Nuclear Deterrence Theory, Nuclear Proliferation, and National Missile Defense

On December 17, 2002, President George W. Bush ordered the deployment of a national missile defense (NMD). Proponents of missile defenses, both inside and outside the Bush administration, argue that, absent NMD, the proliferation of nuclear weapons and the greater U.S. vulnerability that this entails will significantly limit the United States' ability to secure it foreign policy goals. "A policy of intentional vulnerability by the Western nations," Secretary of Defense Donald Rumsfeld argues, "could give rogue states the power to hold our people hostage to nuclear blackmail—in an effort to prevent us from projecting force to stop aggression."¹ Similarly, Walter Slocombe as undersecretary of defense in the Clinton administration asserted, "Without defenses, potential aggressors might think that the threat of strikes against U.S. cities could coerce the United States into failing to meet its commitments."²

To what extent do the spread of nuclear weapons and the means to deliver them threaten U.S. interests and impede the United States' ability to pursue its interests? To what degree would a national missile defense offset any adverse effects of this spread and at what political and economic cost? The answers to these questions depend at least in part on one's ideas about how nuclear deterrence works.

Surprisingly though, most analyses of the effects of the spread of nuclear weapons and of the corresponding value of NMD have not grounded themselves directly in nuclear deterrence theory.³ These studies have generally not drawn explicitly on the foundational work of Bernard Brodie, Hermann Kahn, Thomas Schelling, Glenn Snyder, Albert Wohlstetter, and a few others. Indeed some analyses have explicitly rejected nuclear deterrence theory, seeing it as an obsolete and possibly dangerous kind of Cold War thinking.⁴ Other analyses, focusing more narrowly on specific policy or technical issues, have typically relied on a gloss on nuclear deterrence theory, variously asserting either that the threat of retaliation will deter rogue states as it deterred the Soviet Union during the Cold War or that this threat may prove insufficient to deter rogues.⁵ And still other more theoretically oriented studies have looked to different theories and conceptual frameworks and not to nuclear deterrence theory.⁶

This article extends some of the fundamental ideas developed by Brodie, Snyder, and especially Schelling in the U.S.-Soviet context to examine the effects of the spread of nuclear weapons and of NMD. Three broad conclusions emerge. First, although nuclear deterrence theory remains useful, its implications vary with the conditions in which it is applied. Therefore, the relative stability between the United States and the Soviet Union during the second half of the Cold War following the 1963 Cuban missile crisis may provide a poor guide to the stability of a crisis between the United States and a new nuclear state (or, for that matter, between two new nuclear states such as India and Pakistan). Second, NMD would give the United States somewhat more freedom of action and make a rogue state more likely to back down in a crisis. But these effects will be modest unless the defenses are very good. Finally, NMD, unless it is extremely effective, is likely to raise the risk both of a nuclear attack on the United States and of nuclear weapons striking the United States. These greater risks, moreover, are not the result of a mistaken overconfidence in the effectiveness of NMD. They are the direct consequence of a greater U.S. willingness to press its interests in a crisis harder.

There are three parts to the analysis. The first revisits and elaborates some of the elements of nuclear deterrence theory. It focuses especially on the credibility problem inherent in nuclear deterrence, the way that conflicts of interest play themselves out in the presence of nuclear weapons, and the dynamics of brinkmanship. It also shows that "rational deterrence theory," despite some claims to the contrary, actually helps to explain how deterrence between rational actors might fail and what makes this failure more likely. The second part discusses the dynamics of escalation and the likelihood of U.S. intervention...