Limited National and Allied Missile Defense

To the Editors:

In their article "National Missile Defense and the Future of U.S. Nuclear Weapons Policy," Charles Glaser and Steve Fetter perform a valuable service for readers of International Security and, more generally, the U.S. debate on national missile defense (NMD). ¹ Their nonpolemical treatment of the technical, military, diplomatic, and strategic issues in the missile defense debate is unusual for its rigor and thoughtfulness. They argue that deploying defenses against the possible rogue-state missile threat would have some value—especially if the defenses emphasized boost-phase systems on land, at sea, or in the air that could shoot down enemy missiles early in their flight before most countermeasures could be deployed. But at the same time, they wisely argue that missile defenses could do more harm than good for U.S. security if Russia and China are not reassured in the process. ² On balance, given these latter concerns, the authors offer a decidedly ambivalent overall assessment of the desirability of NMD, but much sage advice about how NMD should be deployed if it is to be built.

Glaser and Fetter, however, tend to underestimate the potential importance of missile defenses. They do well to avoid the mistake of many NMD critics when they note that missiles have a certain cachet not possessed by "suitcase bombs." Missiles need not be predeployed by agents of questionable trustworthiness who must get past border inspectors without being caught. Moreover, their very existence can serve a political purpose even if their owners do not explicitly threaten to employ them. The authors are also surely right not to go to the other extreme and portray missiles as the top security threat facing the United States in the years ahead.

That said, there are three main points that Glaser and Fetter brush over too lightly or ignore altogether. All bolster the case for limited national missile defense, making the desirability of such a system greater than the authors allege—even if they are still right to argue that any NMD system must be limited and nonthreatening to Moscow and Beijing.

Issue #1: The Possibility of Deterrence Failure

The most important issue that Glaser and Fetter neglect is the question of when deterrence can fail. They do note that such deterrence failure is possible, and briefly give plausible reasons why it might occur. But ultimately, they conclude that such an outcome is "probably unlikely even relative to the [other] scenarios" that they outline (pp.67-68, at p.68). The subject merits much greater discussion.

Under most circumstances, rogue-state leaders are likely to be just as deterrable as were Soviet leaders. North Korea has been dissuaded from launching a second Korean war for nearly half a century. Iraq has tested the United States on numerous occasions since the end of Desert Storm—but has not pushed so far as to provoke war.

Nonetheless, deterrence is not totally reliable. It is imperfect because crises have internal dynamics capable of producing outcomes that would not have been foreseen or desired at their outset. ³ There are also scenarios in which classic deterrence theory is irrelevant. Most notably, the logic of deterrence does not apply to a national leader who believes that he is highly likely to be captured or killed during a war.

If the United States were to fight Iraq or North Korea again, as current Pentagon plans assume that it might, Washington would want to have the option of overthrowing their regimes and occupying their territories. Indeed, it is largely for such reasons that Pentagon strategy for the last ten years has postulated the possibility of two nearly simultaneous military operations each involving half a million U.S. troops. The decline of the Iraqi and North Korean threats, together with the ongoing strengthening of the South Korean military, make it unlikely that such large forces would be needed just to protect...