MAD is Not Bad

Robert McNamara was US secretary of Defense under John F. Kennedy and Lyndon Baines Johnson, when he conceived the theory of Mutually Assured Destruction (MAD). He spoke with NPQ in Washington.

NPQ | President Bill Clinton has proposed a limited missile shield for the United States. Presidential candidate George W. Bush has called for a broader national shield as a more ethical substitute for the Cold War strategy of "mutually assured destruction" as well as unilateral US nuclear reductions.
As the person who conceived the MAD theory, what is your response?

ROBERT McNAMARA | Bush’s proposal for unilateral reductions of nuclear weapons vis a vis the Russians—disassociated from his position on ballistic missile defense—is a major step forward because he seems to understand that deterrence does not require an equality in the number of missiles each side has. He is quite correct to say that we should not let our force levels be determined by that of the Russians. I also give him great credit for wanting to reduce the danger of accidental launch of nuclear warheads by removing hair-trigger alert.

Bush doesn’t put it this way, but he is really asking what the US needs minimally to achieve stability of deterrence. To answer this question, one needs to understand the basic intellectual foundations of the mutually assured destruction concept: There is no other basis for stability of deterrence between two nuclear-equipped opponents than the confidence on each side that they have the capability to absorb a first strike from the other side with sufficient weapons surviving to inflict unacceptable damage on the opponent when launching a second strike.

When I developed this concept in 1961, I determined that the United States needed 400 surviving warheads to deter the Russians from a first strike. How did we end up, then, with 15,000 on each side at the height of the Cold War? Because we didn’t have enough spies, enough information, to know their intentions. We built up according to a worst-case scenario of Soviet production capabilities down the road (which, it turned out, exceeded their intentions at the time). And once they saw us building up more than they thought we would, they responded in kind. That is where MAD gave way to an arms race.

However, the idea of supplanting MAD with a ballistic missile defense is just dangerously wrong. It is clear from experience that any attempt by a nuclear-equipped power to erect a defense will cause a buildup on the other side.
In November 1966, when I met with President Lyndon Johnson in Austin, Texas, to go over the proposed defense budget, we had photographs that showed the Russians had begun to deploy an anti-ballistic missile system around Moscow.

We assumed it would be insane for them to deploy it just around Moscow and, therefore, concluded this was a first step toward a nationwide system. In response, the Congress wanted appropriations for the deployment of a US ballistic missile defense system.

I opposed this, arguing that the proper response was a further buildup of the US nuclear force to compensate for any losses caused by their defense. When President Johnson made this case to (Soviet Premier Alexei) Kosygin in 1967, I remember Kosygin pounding the table and making the opposite case: "Defense is moral, offense is immoral."
But that is absurd.

In the end, we pursued a joint track of no defense and limits on offense. Ultimately, that led to the Anti-Ballistic Missile Treaty and the SALT arms reduction treaties.
My theory was then, and my belief is today, that you cannot limit offensive weapons by treaty in the face of an unlimited defense. If one side is limited by treaty to, say, 3000 warheads and the other side installs a defense system with no restraints, at some point that unlimited defense will kill so many of those 3000 warheads that "unacceptable damage" cannot be assured and stability of deterrence is lost.
That is what underlies the linkage between the antiballistic missile (ABM) treaty and limits on offensive weapons.

And that is why it is crazy when President Clinton, George W. Bush or anyone else talks about a rather vague, poorly described US deployment of defense in the face of a statement that we will introduce unilateral reductions in offense, or in Clinton’s case, seek reductions with the Russians’ to 2500 missiles. The other side will never go for it.

NPQ | What went through your mind in 1966 when the Russians deployed a defensive system is, then, exactly what is going through the minds of Chinese and Russian strategists today?

McNAMARA | Exactly. Why? Because the United States is not clear what it means by a missile defense. Clinton says he is only talking about 100 missiles in Alaska to defend against North Korea, but he has also said that might be followed with another deployment a few years later to strengthen the system.

This is not what Bush has in mind, and it is certainly not what many Republicans have in mind. Bush has mentioned a three-stage program: first, 100 in Alaska; then 100 in North Dakota; and, ultimately, a space-based element. This space-based element is the most ill-defined and carries a potential for a "kill capability" of any enemy’s missiles far greater than the first two combined.

That is what the Russians see. They think, "My God, these people are out of their minds. The Americans are really thinking about a missile shield that will make our offensive weapons inadequate" and destabilize our deterrent relationship.

NPQ | From the standpoint of the national interests of Russia and China, then, they would be justified in responding to the US plans with a nuclear arms buildup?

McNAMARA | If we were in their shoes, I can tell you, that is exactly what we would do. We Americans will argue among ourselves, some saying this defense system will make us perfectly secure and others saying it won’t work.

Not knowing, how will the Russians or Chinese ever accurately appraise the "kill capability " of our defense system? How can a leader charged with defending his own country respond? He can only do the responsible thing in the face of uncertainty: assume a worst-case scenario. That is not irrational, or illogical.

If you are dealing with the security of your country, with the possibility of nuclear war, you have to look at worst cases. For them, the response to the deployment of a US missile defense will be the expansion of their offense. That is why this situation is so dangerous.

NPQ | If America had had a limited missile defense, maybe the Cuban missile crisis in which you were so intimately involved would have been avoided?

McNAMARA | On the contrary. What we came to know 30 years later was that the real threat was not from missiles based in Cuba, but nuclear-armed bombers that could have attacked the East Coast of the US, against which a missile defense system wouldn’t have been useful.

This is illustrative of the problem with a missile defense: The threat can come from elsewhere. That is why the best course is to foreswear defense, and cut offensive weapons to the minimum.

NPQ | How, then, ought nuclear powers relate to one another?

McNAMARA | First, there must be transparency among all nuclear powers so we all know each other’s intentions and capabilities. Second, reduce nuclear weapons to the minimum required to maintain an appropriate qualitative balance with one’s opponent so there is deterrent stability.
Deterrent stability can only be achieved when each party believes it can absorb a first strike with enough surviving weapons to inflict unacceptable damage on the other.

Back to Index