The first problem with this argument is that the Cuban missile crisis wouldn’t have happened at all if nuclear weapons hadn’t existed. So, unless Tepperman wants to argue that the crisis posed no danger whatsoever (a very high bar to clear), nuclear weapons clearly did not serve us well in that instance. Second, while both sides may have behaved rationally in pulling back from the brink, the point is that neither wanted to approach the brink in the first place. The lesson of the crisis is that human beings, acting in what they perceive to be a rational manner, can produce outcomes that are wildly out of step with their self-interest. After all, Khrushchev was perfectly sane when he ordered missiles placed in Cuba, but his actions had unintended consequences. And if rationality can produce unintended outcomes, then Tepperman is wrong in suggesting that MAD is foolproof: The Cuban missile crisis could have ended very differently.

That certainly seems to be the case when you consider all the near-misses that occurred during those 13 days. Contra Tepperman’s assumptions, the Cuban missile crisis showed that, even when they try to behave rationally, states are not always in control of their own behavior. For example, as the crisis neared its climax, Air Force officers in California launched an ICBM over the Pacific in a pre-scheduled test--not the sort of thing one (rationally) wants to do during a nuclear stand-off and certainly not something President Kennedy approved. What’s more, the following day, a Soviet officer in Cuba shot down an American reconnaissance flight in direct contradiction of Khrushchev’s orders, further escalating the conflict. Worst of all, during the crisis the Soviets had dozens of tactical nuclear weapons on the island--i.e., nukes meant for battlefield use--and for a time their field commanders had pre-delegated authority to use them in the event of a U.S. attack. (For more on this, see here.) In fact, Castro had told Khrushchev that, if the Americans attacked, he should use the nuclear weapons even though he recognized that Cuba would likely be destroyed in retaliation. This contradicts Tepperman’s assertion that “even the craziest tin-pot dictator is forced to accept that war with a nuclear state is unwinnable and therefore not worth the effort.”

There are a lot of other problems with Tepperman’s piece, but he makes perhaps the most damning arguments against its thesis himself. Toward the end of the article, having spent more than 2,000 words explaining how nukes protect us, Tepperman adds that a key problem with the “dreamy ideal” of disarmament is that it distracts us from the more important problem: “making the world we actually live in--the nuclear world--safer.” Given that his point was that nukes made us safe, it was a little jarring to read that it needs to be made safer, because that suggests there is some danger now.

Indeed, Tepperman goes on to note that, to be truly safe, we need to make sure that every nuclear-armed country has a secure retaliatory capability, presumably so that in a crisis it doesn’t feel the need to attack first because of its perceived disadvantage. Alas, this problem of perceived insecurity remained unresolved throughout the Cold War despite the thousands of weapons the United States and the Soviet Union each had—neither side ever felt truly safe and often feared a first strike. What’s more, Tepperman’s caveat suggests there might be a dangerous instability between states, like India and Pakistan, that have only small arsenals but that Tepperman had earlier assured the reader were less likely to fight each other if they had nukes. (And what are we going to do, give North Korea a secure retaliatory capability?) Finally, after arguing that there’s little risk of nukes falling into the hands of terrorists, he then acknowledges that we ought to continue helping Russia and Pakistan to safeguard their arsenals and to secure “loose nukes.” (Wait, loose nukes?!) This, Tepperman assures us will help prevent the danger of an “accidental launch.” (Accidental launch?!)

The implication here is that nuclear weapons can pose significant dangers--including in their current disposition. Which, of course, is the precisely point made by President Obama and others advocating disarmament.

Obama's hunger for a diplomatic success, such as it is, allowed the Russians to exact a price: linkage between offensive and defensive nuclear weapons. This is important for Russia because of the huge American technological advantage in defensive weaponry. We can reliably shoot down an intercontinental ballistic missile. They cannot. And since defensive weaponry will be the decisive strategic factor of the 21st century, Russia has striven mightily for a quarter-century to halt its development.

When McNamara joined the Kennedy administration in 1961, American nuclear "strategy" called for launching the entire nuclear arsenal-nearly 3,500 weapons-at the communist bloc if the Soviets made any move against Western Europe. This approach had severe flaws. For one thing, it meant that the United States would kill hundreds of millions of civilians. Indeed, it would decimate nations, like China, that were unlikely to even be involved in a Soviet attack. What's more, such an attack would leave the United States open to nuclear retaliation. After all, there was little chance that our first strike would destroy every Soviet weapon, and even a modest number of warheads could take out most major American cities.

McNamara, understandably, wanted an alternative, and initially he settled upon a strategy known as "counterforce." In the event of a Soviet attack on Western Europe, counterforce dictated that the United States would first target Soviet military forces. The strike need not be an all-out "war orgasm," as Herman Kahn memorably put it. Rather, the idea would be to stop the conventional assault and degrade the Soviet ability to retaliate. We would keep some nuclear weapons in reserve and hold Soviet cities hostage to a second strike to deter them from responding in kind. In essence, we would treat nukes much like regular weapons. As McNamara said on June 17, 1962, "[Our] principal military objectives, in the event of a nuclear war stemming from a major attack on the Alliance, should be the destruction of an enemy's forces, not his civilian population." The implication, as I noted in U.S. vs. Them, was that, like a conventional war, a nuclear war might be winnable.

But McNamara soon realized that, although it might be better to have options besides launching a total nuclear holocaust, counterforce would not meaningfully limit damage to the United States in the event of a nuclear conflict. It took only a few enemy weapons to stop the United States from functioning as a society; we wouldn't be able to take out all the Soviet weapons in a first strike; and if we used nukes, the Soviets were unlikely to resist their impulse to retaliate. Combined with studies showing that it was impossible to effectively defend against a nuclear attack-either by shooting down incoming missiles and bombers, or by sheltering civilians from radiation-that logic forced McNamara to conclude that the United States would never benefit from using nuclear weapons against a nuclear-armed adversary. Doing so would be suicidal.

So in September 1967, McNamara gave a speech that emphasized deterrence through the ability to inflict "unacceptable damage" on the enemy even after a surprise first strike. He called this an "assured-destruction" capability. McNamara's conversion by no means resolved debates over U.S. nuclear posture. Conservatives never accepted the vulnerability imposed by mutual assured destruction, which they parodied with the acronym "MAD." They fought for missile defense, fallout shelters, and war-fighting strategies. Democrats and Republicans alike constantly worried whether the Soviets might eke out some nuclear advantage. And the United States continued to target Soviet nuclear forces because, in the event of a nuclear war, one might as well try to win. But McNamara had long ago realized the impossibility of victory. Mutual assured destruction was not a policy that American officials could accept or reject; it was an inescapable condition of large nuclear arsenals. The best that policymakers could do-indeed, the responsibility they had-was to manage that condition so that it did not lead to war.   

In his later years, McNamara advocated efforts toward nuclear disarmament, and he strongly criticized George W. Bush's flaccid efforts at arms control and his plans to build new, bunker-busting nuclear weapons. Such positions put him in the company of other high-level nuclear skeptics not known for their dovishness, like Paul Nitze and, more recently, George Shultz, Henry Kissinger, Bill Perry, and Sam Nunn. They also put him in line with Barack Obama's vision of a world without nuclear weapons and his agreement yesterday with President Medvedev to further reduce the U.S. and Russian arsenals. But none of those efforts would be possible were it not for the realization, officially articulated by McNamara, that a nuclear war cannot be won and must therefore never be fought.  

--Peter Scoblic