The United States is in the midst of its third major debate on nationwide ballistic missile defense-the first culminating in the 1972 ABM Treaty and the second sparked by President Reagan's "Star Wars" speech in 1983. This time the Cold War is over, the objectives for the defense are limited, and technology has advanced to the point where some options may be technically feasible.

However, intercontinental ballistic missiles (ICBMs) are not the primary threat to the United States, as events since September 11 demonstrate. Other homeland defense programs, especially civil defenses against bioterrorism, are more important. Yet emerging missile states may acquire ICBMs some day. To the extent that this is a concern, diplomatic efforts can limit the spread of ballistic missiles, and deterrence can dissuade their use. National missile defense (NMD), then, is insurance against the relatively unlikely event that ICBMs will be launched against the United States.

If the United States decides to deploy a limited NMD, the questions become what type and how much? A midcourse NMD system (one that attempts to intercept missile warheads as they fall through outer space) of the sort proposed for deployment in Alaska is the most technically mature option and would probably work well enough against emerging ICBM threats to justify limited deployment, assuming that the threat materializes. However, such a defense should contain only about 20 interceptors to minimize adverse political reactions from Russia and China. Over the long run, midcourse defenses may be vulnerable to sophisticated countermeasures. Therefore, the United States should place greater emphasis on land, naval, and air-based boostphase intercept options (defenses that attempt to intercept the ballistic missile while its rocket motors are still burning) because they are more robust to countermeasures and they pose relatively little threat to Russia and China. Space-based boost-phase NMD systems have the advantage of global coverage; however, they are technically more challenging, probably more expensive, and more destabilizing.