ‘predictability issue’. Since deterrence is a state of mind in your opponent you are always tempted to try ways of making that opponent’s behaviour more predictable, i.e., in accordance with what you want. Rather than simply hope that fear of your nuclear arsenal will make the opponent behave the way you want it to, you try and replace hope with some form of compulsion. To demand a high degree of predictability about an opponent’s behaviour means a demand for repeated, regular, institutionalised predictability and symmetry by each nuclear player vis-à-vis the other. The reality is that this is not possible. But as long as one nuclear player or the other seeks to establish what it thinks can be the conditions in which such assured and predictable behaviour by the other side will be forthcoming, then this slides easily into a strategy of not just simple deterrence but aggressive compellence.

This strategy of compellence involves certain kinds of nuclear preparations and their associated political implications and signals whereby the other side is supposedly ‘compelled’ to follow the pattern set by, and more ‘controlled’ by the first nuclear player. Not only is nuclear arms escalation written into this ‘compellence’ script but nuclear tensions are even more heightened than otherwise. The ‘certainties’ of compellence are substituted for the ‘uncertain certainties’ of ‘normal’ deterrence. Something like this happened with the US pursuing such a compellence strategy as part of its efforts to ‘stabilise’ and ‘control’ to its own ‘advantage’ the nuclear arms race of the late seventies and eighties. It is also a script written into the BMD, which represents the US ambition nuclearly/militarily to dominate the world via domination of space itself. Compellence is a more aggressive form of seeking ‘political advantage’ through nuclear weapons and there is no natural firebreak between the dangerous instability of deterrence-based nuclear behaviour and the more dangerous instability of compellence-based nuclear behaviour.

It is not enough just to have nuclear weapons or to claim that they are never going to be used and that their purpose is only to prevent a nuclear conflict. Your nuclear threat must be credible. Indeed, an opponent will not be deterred if it believes that the deterrer will never use his nuclear weapons. Thus the ‘capability’ and the ‘will’ of the deterring country must not be doubted. The ‘enemy’ should be convinced that its nuclear opponent will use nuclear weapons if pushed to the brink, or at least uncertain whether or not they will be used, but never certain or confident that they will not be used. However, any second-use of nuclear weapons is not, and cannot be, an act of security retrieval or enhancement. Once an opponent has launched a first-strike against you then your security has gone. By using your nuclear weapons second, all you are doing is engaging in an act of revenge – to make the other country suffer too. This is also a senseless act of revenge because it only sets off a chain of further launches and counter-launches, which further devastates both countries. This being so, the side using nuclear weapons first can entertain ideas that a second-strike is not credible.

Thus the question of ‘credibility’ or willingness to use nuclear weapons becomes very important. So governments that have nuclear weapons despite occasional pronouncements that these will never be used do not actually want the public or other governments to seriously believe them. This would undermine that country’s credibility. In the face of all kinds of challenges and uncertainties – technological, military, etc. – the capability to use nuclear weapons must be constantly updated and fine-tuned and available for showing in a variety of circumstances. Also, the political will to use nuclear weapons (presumably in the last resort) must be periodically displayed. Thus the requirements of making one’s nuclear deterrent credible creates powerful pressure for the generation and sustenance of both an enduring politics of nuclear-related hostility – including nuclear brandishing and brinkmanship – and of arms racing between rivals.

Nuclear perspectives/behaviour are the prisoner of, and subordinated to, the more fundamental and overarching framework and context of political hostility between nuclear-political rivals. Nuclearisation/militarisation are themselves the symptoms or results or expressions of this prior hostility. They are not their primary causes and therefore cannot be the solutions undoing this hostility. Indeed, they exacerbate such tensions/hostility because nuclearisation is itself the announcement that one is willing to inflict the utmost devastation to the opponent country and society. It was not the supposed deterrent qualities of the US-Soviet nuclear stand-off that eroded political hostilities but the other way around. It was the Gorbachev initiated unraveling of Cold War politics that paved the way for the erosion of nuclear tensions. The same principle of the prior importance of politics applies to the India-Pakistan situation. And nuclear weapons, including the presumed powers of nuclear deterrence, operate within this wider and more determining political context. The rhythms of deterrence behaviour are subordinated to the more powerful rhythms of political behaviour between mutually hostile countries.

This gives rise to a classic paradox in the search for stable deterrence: the conditions that are thought to make it necessary to apply deterrence guarantee that it will not be stable. The extent to which deterrence is genuinely stable is the extent to which it is unnecessary! India-Pakistan cannot have a stable deterrence equation nor did the US and the Soviet Union. But Britain and France can have a stable deterrent equation with each other precisely because it is unnecessary for them to have such a nuclear equation.

Constant technology advances in the development of nuclear warheads and in the designing of the range and accuracy of delivery systems is also a major input in ensuring the degenerative logic of deterrence-based thinking and behaviour. The more inaccurate and relatively invulnerable or undetectable the nuclear missiles are the more ‘stabilising’ they are. That is to say, if the missiles are not very accurate then they can be used to attack cities but not specific military targets. This makes them les useful as first-strike weapons aiming to, or capable of, knocking out an opponent’s military installations (including its nuclear missile installations and air bases) and more useful as second-strike weapons able to devastate cities. In the jargon this is called the difference between counter-value targets (e.g. cities) and counter-force ones (e.g. military/nuclear infrastructure). Moreover, the best second-strike nuclear weapons are considered to be those that combine low accuracy with relative invulnerability to a first strike. Thus submarine-based nuclear-tipped missiles are seen as the best guarantor of second-strike capacity.

What happens, however, when land missile systems become more mobile (always being moved around on rail or road systems rather than being stationed in fixed silos), when there is 24 hour rotation of airplanes carrying nuclear weapons, and when submarine launched missiles become ever more accurate (which is happening), and when efforts at detection of submarines and anti-submarine warfare make steady technical advances? What all this means is that technology advances are themselves undermining the distinction between ‘stabilising’ second-strike weapons and ‘de-stabilising’ first-strike weapons since the former can increasingly double up to do the job of the latter. In short, this undermines hopes of stabilizing the nuclear equation. What is more, new technological breakthroughs, like a developed Star Wars project of the kind envisaged by the US, are profoundly destabilizing. One would need a separate essay to explain this complexity fully. In short, because technology does not stand still neither does nuclear arms development. This is not just a matter of replacing the old with the new but also of creating new problems, difficulties and dangers.

The fear of a first-strike leads to the search for what are called ‘survivability enhancing practices’. These are measures aiming to ensure that one’s second-strike capacity even after a first-strike is relatively unimpaired. But these very practices themselves undermine the stability of the deterrence equation. The two most important such measures are a) to disperse one’s nuclear arsenal as widely as possible; and b) to adopt what is in effect a ‘launch-on-warning’ posture for one’s missiles. Both give rise to grave problems. The first involves a centralization-decentralization dilemma. It is not enough to disperse the locations of nuclear delivery systems. There is also the problem of decentralizing command and control over such systems. This is because there is always the danger of what is called a ‘decapitating first-strike’. That is to say, an opponent can not only strike first, but seek to decapitate the command and control system of the opponent, i.e., finish off the key decision-makers who are the apex of the chain of command over the nuclear arsenal. Even the establishing of ‘redundant’ or multiple chains of command to become operative in wartime may not prove adequate to the impact of an effective decapitating first-strike. To avoid this dilemma one has to greatly decentralize command and control to junior levels and more localized personnel at the much lower rungs of the chain of command so they can carry out a second-strike. But any such decentralization greatly enhances the possibility of an accidental launch or a miscalculation (especially in wartime situations) that leads to a launch on the presumption that an enemy nuclear attack is taking place or about to take place. We now know from recently disclosed official documents how close a Russian submarine was to launching a nuclear attack on a US ship during the Cuban Missile crisis on the basis of just such de-centralised authority because it thought that ship was torpedoing its submarine. [This event is now to be enshrined in a Hollywood movie.]. Pakistan, as a much smaller country fearful of a possible decapitating first-strike by India, will have to face a particularly acute dilemma of centralization-decentralization. But the problem is acute enough for India as well.

As for adopting a ‘launch-on warning’ posture to ensure second-strike capacity, this was the dominant form taken by the stationing of land-based missiles of both the USSR and the US during most of the Cold War period. What this means, of course, is that there is an inescapable trade-off between the requirements of nuclear safety and guarding against the risks of a launch by accident or miscalculation, and the requirements of deterrence efficacy. During the Cold War period (even afterwards) there were various occasions when there were false alarms and in some of those cases, matters came close to a head with the near launch by one side of its missiles. [The best study of this is ‘The Logic of Accidental Nuclear War’ by Bruce Blair, Brookings Institute, 1993.]. But another rarely noticed point is that launch-on-warning raised a fundamental question about nuclear deterrence. This question is not about whether or not deterrence actually worked but whether it even existed!

A launch-on-warning posture meant that the US had less than 25 minutes before a Soviet missile could hit it and therefore a maximum of only that much time to decide whether the alarm was a false one or not. Since the US had submarines much more closely positioned to the Soviet mainland, the latter had only around 10 minutes before there would be a hit, so that much less time to decide on its response. In the US case from the time the alarm is raised about a possible Soviet launch, 10-12 minutes would have to elapse for the missile to be identified, its path tracked and the necessary information relayed to top command. Another 2-3 minutes would elapse before this could be communicated to the President. Any decision (whether this was a false alarm or a real one needing the order to fire off) taken by the President if it was to be communicated to all necessary stations so as to be carried out would require another 8-10 minutes. In short, out of the roughly 25 minutes in which a decision by the President has to be made — even if the President was in direct telephonic communication with key aides — he has literally only one or two minutes in which to take a truly momentous decision. When the space and time allowed for human decision is so shortened by adopting the posture of launch-on-warning, what then is one to make of the claim that deterrence is in operation when there can never be any foolproof check on an accidental launch by an opponent or false alarm by one’s own system?

In the case of India, the best survivability practice would be for both sides to move towards a launch-on-warning posture, which is something they may well do in the future. But the missile flight time between the two countries at 5-8 minutes is so much shorter than even in the Cold War case for the US. There is no way there can be even the illusion of being able to maintain a proper check on preventing launches by accident or miscalculation. The trade-off between wanting greater nuclear safety and deterrence efficacy in South Asia is an even starker one than was the case between the great Cold War rivals.

According to the logic of nuclear deterrence thinking, it is not enough to have a second-strike capacity that survives a possible first-strike. This capacity must be able to inflict what is called “unacceptable damage” on the opponent. If after a first-strike you only have a few weapons left over then your opponent may be prepared to take the risk of a first-strike and ‘absorb’ the second-strike. But what is “unacceptable damage”? And how do you ensure that you have it after a devastating first-strike? The simple answer is that the concept is inescapably vague and impossible to quantify and there is no assurance that you can retain the ability to inflict such “unacceptable damage” after an enemy first-strike on you or that the damage you might be able to inflict would be unacceptable to the opponent. All that happens is that both sides have to embark on the escalator of making more and better nuclear weapons in the futile search for achieving such a second-strike capacity. George Lee Butler, who for 12 years headed the US’s Strategic Air Command (which is the service that has overall control of the US nuclear arsenal) and who between 1992-5 was the key Presidential adviser (the one person the President must consult before pressing the nuclear button) and subsequently turned nuclear disarmer, said quite correctly in regard to the “unacceptable damage” issue that this was impossible to quantify and impossible to operationalize. The US ended up targeting over 16,000 locations in the USSR with its ready delivery systems but could still never be sure that they could have an ‘adequate’ second-strike capacity to cause “unacceptable damage”. Butler revealed that he was himself so shaken, when he took over supreme command, by the revelations of the insane logic that was operating in US nuclear preparations in the name of deterrence efficacy that he began to systematically question the basic assumptions of such thinking and the security paradigm based on such thinking.

Seeking security through nuclear weapons and the supposedly wondrous powers of deterrence is itself reflective of a particular political approach to matters of security for a country. This approach has a name – it is called Realism or Realist thinking in international relations. This approach gives prime importance in politics to military power (even more than to economic power) and believes that great military power provides the leverages for getting great political advantages. It is this kind of thinking that lies behind the belief in the efficacy of nuclear weapons through its ‘threat power’. Given this mental-intellectual predilection, invariably the justifications for the importance of nuclear weapons go beyond the simple claim that it provides through deterrence security against a nuclear attack. Since nuclear weapons are the supreme form of military threat power, and since military threat power is considered to provide so many other potential political advantages, even greater value is attributed to nuclear arsenals and it comes to be seen as the means for pursuing many more political goals.

Nuclear weapons are then seen as not just preventing nuclear war but as helping to achieve crisis-stability; to prevent even conventional war; to provide general foreign policy support; to provide global status and prestige; to help secure arms control; and even to secure eventual nuclear disarmament. Now if one thing – nuclear weapons – can help give us all these advantages then how on earth can one oppose having them? Alternatively, one can see this ridiculous expansion of the qualities attributed to having nuclear weapons as reflecting a profound inability to think through the complexities of the notion of power, the inability to perceive its different forms and the specific limitations of each form. What we have instead is a set of Godworshippers where nuclear weapons have become the new God that can magically deliver all kinds of political goodies. This unwarranted expansion of the ‘value’ of nuclear weapons then makes it even more difficult to want to, or work to, get rid of them.

Finally, there is the ‘What if?’ question! What if nuclear deterrence breaks down and there is a nuclear exchange? In fact, those involved in running a nuclear weapons system, for all their publicly pronounced assurances, know that neither they nor anyone else can ever guarantee that someone somewhere, including their enemies, will not use such weapons. All countries having nuclear weapons must think about, and have some plans for, what to do if deterrence breaks down. That is to say, they have to make some kind of preparations to actually fight a nuclear war. This itself mocks their claim that since one has nuclear weapons there is less danger of having a nuclear exchange or war. India, for all its pretentious claims about the efficacy of deterrence after Pokharan II, had to subsequently launch ‘Operation Parakram’ with Gen. Padmanabhan, the first chief of army staff of a nuclear India, publicly declaring that these were exercises to help equip the Indian forces to fight in actual nuclear warfare conditions. But these very warfighting preparations themselves undermine confidence in deterrence postures and the claim that one can rely on the efficacy of deterrence.

At the end of it all, one can still imagine Indian nuclear hawks resorting to the counterfactual and saying, `Look, deterrence works because after 1945 there has still been no nuclear exchange between nuclear weapons powers.’ This is not a serious form of reassurance. Indeed, no one who has seriously studied the record of the nuclear age since 1945, would ever try to pass off this absence of nuclear war as a source of reassurance. There have been some very close misses with the Cuban Missile crisis bringing the world to within a hair’s breadth of a nuclear holocaust. On a number of occasions there were key people advocating the use of such weapons but fortunately each time those who opposed them got the better of the decision, if not always the argument. There is never any guarantee that such a situation will always prevail in the future. Indeed, some of the key former believers in the efficacy of nuclear deterrence have turned against their earlier beliefs, and many like George Lee Butler have said that it was only “by the grace of God” that a nuclear disaster was averted during the Cold War era. Today, the danger of a nuclear conflict has taken a more regional turn. If the chances of a global nuclear holocaust have receded due to the end of the Cold War, the chances of a ‘regional’ nuclear exchange (with South Asia the current front runner for fulfilling this possibility) have become greater.

Moreover, the test for whether nuclear weapons bring about greater security is not established simply by asking whether or not deterrence has broken down completely through the eruption of a nuclear war. There is another more routine test question for assessing whether or not nuclear deterrence works. And that question is simply whether the countries that have such weapons are or feel more nuclearly secure from their rivals or enemies, actual or potential! Has the world become more nuclearly safe over all these decades? If nuclear deterrence is efficacious then their spread to more countries should logically make the world safer. If it works to make India and Pakistan more secure then surely the same logic applies to other countries striving to make themselves more nuclearly secure? Does anybody in his or her right mind think this is actually the case? That their spread to Iran, Iraq (Israel already has them) and elsewhere, or their possession by non-state actors, will mean greater safety?

Why is it so difficult for people, especially our bomb-supporters, to see what is staring them in the face – that the world is in a deep nuclear mess? The single biggest culprit for helping to create this mess is the mind-set of those who believe in the efficacy of nuclear deterrence. This is the mind-set that must be delegitimised, undermined and exposed for the bankrupt politics that it leads to, if we are ever to secure a truly nuclear safe world. This can only be a nuclear free world.