Will the threat of Mutually Assured Destruction continue to prevent nuclear war?

The Mutually Assured Destruction (MAD) thesis is the argument that two states armed with nuclear weapons are unlikely to engage in a nuclear war, on the basis that it would be so destructive for both countries. In practice this also makes it less likely that these states would engage each other in a conventional war, as it risks nuclear escalation. There is clearly some truth to this. The balance of probability, in my view, is that some form of conventional war would have taken place during the 1945-90 Cold ‘War’ had the main actors not been nuclear armed. There might have been a war between the American led anti-communist block and the USSR, or communist China, or indeed a more intense conflict between the USSR and communist China (as opposed to the relatively brief border war which they actually fought).  But whilst, because the stakes are so high, nuclear armed adversaries are less likely to go to war, this doesn’t make it impossible nor, given enough time, necessarily improbable. MAD is only partially correct. Its behaviour is like that of an old and fat cat to a rat infestation. It makes it less likely, but by no means impossible. There are I think three primary scenarios in which MAD fails – making a nuclear attack from one nuclear armed power on another plausible. These are:

1. Insanity/hate: MAD works on the assumption that those with access to nuclear weapons behave rationally, and care about preventing the destruction of their supporters at least as much as they want to destroy their opponents. A leader who hates his enemies more than he loves his own people, and history has thrown up plenty such examples, may decide to deploy nuclear weapons even if likely to invite a nuclear reply. This is particularly plausible in the case of a power-crazed dictator, confronted with either a popular uprising or foreign invasion. If Hitler had been able to deploy nuclear weapons from his bunker in 1945, even if his enemies had been similarly armed, my guess is he would have done so. We should also recognise that world leaders do occasionally go mad, by any reasonable definition of the term. If they have access to nuclear weapons, and there aren’t sufficient checks-and-balances to stop them being deployed (most likely in an autocratic dictatorship) it’s possible that they might decide to use them.  
2. Belief in an afterlife: MAD works on the assumption that those with access to nuclear weapons care more about this life than any potential ‘afterlife’. For a decent proportion of humanity this is simply not the case, and of this a tiny proportion believe that destroying their enemies facilitates a pleasant afterlife.  If you believe that killing large numbers of your Gods ‘enemies’ makes your assent to heaven more likely, and that it either doesn’t matter or is a bonus if your killed in the process, then mutually assured destruction is no longer an irrational act. Almost anything becomes rational if it guarantees you an eternity in paradise (indeed beyond getting into paradise this life becomes almost irrelevant), and if large numbers of your supporters die as well that’s not necessarily a bad thing – you may have arranged their transition to heaven as well. The world’s two biggest religions by number of adherents, Christianity and Islam, have both produced bodies of believers who believe that killing, and dying fighting, the enemies of God guarantees entry into Heaven (for example Christian Crusaders and a variety of fundamentalist Islamic groups). This could well lead to groups or individuals who would launch a nuclear attack if they could, in the full knowledge that it would lead to MAD. There are a number of, generally rather small, Islamic fundamentalist groups who would probably deploy a nuclear weapon if they could, even if it were to invite a nuclear reply. Groups like ISIS have produced large numbers of suicide bombers prepared to die in order to harm their enemies, and it’s likely they would be prepared to deploy nuclear weapons using the same rational.
3. Disparity of destructive impact: Some countries are much bigger, and much more densely populated, than others. As a result it’s possible to imagine a scenario in which, should a nuclear war occur, one combatant would suffer significantly more than the other. In practice this is only likely to occur if the countries concerned only have a handful of nuclear weapons, say less than ten, and most currently nuclear armed states have considerable or vastly more (the exception being North Korea which is estimated to have around eight). However if additional countries develop nuclear weapons they may do so in small numbers, activating this scenario. Iran is around 1,648,195 km2 in size. Israel by contrast only has 22,072 km2. If both countries only had a few nuclear weapons you can just about imagine a scenario in which Iran decides to launch, on the basis that it will suffer a lot less than Israel (though this may be complicated by having one of the holiest shrines for Islam in Jerusalem).

What should we conclude from the obvious flaws to mutually assured destruction? That, whilst it certainly makes nuclear war less likely than conventional war, it far from prevents it. There are two main, somewhat interlinked, conclusions. Firstly it emphasises the importance of nuclear non-proliferation. Broadly speaking the more countries have nuclear weapons, the more likely one of the scenarios when MAD fails will come to pass, leading to nuclear war. It’s worth noting that the failure to stop one country developing nuclear weapons risks several countries developing them. Pakistan decided it needed nuclear weapons, which it eventually got, after the first Indian bomb was tested in 1974. This is one of the reasons why it was so important to stop Iran acquiring a nuclear bomb. If Iran got the bomb it’s likely that at least one of the Sunni Muslim states of the Middle East would have attempted to reciprocate. The most likely candidates would be Saudi Arabia, Egypt, Turkey and maybe even the UAE (whose rulers seem to like big shiny things which convey status). And to be blunt a Middle East with multiple nuclear armed powers would be a powder-keg.

The second conclusion is that it matters enormously what sort of political systems nuclear armed countries have. It is especially important to keep nuclear weapons out of the hands countries which are led by, or are likely to be led by, Government which are likely to trigger one of the three scenarios outlined above. Broadly this means regimes which are dictatorial/autocratic, allow few checks on executive power and/or are led by those who value an ‘afterlife’ over this life, and think violence can help them secure it. I would say that, because nuclear weapons are so uniquely awful, it would almost certainly be better to live in a world without them. I say that recognising that this would make conventional warfare more likely. But this isn’t an ideal world, and nuclear weapons are here to stay. As such it’s crucial that nuclear arms are maintained by a small number of liberal-democratic states, to avert a world dominated by nuclear armed tyrants. At present five states which could reasonably be described as liberal-democratic are nuclear armed (America, Britain, France, India and Israel). I would say three or four is the ideal number. It certainly needs to be more than one – the rise of Trump in America has shown how potentially unstable even the most advanced democracies can be, and we’d be foolish to put all our eggs in one basket.

The counter-point to this is that, as I mentioned earlier, it’s of vital importance to limit the proliferation of nuclear weapons amongst unstable/authoritarian/theocratic states, if need by via the use of force. To put it bluntly it’s my view that if nuclear proliferation expands significantly beyond the current nine nuclear armed powers, then nuclear war becomes all but inevitable. And a nuclear war would be so horrific that it’s worth almost any price, including a bloody conventional war, to avoid this danger. No one knows exactly how many people were killed when American planes dropped nuclear bombs on Hiroshima and Nagasaki in August 1945, though a total initial death toll of 120,000 is almost certainly on the conservative side. The nuclear bombs advanced nations currently possess are significantly more powerful than those deployed in 1945. I’m very glad the Iranian Government came to its senses, and agreed to a reasonable settlement with some of the main world powers in July 2015 over its nuclear programme. But if it hadn’t agreed I think there would have been a strong moral case to fight a conventional war with Iran, to prevent nuclear proliferation amongst autocratic, unstable and theocratic regimes. In short, if we allow significant nuclear proliferation, we will probably at some point have to face nuclear war.