In going after terrorist groups or rogue states armed with or seeking to acquire nuclear, biological or chemical weapons, conventional forces remain the United States' most effective tool. These forces include special operations units, precision-guided conventional munitions tailored to penetrate or disable underground facilities and destroy the special weapons within them, and the means of promptly delivering such capabilities anywhere in the world.

Using nuclear weapons against these types of threats might be militarily feasible, but it seems very unlikely that any U.S. president would decide to authorize a nuclear strike to pre-empt a suspected terrorist threat or an assault from a WMD-armed power. In theory, if the president believed an attack on the homeland were imminent and likely to be devastating, he might consider a pre-emptive use of a nuclear weapon. In reality, however, the likely uncertainty of any intelligence on the threat, the availability of alternative military actions, the prospect of large losses of innocent lives, and the enormous political ramifications of the first use of a nuclear weapon since Nagasaki would make any president extremely reluctant to take such a step. This is why it is so important to develop long-range conventional weapon systems that could be used to respond promptly to a WMD threat. Even in response to an enemy's first use of nuclear weapons, say, in an Iran-versus-Israel scenario, the ability to respond conventionally and win should be rated higher than the ability to retaliate with nuclear weapons.

One argument put forth in favor of developing new low-yield nuclear weapons is that the fewer innocent casualties likely to result from their use would dampen this self-deterrence. When one examines how presidential-level decisions are made, however, the specific characteristics of the nuclear weapon to be used are unlikely to make a difference. Of primary political significance would be the decision to use a nuclear weapon; questions of yield and the specific forecast of casualties would be lost in the weeds.

The only powerful argument for developing new types of nuclear weapons is their potential to penetrate deeply buried facilities that lie outside the destructive reach of current U.S. conventional capabilities. This argument justifies continuing research on these and alternative means to neutralize hard and deeply buried targets. Any decision on the full development of such capabilities, however, should await the results of more serious research efforts than have been possible to date.

Even so, it should be recognized that any U.S. program to develop new types of nuclear weapons undercuts the president's proclaimed goal to stem and reverse WMD proliferation, and their potential military utility thus has to be weighed against the negative consequences of such a decision for the fight against proliferation. When the United States asserts a national security need to acquire new types of nuclear capabilities, while at the same time continuing to maintain large, if declining, nuclear stockpiles from the Cold War era, the diplomatic power of its demand that others abstain from or even give up nuclear capabilities is greatly weakened.

There is also an argument that researching new types of nuclear capabilities diverts resources from more relevant military capabilities. But the cost of the nuclear weapons research currently on the table is chump change by weapon-system standards. A more relevant economic question concerns nuclear delivery systems. Programs are ongoing both to improve existing land-based Minuteman ICBMs and sea-based Trident ICBMs, and to begin developing successors to them. The real spending will begin as these latter programs move into more advanced development. Just when that should happen, and whether there might not be alternatives to current thinking on them, are questions that should receive high-level attention in the Department of Defense and Congress.

Missile Defense and the Great Powers

There is also a third threat scenario involving the use of WMD: the emergence, down the line, of a hostile great power that would challenge the United States. The only such potential rivals in the foreseeable future are a back-sliding Russia and an openly confrontational China, perhaps even aligned with one another.

It goes without saying that the United States should pursue diplomatic, political and economic strategies to reassure these nations that we do not intend them any harm and that they will derive benefits from cooperative relations with us. But it is only prudent for the United States also to maintain a veiled military posture that makes clear to them that attempting to compete militarily with the United States would entail enormous costs and no success. This is essentially the strategy the Reagan Administration pursued when it threw billions of dollars into defense programs to compel the Soviet Union to give up the competition.

Current U.S. efforts to develop a missile defense system have raised repeated skepticism as to its feasibility, cost and ultimate use. From a technical point of view, it is no doubt feasible to develop at a reasonable cost a highly reliable missile defense system that can protect the country against small numbers--say, tens--of incoming missile warheads. Granted, there have been test failures and operational problems over the past year. But this is to be expected in any new system development and is primarily an engineering challenge, not a question of the program's viability or practical utility. The cost and time required to develop a defense system that can reliably defeat a more extensive threat--say, a few hundred warheads--is more problematic. Effective protection against a threat of this magnitude would require the interception of hostile missiles during the boost phase of their flight, a technology that is less well known and more difficult to master, and may cost more than the nation is willing to spend on this type of threat. Rationally, in the mid-term it might make more sense to continue to depend on deterrence against the types of threats that could be posed by Russia and China--imperfect though deterrence may be--and to allocate more resources instead to protecting the country against terrorists and rogue states.

The challenges of developing a missile defense system are also politically critical. More specifically, in the case of Russia--the only potentially hostile country still possessing a large nuclear force and extensive nuclear infrastructure--U.S. strategy should make clear that no missile defense we could conceivably deploy for decades would jeopardize Russian retaliatory capabilities. Negotiations with Russia on continuing mutual nuclear force reductions got off to a good start during the first Bush term and should be pursued further. So long as Russia retains confidence in its own deterrent capability, cooperation can continue without the need for a new arms control treaty.

China, on the other hand, with its still-limited long-range nuclear forces and only rudimentary technology and infrastructure, calls for a different approach. The United States currently has the means to deny China a retaliatory capability, meaning a capability to deter the United States from acting in defense of its interests in regions close to China's shores. It is essential that the United States retain its freedom of action. The short-lived 1995 crisis over Taiwan, for example, was resolved in part through diplomatic efforts and mutual realization that each side had much to lose from continued tensions, but also because U.S. military activities in the region indicated a willingness to intervene.

In a scenario ten to twenty years from now in which China already possesses a significant long-range nuclear arsenal and the capability to strike the U.S. homeland, its willingness to compromise might be less pronounced. In such circumstances, Chinese leaders might conclude that given the centrality of the Taiwan issue to them and its marginality in U.S. calculations, Washington would wish to avoid escalation to a nuclear exchange. As a result, for the next ten years at least, the United States may wish to maintain a substantial lead in offensive capabilities vis-Ã--vis the Chinese, while building defenses capable of stopping the relatively small long-range force that China is likely to deploy during this period.

In considering the U.S.-China strategic relationship, it is also important to look at the question of nuclear testing. Since 1945, the United States has conducted some 1,030 nuclear tests, more than all other nuclear-armed nations combined. China has conducted only 45. As a result, the United States holds a considerable technological edge over its potential rival. Having already hit the flat part of the learning curve on nuclear technology, the United States has little to benefit from additional tests in terms of new capabilities, unlike China, which still has a lot to learn. In this context, it is essential for the United States to avoid giving China a pretext to advance its nuclear weapon designs and gain new capabilities by breaking the more than ten-year-old moratorium on nuclear testing. China may well break the moratorium on its own, but we should studiously avoid giving it an excuse to do so.

Strengthening Supply-Side Restraints

Finally, any effective U.S. strategy to contain and reverse WMD proliferation must place greater emphasis on the supply side of the WMD equation. Pre-emption and interception are by themselves insufficient; they must go together with effective preventive diplomacy. This is particularly true in the case of terrorist groups, given the difficulties in finding and deterring them. The administration should take advantage of the existing bipartisan support behind the 1991 Nunn-Lugar Act and the Comprehensive Threat Reduction program, aimed at securing Russia's vast stocks of WMD materials, to improve upon and further expand present initiatives. The 2002 G-8 non-proliferation initiative pledging $20 billion over ten years for Russia and other former Soviet states is one recent example of how U.S. leadership can promote multilateral cooperation in this field.