9 March 2015

Deterrence, without nuclear winter

Seth Baum

Seth Baum

Seth Baum is executive director of the Global Catastrophic Risk Institute, a nonprofit think tank that Baum co-founded in...


The biggest danger posed by today’s large nuclear arsenals is nuclear winter. One or two nuclear strikes could wreak devastating destruction on a few regions, but would not destroy human civilization as a whole. The roughly 16,300 nuclear weapons that currently exist, though, are more than enough to cause nuclear winter, which, through extreme cold conditions, ultraviolet radiation, and crop failures, could threaten the whole of humanity. If we fail to avoid nuclear winter, we could all die, or we could see civilization collapse, never to return.

That makes avoiding nuclear winter paramount. But the world’s major powers, in particular the United States and Russia, have long argued that their large nuclear arsenals are required for deterrence. Deterrence means threatening another party with some sort of harm in order to persuade it not to do something. In this case, it means threatening massive nuclear retaliation to dissuade another country from launching an attack itself. If two countries were to follow through on their threats of nuclear retaliation, mutual destruction would be assured. That deters both sides from starting a war. But nuclear deterrence can fail, as demonstrated during events like the Cuban missile crisis, when there have been escalations towards nuclear war. (Martin Hellman, Ward Wilson, and others have documented such events.)

As things stand now, a failure of deterrence could result in nuclear winter. It may be possible, though, for the world’s biggest nuclear powers to meet their deterrence needs without keeping the large nuclear arsenals they maintain today. They could practice a winter-safe deterrence, which would rely on weapons that pose no significant risk of nuclear winter.

Since the 1980s, the standard policy stance of those concerned about nuclear winter has been to call for unconditional disarmament—without such requirements as having alternative weapons in place or achieving improved relations with other nuclear-armed states. And there have been some disarmament successes. The United States and Russia have significantly reduced their nuclear arsenals. The worldwide initiative to ban nuclear weapons based on their humanitarian impact has also done important work in recent years, benefiting nuclear and non-nuclear states alike.

But calls for unconditional nuclear disarmament can only get so far. Nuclear-armed countries are unlikely to make drastic reductions in their arsenals if doing so would compromise their perceived national security interests, in particular their ability to deter adversaries. In order to avoid nuclear winter, we need to work from the perspective of those who actually have the weapons. We must seek solutions that they will actually want to implement.

Drastic cuts required. How significantly do arsenals need to be cut to avoid nuclear winter? They don’t need to be reduced to zero, but the short answer is that nobody knows. Humanity’s ability to survive massive nuclear attacks is not well-studied. One recent study found that two billion people—nearly a third of humanity—could be at risk of starvation from a nuclear war involving just 100 weapons. The actual death toll could be even larger, because this study only looked at starvation. Secondary effects such as pandemics and violence could kill even more.

Based on current knowledge, then, I propose that the worldwide total of allowable nuclear weapons should be no more than 50. There is no guarantee that human civilization would survive a 50-weapon war, but the risk of nuclear winter would be very low.

Getting from today’s 16,300 nuclear weapons to 50 would require drastic cuts in the arsenals of every nuclear weapon state except North Korea, which is believed to have fewer than 10 nuclear weapons and may not have any. The eight major nuclear weapon states will not want to simply abandon their arsenals, though, without some ability to meet the same perceived need that the weapons fulfill.

That’s where winter-safe deterrence comes in. Deterrence doesn’t have to rely on large arsenals of nuclear weapons, or even on nuclear weapons at all. A range of candidate weapons could conceivably achieve the same goal without risking global catastrophe.

Alternatives to nuclear weapons. The ideal weapon would have a suite of attractive properties. It would not pose a significant proliferation risk. It would be affordable, technologically feasible, and politically acceptable. It would not significantly shift geopolitical power or destabilize the international system. And it could be used as a retaliatory second-strike weapon, which is crucial for deterrence.

Finding such a weapon, or system of weapons, is a tall order, though some arms are easy to rule out. Cyber weapons are bad deterrents because they work mainly in surprise attacks, and one cannot be deterred by weapons of which one is not aware. Contagious biological weapons pose as large a global catastrophic risk as large nuclear arsenals, so would not make the world safer.

Other weapons show more promise. For example, conventional prompt global strike systems now in development, which can reach anywhere on earth in as little as an hour, can be used to threaten small targets with much less total destruction than nuclear weapons. The United States is at the forefront of developing these weapons for this purpose, and Russia has also indicated interest. But the role of conventional prompt global strike and other conventional weapons will remain limited because they only threaten small targets. This compact focus is great insofar as it means less total destruction, but it can be bad for deterrence, which often leans on much larger threats.

Exploring “good” options for threatening large destruction is a peculiar and regrettable task in which no civilized person should take any joy. But if doing so can save many lives, and indeed save civilization itself, then it should be done. The two weapons that stand out are non-contagious biological weapons and nuclear electromagnetic pulse. The former could work well if deterrence requires threatening large human populations. The latter could work well when deterrence requires threatening large amounts of infrastructure. In both cases these are tentative conclusions, backed by only by my limited, preliminary study. Governments should not move forward with either weapon without more careful examination.

What is clear is that winter-safe deterrence appears feasible. This is excellent news. It gives some hope that the risk of nuclear winter can be minimized even while countries feel the need to deter each other. Ultimately, we should seek a world at peace, in which deterrence isn’t seen as necessary. While that world is still a work in progress, though, we should keep the world we live in now safe from nuclear winter and other global catastrophes. If we fail at this, the rest will be for nothing.

Editor's note: The author’s paper on the issues addressed here, “Winter-Safe Deterrence: The Risk of Nuclear Winter and Its Challenge to Deterrence,” will be published in a forthcoming issue of Contemporary Security Policy.