Is It Time for the United States to Get MAD?

Is the use nuclear warheads appropriate as retaliation in the war against terrorism?

November 15, 2001

Is the use nuclear warheads appropriate as retaliation in the war against terrorism?

Let’s get some things straight right at the outset. For all of bin Laden’s rhetoric, there is no prospect of any U.S. nuclear strike, far from it. For starters, there is no target in Afghanistan worth a nuclear weapon. And the United States would be foolish in the extreme to use nuclear weapons in current circumstances. Doing so would destroy the U.S.-led coalition and turn the globe’s sympathies toward bin Laden and Afghanistan.

Still, bin Laden’s threats are bound to evoke a debate in the United States about images of assured destruction, mutually assured destruction, or MAD, of Cold War superpower nuclear relations. No matter how mad the United States is, this is no time to get MAD over nuclear weapons.

But this is very much a time to rethink how, if at all, U.S. nuclear threats might be relevant to the campaign against terrorism.

In the bad old days of the Cold War, the United States and the Soviet Union were both MAD over nuclear weapons. MAD was sometimes conceived as a strategy — Robert McNamara as Secretary of Defense in the 1960s used it that way to try to limit the military’s appetite for more new nuclear weapons.

But MAD was mostly a fact of life. Once both nuclear superpowers had what were called secure second-strike forces, both could destroy the other even if it was struck first.

MAD was not appealing, but it was mutually deterring, hence made for stability. It gave rise to the awful aphorism about the superpower nuclear stalemate: offense is defense, and defense is offense; killing people is good, but killing weapons is bad.

Offense was defense because, so long as the other side knew that you had the capacity to devastate it even if it struck first, it would not strike at all. It would be deterred. Defense was offense because if you could deny the other the ability to destroy you even if you struck first, then you might be tempted to strike first.

Killing weapons was bad because if you could target the other side’s nuclear weapons and destroy enough of them, you might be able to deny it the ability to retaliate, and thus you might be tempted to strike first.

Under the MAD aphorism, the ideal nuclear weapons were, like early submarine-launched weapons, neither very vulnerable to a first strike nor very capable of carrying one out because they were not very accurate.

The MAD logic was indeed mad, so the world can do better now that the Cold War is history. Alas, that process has been complicated by the bitter polemics over the Bush administration’s proposal for defense against ballistic missiles.

What is relevant from all this past history now is not “MAD” but “AD,” — not mutually assured destruction, but assured destruction. No conceivable U.S. foe in the campaign against terrorism has any capacity to defend against U.S. nuclear strikes, so the question is not what the United States can do — but what it should do, and more pointedly, what it should say it will do.

Surely, the United States will not want to rule out nuclear retaliation against nuclear terrorism. The rub is how credible that could be. If terrorists exploded a “suitcase” bomb, nuclear retaliation would be proportionate. But retaliation against what?

Or suppose, somewhat more probably, that terrorists exploded instead a much cruder “radiation” bomb. The purpose of such a bomb would not be to seek a nuclear explosion, but “only” the release of lethal radioactive material.

Such a bomb might kill tens of thousands of Americans. But would the United States want to escalate by crossing the fully nuclear threshold, even if it could decide on its own terms what to strike? An exchange of Mecca for Washington — or Medina for Philadelphia?

Probably, the United States will want to modestly brandish nuclear weapons — and leave some ambiguity about exactly when it might use them. In principle, it surely might threaten assured destruction in response to a nuclear strike of any sort that could be traced back to a terrorist group. It might do the same for any state that harbored or assisted the terrorist’s use of weapons.

But even here there are clear limits. What, for example, if the insinuating material were traced back to rogue elements in the former Soviet Union? We surely wouldn’t contemplate nuking Moscow.

The still thornier question is whether the United States should make similar threats in response to chemical or biological attacks that killed tens of thousands of Americans. Attribution would be harder than for nuclear attacks because — as the current anthrax scare vividly demonstrates — there are more sources of chemical and biological material.

And yet, it is argued that Saddam Hussein’s Iraq was deterred from using chemical weapons during Desert Storm, even against Israel, because it worried that the United States would escalate to nuclear retaliation.

So there is value to ambiguity, to refusing to rule out an AD-type nuclear response — and to stressing that if large numbers of Americans were killed, all weapons would be considered in responding.

Finally, the Cold War’s nuclear arms control negotiations had the side-effect, partly intended, of seeming to put them in a class by themselves. In responding to terrorism, the United States will want to blur somewhat that separation, by keeping the focus on how destructive chemical and, especially, biological weapons can be.

No, this is no time for the United States to get MAD over terrorism. It is the time, though, to think about what nuclear threats might AD to the weapons it has in combating terrorism.