If a war between China and the United States were to go nuclear, who would win?

In the scenario of the exercise, it is the year 2032 and a war has been going on over Taiwan for 45 days. China uses “theater” nuclear weapons (with shorter range and lower yield than city-destroying “strategic” missiles) to shorten the war by coercing the United States into submission. The targets include Guam and Kwajalein Atoll, a pair of islands vital to the US military posture in the Pacific, as well as a US aircraft carrier strike group.

This is distressingly plausible. One reason is the geography of the Asian battlefield. During the Cold War, America and the Soviet Union planned to use many tactical nuclear weapons to destroy large and dispersed troop formations, often in the vicinity of towns and cities. “Today in the Pacific,” the study notes, “naval ships at sea and military air bases on small islands are a very different target.” Fewer nuclear bombs would be required and there would be less civilian damage than in Cold War strikes.

This is related to a second reason: the evolution of weaponry. Most people, unreasonably, think that conventional weapons are less escalatory and therefore more usable than nuclear weapons. But today’s low-yield nuclear bombs (20 kilotons of explosive power, about the size of Hiroshima) can be delivered with extreme precision and less collateral damage. “The line between low-yield tactical nuclear weapons and precision-guided conventional weapons in terms of both their operational effects and perceived impact is blurring,” says CNAS.

The third factor is the effect of a long war. Weeks into a conflict, both sides would run out of conventional weapons. Theater nukes would become more attractive. “On a weapon-by-weapon basis,” the authors note, “nuclear weapons are more efficient at destroying large-area targets.” Their immense power means they would still work even if weeks of war had degraded the command, control and intelligence systems on which conventional munitions depend.

The result of all this, in the war game, was a strange kind of nuclear war: China was incentivized to use nuclear weapons first, despite its formal “No First Use” pledge, but once it did , and in contrast to expectations of how a war between the United States and the Soviet Union in Europe would have played out, things have not necessarily turned into an apocalyptic exchange of strategic nuclear weapons, that’s what counts as to good news

The exercises suggested China had more reason to cheer. Experts and officials playing as China had a wide range of military targets: Asia is full of US naval facilities and assets. (Although there is little evidence that China has low-yield nuclear weapons at this point.) The US team, on the other hand, struggled with the fact that many of the most attractive targets for retaliation were in mainland China. Attacking those with tactical nuclear weapons would carry a much greater risk of escalation to a general nuclear war.

In addition, players discovered that the United States lacked the weapons needed to hit the “very small number” of lower-risk targets, notably Chinese warships and bases on disputed reefs in the South China Sea. Its most advanced non-nuclear missiles would have been phased out by 45. America, unlike Russia, no longer has a nuclear-tipped anti-ship missile A new submarine-launched nuclear cruise missile is planned for the 2030s, but it could not be used to deter Chinese nuclear use before the fact without giving away where it was it would also tie up the scarce attack submarines in the middle of a naval war.

Nuclear strategy has its own macabre grammar, steeped in the assumptions and experience of the Cold War and reshaped by the march of military technology. However, it comes down to politics. Faced with the nuclear annihilation of 5,000 US sailors on an aircraft carrier or a nuclear attack on US territory like Guam, a US president would respond with nuclear force, achieve what would be an ever-shrinking quiver of conventional weapons , or would you fold? This, the authors acknowledge, is “the fundamental and unknowable component”.

© 2024, The Economist Newspaper Limited. All rights reserved. From The Economist, published under license. Original content can be found at www.economist.com