What the Bomb Prevents

Begin with the record, because it is stranger than it looks.

Two nuclear weapons were used in war, on Hiroshima and Nagasaki in August 1945, and none since. In the eight decades that followed, the number of states with the bomb rose to nine, arsenals reached tens of thousands of warheads, and the United States and the Soviet Union spent forty years preparing to fight each other. They never did. Whatever else is true about nuclear weapons, that absence needs explaining.

The first explanation was written almost at once. In 1946 Bernard Brodie, an American strategist, edited a short book called The Absolute Weapon and argued that the bomb had changed what armed forces were for. Until then their chief purpose had been to win wars. From now on, he wrote, it must be to avert them, and it could have almost no other useful purpose.

That is deterrence in its plainest form: a war that cannot be won will not be started, provided each side can still destroy the other afterwards.

The proviso turned out to be the hard part. In 1959 Albert Wohlstetter, working at the RAND Corporation, published an essay called The Delicate Balance of Terror, and its argument was that deterrence was not automatic. A force that could be destroyed on the ground by a surprise attack deterred nothing. Worse, it invited the attack, because in a crisis each side would have reason to strike first rather than lose its weapons. What deters is not having the bomb but having a bomb that survives.

Much of the expensive engineering of the Cold War follows from that sentence. Missiles went into hardened silos, bombers onto airborne alert, and from 1960 a growing share of the American force went to sea, in submarines nobody could find. The point of the submarine was not that it was powerful. It was that nobody could destroy it first, which made a first strike pointless.

Now the part the reassuring version leaves out.

In 1965 the political scientist Glenn Snyder noticed a consequence of all this stability. If both sides are confident that neither will risk a full nuclear exchange, then a smaller fight looks safer than it otherwise would, because each expects the other to stop before things get out of hand. Stability at the top makes instability lower down less dangerous to try. The idea is called the stability-instability paradox, and the record since has given it plenty of material.

China and the Soviet Union, both nuclear-armed, fought on the Ussuri river in 1969. India and Pakistan tested weapons in May 1998, and within a year Pakistani forces had crossed the line of control at Kargil and fought a limited war in the mountains. Indian and Chinese soldiers died in the Galwan valley in 2020. In May 2025, after the attack at Pahalgam, India struck targets in Pakistan and Pakistan struck back for four days before a ceasefire. None of these became a general war. All of them were fought by states that might have been more cautious without the bomb, or might not have fought at all.

Thomas Schelling, another strategist of the period, supplied the piece that explains how such crises are actually conducted. A threat to destroy the world is not believable, because nobody would carry it out. What is believable is a threat that leaves something to chance: a willingness to let a situation become dangerous enough that neither side is fully in control. Crises, on this view, are competitions in risk-taking, and the side that seems readier to tolerate the risk tends to get its way.

Put the three ideas together and the comfortable conclusion becomes harder to hold.

The bomb has very probably helped prevent a third world war between the great powers, and it is reasonable to count that among the largest facts of the century. It has not prevented war between nuclear-armed states. It has shaped those wars, kept them limited, and given each side a reason to believe the other will stop. That belief has held every time so far. It is a bet on mutual restraint under pressure, renewed in each crisis, rather than a law of nature, and the dangerous moments are precisely the crises in which the bet is being placed.

Brodie's sentence is still the best summary of what the weapon is for. The rest of this book is about what it costs to keep it unused: a treaty that tries to stop it spreading, the states that chose to give it up, and the nights when keeping it unused depended on one person's judgement.