Nuclear War Games

Over the last few months, the New York Times has run articles on nuclear weapons and nuclear war. The latest is by William Langewiesche, about a classified war game in which nuclear war escalated much more quickly than the participants, including the Secretary of Defense, expected.

I am not a fan of Langewiesche’s writing, so I’ll divide my review into two parts: Substance and Style. This post will cover Substance.

It should not be surprising that nuclear war escalated quickly. Nuclear weapons do not allow for subtle strategy. There are two options:

1. Hit the other side first with surprise. Take out their nuclear capability completely and as much of the conventional as possible.
2. Work nuclear weapons into a conventional war in some way. The bar might be the other side’s use of chemical or biological weapons, or it might be existential danger to the country.

Policy discussions oscillate between the two. The problem with the second is that there is no natural breakpoint at which escalation can stop. The problem with both of them is that the end is the same: full-up nuclear war with all its destruction.

John Foster Dulles’s massive retaliation (option 1) of the early 1950s was countered with the development of an enormous variety of buried nuclear mines, the Davy Crockett artillery shell, and of course air-delivered bombs that might be used to fight a nuclear war that might last more than a half-hour. Then the question of escalation arose, and the idea of strewing nuclear weapons all across Europe seemed unwise. In the 1960s, Robert McNamara was horrified by the numbers of nuclear weapons, some with megaton yields, aimed at single targets, over 600 at Moscow, for example. He tried to find alternatives short of general war. Option 1 continued through the 1980s.

Part of that was the military’s preferences. Some wanted to believe that if the Soviets had one person alive at the end, and we had two, we had won.

Nuclear-armed submarines provided a second-strike force that threatened destruction no matter what the first strike was like and stabilized the situation of option 1 as much as it could be stabilized. Attacks out of the blue became less likely. That threat ended when the Soviet Union did. Now we again have folks musing that maybe we could use a little nuke here or there in a regular war. Vladimir Putin has implied such things, and some Americans would like to test those theories. When they see our nuclear strength illustrated by a nuclear strike, they will back off!

Langewiesche describes a wargame played over several days in 1983 that included Casper Weiberger, Secretary of Defense at the time. The war starts with a chemical-biological attack in Germany by the Soviets and a nuclear response by NATO. The nuclear response is with the small nonstrategic nuclear weapons that are believed to be able to control escalation. But the Soviets escalate slightly, NATO responds, and things get out of control, despite the good intentions of the players.

…the players at Fort McNair had bought in and grown genuinely angry. Paris, London, Amsterdam, Rotterdam and Brussels were already gone. Every major German city was gone. Every major Polish city. And many others. Beyond the strikes shown on the map above, Sweden had been hit, as had Belarus, the Baltics, Japan, South Korea, the Philippines, Singapore and American appendages including Hawaii and Alaska.

Care had been taken to spare continental United States and European Russia, but casualties had already exceeded those of World War II, and this was when the fighting still remained mostly “tactical.” Surrender was out of the question for either side. 

That the players in a war game had grown genuinely angry suggests that the “players” in a real situation might well share that emotion. But nuclear strategy posits rationality. Additionally, communications and damage assessment will be much worse in a real situation, limiting the information available.

Another report on a wargame study of war between the US and China has just been released. It is more limited than the 1983 game, intended to identify characteristics of a war that would make nuclear use more probable. The games did not include ethics or broader political considerations. The game was run 15 times, and 3 of those ended in general nuclear war. Other outcomes were not necessarily favorable. I’ll come back later to this conclusion:

U.S. diplomacy was much more important than nuclear brinksmanship.

Those who believe that escalation can be controlled in nuclear war must make a stronger case than they have. Several attempts over the last 80 years to argue that and build appropriate war plans have failed. The evidence now is that nuclear escalation, once started, cannot be controlled by actions on the battlefield. Diplomacy must be considered.

We have heard the Reagan-Gorbachev formulation many times.

A nuclear war cannot be won and must never be fought.

But it means little while the weapons are still there, ready to be used.

Cross-posted to Nuclear Diner