Tony Schwartz, ghostwriter of Donald Trump’s book The Art of the Deal, recently issued a dire warning about the man he worked for, declaring, “I genuinely believe that if Trump wins and gets the nuclear codes there is an excellent possibility it will lead to the end of civilisation.”

Whether Schwartz is exaggerating the danger, talk of the risk of nuclear war seems to be more common these days. With tensions rising with China and Russia, it’s a good time to revisit a grim topic that many wish had vanished with the end of the Cold War.

First, the good news, at least at some level: Humanity is no longer in danger of extinction. Nuclear stockpiles in the United States and Russia have declined a lot since the end of the Cold War. In the late 1980s, the two superpowers had about 64,000 nukes between them. Now, it’s down to less than 10,000. Of those weapons, however, only a subset are actively deployed and can thus be used at a moment’s notice.

So, if for some awful reason there were a full nuclear exchange between the US and Russia, it would involve at the most about 3,000 nukes. In reality, some of those would probably be destroyed on the ground by “counterforce” strikes against submarines and missile silos, a few would be taken out by missile defences and some would simply fail to explode because of decayed trigger mechanisms and fissile cores. That level of nuclear destruction, though, would be enough to wipe the US and Russia off the map, cause global crop failures, send worldwide cancer rates soaring and kill hundreds of millions — or billions — of people. Yet, horrible though that would be, we are no longer staring into the teeth of a world-ending apocalypse.

Of course, that’s cold comfort to those in the US or Russia who would die in the event of nuclear war. Leaders in both countries always must think very carefully about how to minimise the chances of it happening.

Here, economic theory may be able to help. That’s because nuclear warfare was one of the first applications of game theory — mathematical models of how parties interact strategically. One of the ideas to emerge from this was the concept of mutually-assured destruction — in particular, the importance of second-strike capability — in which both sides understand that they’ll lose in a conflict. In other words, if initiating a nuclear attack guarantees your country’s destruction, there’s never an incentive to start a war. As the military computer in the 1983 movie WarGames so aptly put it, the only winning move is not to play.

Paradoxically, though, this means there’s a danger to having too few nukes. If stockpiles decline to the point that one side thinks it can wipe out the other with a first strike and suffer an acceptable level of retaliatory damage, the situation becomes unstable. Each side has an incentive to risk an attack. Improvements in the US ability to track nuclear-armed submarines contribute to this danger, since that raises the tantalising possibility of destroying the enemy’s second-strike capability in a surprise assault.

Now, what about the danger of an unsteady hand on the nuclear button? One part of the game theory of nuclear war is the so-called madman theory. This is the idea that if the leader of a nuclear-equipped country acts a little crazy, the opposing superpowers will be reluctant to do anything provocative that might tip him over into irrationally launching a total war. This idea, which makes sense in some models of the nuclear game, induced US leaders such as Richard Nixon to play up their own aggressiveness during the Cold War in order to deter Soviet conventional attacks.

But in a world of shrunken nuclear stockpiles, a borderline madman could be much more dangerous. When the temptation of a potential first-strike victory is present, an unstable president may not need a severe provocation in order to push the big red button. Instead of simply deterring the enemy, an erratic president may actually just decide to throw the dice and try to eliminate America’s enemies — with disastrous results for both the US and the world as a whole.

So, if the danger in the 1980s was that the world had far too many nuclear weapons, the danger in the 21st century may be that the US and Russia have too few. Although US President Barack Obama’s recent speech in Hiroshima, calling for the eventual abolition of nuclear weapons, makes for good optics, actually pursuing that policy may remove the strongest protection against nuclear war. Meanwhile, shrunken nuclear stockpiles should make us a little more fearful of ever electing a president who may think that launching a first strike will be anything less than madness.

— Washington Post

Noah Smith is an assistant professor of finance at Stony Brook University and a freelance writer for finance and business publications.