To the editors:
In his essay, Jean-Pierre Dupuy correctly criticizes nuclear deterrence and Mutually Assured Destruction (MAD). “If MAD requires madmen to fail,” he concludes, “it also requires madmen to succeed. When it comes to nuclear strategy, madmen are needed all around.” This letter elaborates on that statement.1
Nuclear strategy is discussed much less frequently today than during the Cold War, but nuclear deterrence is still highly regarded within the national security community. In a 2013 speech, Secretary of State John Kerry declared that “a strong nuclear deterrent remains the cornerstone of US national security.”2 This letter outlines five faults that have the potential to transform nuclear deterrence into the end of civilization.
Nuclear strategy places great emphasis on the credibility of deterrence. But is it really credible to threaten an action that would result in our own destruction? Consider the following exchange between Senator John Glenn and Secretary of Defense Harold Brown that took place during top-secret hearings before the Senate Foreign Relations Committee on September 16, 1980:
Glenn: I get lost in what is credible and not credible. This whole thing gets so incredible when you consider wiping out whole nations, it is difficult to establish credibility.
Brown: That is why we sound a little crazy when we talk about it.
Glenn: That is the best statement of the day.
Dupuy examined the need for irrationality in his paper. This letter merely adds to his argument. For nuclear deterrence to work, our adversaries must be rational enough for our threats to deter them, but we must be sufficiently irrational to remain undeterred by comparable threats. A 1995 US Strategic Command report, “Essentials of Post-Cold War Deterrence,” was unusually candid on this point:
The fact that some elements [in the American nuclear command] may appear to be potentially “out of control” can be beneficial to creating and reinforcing fears and doubts within the minds of an adversary’s decision makers. This essential sense of fear is the working force of deterrence. That the US may become irrational and vindictive if its vital interests are attacked should be part of the national persona we project to all adversaries [emphasis added].3
Most people tend to think of nuclear deterrence and Mutually Assured Destruction as synonymous, but history shows that nuclear threats have often been used when the stakes are much lower. No vital American or Soviet interests were at stake during the 1962 Cuban Missile Crisis. “It doesn’t make any difference,” US President John F. Kennedy remarked in private, “if you get blown up by an ICBM flying from the Soviet Union or one from ninety miles away. Geography doesn’t mean that much.”4 Yet he risked a nuclear war to get the Soviet missiles removed. Soviet leader Nikita Khrushchev was also prepared to risk a nuclear war to deter America from attacking Cuba again.
The threats made by the US against Iran are a more recent example. The Nuclear Posture Review (NPR) produced by the US Department of Defense in 2010 states that, “the United States will not use or threaten to use nuclear weapons against non-nuclear weapons states that are… in compliance with their nuclear non- proliferation obligations.” The NPR also notes that “North Korea and Iran have violated non- proliferation obligations.”5 There is thus an implicit threat to attack Iran with nuclear weapons. No threat is directly stated in the NPR, but the media in both the US and Iran put together these statements to reach that conclusion.
Even when incapacitated to the degree that they would be unable to legally drive, the leaders of nuclear armed states still retain the ability to start a nuclear war. Russian President Boris Yeltsin was frequently disoriented as a result of alcoholism. US President Richard Nixon also had a drinking problem. On October 11, 1973, in the midst of the international crisis sparked by the Yom Kippur War, British Prime Minister Edward Heath requested a phone conversation with Nixon. In a declassified telephone transcript, Henry Kissinger, Nixon’s National Security Advisor, can be seen asking his assistant “Can we tell them no? When I talked to the president, he was loaded.”6 In his memoirs, former British Prime Minister Tony Blair admitted that his daily alcohol consumption was “definitely at the outer limit. Stiff whiskey or G&T before dinner, couple of glasses of wine or even half a bottle with it.”7 Given that a failure of nuclear deterrence could destroy civilization, it is illogical that someone unable to drive has the power to start a nuclear war.
Even though a number of governments risk destroying civilization by relying on nuclear deterrence, none have estimated the risk of a failure producing that result. I no longer hold security clearances and cannot be sure that no classified studies exist, but having spoken with people possessing the relevant clearances, I have not heard of any being undertaken. I have also discussed the matter with high ranking staff at United States Strategic Command, but there are no indications that any studies have taken place.
In a private conversation, an individual who had held a high position in the nuclear command and control structure told me that he did not want the American public to learn just how risky nuclear deterrence really is. If they knew, the public would, he believed, react irrationally. This would, in turn, lead to new national security policies, namely complete nuclear disarmament—an approach he considered even more dangerous than deterrence.
Nuclear deterrence is almost surely far riskier than society would tolerate if it understood the danger. Even if the risk of a new world war was only one-in-five hundred each year, the cumulative risk over the roughly ninety-year life expectancy of a child born today would be close to one-in-six. This is the same level of risk found in Russian roulette.
“When it comes to nuclear strategy,” Dupuy notes, “madmen are needed all around.” If we are to lay any claim to being rational, it is time for us to critically re-examine the questions raised in Dupuy’s essay and this letter.
Martin Hellman
Jean-Pierre Dupuy replies:
Martin Hellman makes essential points that I accept gratefully.
