The preemption concept was further elaborated in the president's West Point speech and then more formally in the National Security Strategy. It threatens to attack so-called rogue states, which pose a danger to the United States, whether or not they are demonstrably linked to terrorist organizations of global reach. The administration argues that the continued spread of weapons of mass destruction (WMD) technology to states with a history of aggression creates an unacceptable level of risk, and presents "a compelling case for taking anticipatory actions to defend ourselves, even if uncertainty remains as to the time and place of the enemy's attack."

However, a broad-based doctrine of preemption carries serious risks. The Bush administration was right to take a strong stand against terrorists and extremist states, but it had already accomplished this goal with its early words in the period after the September 11 attacks and its actions in Afghanistan. It did not need a formal doctrine of preemption to drive the point home. Rather than enunciate a formal new doctrine, it would have been better to continue to reserve the preemptive military tool for a narrow, rare class of situations where inaction poses a credible risk of large scale, irreversible harm and where other policy tools offer a poor prospect of success. Given that the doctrine has now been promulgated, the Bush administration should clarify and limit the conditions under which it might be applied.

Elevating the preemptive option to a policy doctrine can have serious negative consequences. For one, it reinforces the image of the United States as too quick to use military force and to do so outside the bounds of international law and legitimacy. This can make it more difficult for the United States to gain international support for its use of force, and over the long term, may lead others to resist U.S. foreign policy goals more broadly, including efforts to fight terrorism. Elevating preemption to the level of a formal doctrine may also increase the administration's inclination to reach for the military lever quickly, when other tools still have a good chance of working.

Advocating preemption warns potential enemies to hide the very assets we might wish to take preemptive action against, or to otherwise prepare responses and defenses. In this tactical sense, talking too openly about preemption reduces its likely utility, if and when it is employed.

Finally, advocating preemption may well embolden other countries that would like to justify attacks on their enemies as preemptive in nature.

One can argue that a more explicit policy of preemption actually reinforces deterrence by putting other countries on notice about America's seriousness of purpose in addressing threats such as the possession of weapons of mass destruction by rogue regimes. It also allows the administration to argue that its focus on Iraq is part of a broader security concept and does not represent preoccupation with a specific regime. However, linking the real problem of Iraq's weapons of mass destruction (WMD) to a broader doctrine of preemption (defined to include preventive war) complicated the administration's task in gaining international support for its preferred policy last fall, and may be contributing to our diplomatic difficulties at present as well. Bizarrely, many countries seem to worry as much about restraining the United States as about disarming Saddam-not a defensible position on their part, but a reality nonetheless, and one that the preemption doctrine may have helped create.

Many countries worry that the Bush administration will take a similar preemptive and largely unilateral approach in dealing with other cases such as North Korea or Iran or Syria. Further, other countries' frustration with the United States' decision to grant to itself, (though not to others), a right of preemption may chill their willingness to cooperate fully with the United States in the war on terrorism. To date, that does not seem to be a major problem, but the situation could change.

THE NATIONAL MILITARY STRATEGY AND FAILED STATES

Finally, the administration's strategies for dealing with terror do not have a strong plan for helping failed states-the current or future Afghanistans where al Qaeda could find refuge and resources.

There are many tools needed to reduce the prevalence of failed states. But some have to do with the physical capacity to forcibly stop or mitigate conflict in some situations, and to help keep the peace after parties to a war have agreed to a ceasefire on their own in others.

Today, unfortunately, most countries besides the United States do not have the wherewithal to deploy well-equipped troops quickly and effectively to trouble spots, and then to keep them there once they have been deployed. Surveying the world's conflicts, both those now underway and those of the recent past, it would be desirable that the international community have roughly double its current capacity to deploy and sustain forces abroad. It has been averaging about 100,000 forces in various peace and stabilization missions in recent years, but a survey of the world's hotspots suggest that it would often be useful to be able to deploy and sustain 200,000 troops for such missions. As noted, these missions are important not only for humanitarian reasons, but for national security ones as well-to deprive terrorists of sanctuaries and sources of income (from diamonds, drug trading, and the like) that they can often obtain in failed or failing states.

4

Since some countries will choose not to participate in any given operation, and since troops will need to be rotated to avoid exhaustion and burnout, a total pool of perhaps 600,000 personnel would be desirable. That number is not exact; it is hard to know how troop rotations would work in advance. But a three-to-one ratio of available forces to deployed forces has generally been considered appropriate by the U.S. military. If anything, it is optimistic, and even more than 600,000 could be required to maintain 200,000 on deployment. The international community already has about that number of military personnel who can be rapidly deployed and then sustained in overseas theaters. The problem, however, is that two-thirds of the total number now comes from the United States. But there is no reason the United States should be expected

4

See my recent book, Expanding Global Military Capacity for Humanitarian Intervention (Brookings, 2003). In fact, on average the United States maintains no more than 10 percent of its forces on deployment, away from home station, at a time somewhat more than 100,000 out of a total active-duty personnel strength of 1.4 million. Indeed, it considers that level of effort rather onerous. But that aggregate figure of 1.4 million includes many noncombat troops, so the ratio may be misleading. As another means of estimating availability, note that most U.S. military services have a policy of not having individual personnel be absent from home for more than 120 days a year, essentially revalidating the 3:1 rule. See Michael O'Hanlon, Defense Policy Choices for the Bush Administration (Washington, D.C.: Brookings, 2001), pp. 22-58.

to provide most forces for such missions, and as a practical matter, this country will not be willing or able to do so in any event.

For these reasons, if there is to be additional effort in humanitarian and peace operations in the future, most is likely to come from other countries. That means that of the desired pool of 600,000 deployable military personnel, non-U.S. countries should provide about 500,000 of the troops. In other words, countries besides the United States should more than double their aggregate power projection capabilities.

That number should be sobering for those who consider humanitarian military operations to require only relatively modest amounts of force. But it should be within reach for the international community, if not right away, then over time. To begin, not all troops need be equally well trained and equipped. Some missions will be less demanding than others. Some will not require rapid response or long-range transport. Either the peace accords that precede them will be negotiated over an extended period, allowing ample time for preparations, or the operations will be close to home for countries contributing troops. Even if 200,000 forces might be needed at a time, it is unlikely that it would be necessary to deploy more than 50,000 urgently, and unlikely that more than half to two-thirds would need to operate in extremely austere surroundings.

Although their situations vary greatly from region to region and country to country, developing countries face many common budgetary challenges in any effort to expand military capabilities. The costs would follow from the need for more rigorous training and for better equipment.

Particularly in Africa, a continent facing many acute economic problems, the western powers will need to provide many of the resources required to expand and improve regional military capabilities. Programs now underway, such as the U.S. Africa Crisis Response Initiative (recently renamed the African Contingency Operations and Training Assistance program, or

ACOTA), are important steps in the right direction. But they do not involve nearly enough troops or provide sufficiently rigorous training and sufficiently capable equipment.

The need for more rigorous training is evident. Under current assistance programs, exercises and classes typically take no more than a few weeks. Yet creating a highly ready military, competent across a broad spectrum of operations including combat, typically takes

6

many months if not longer. As a U.S. Army field manual puts it, "The most important training for peace operations remains training for essential combat and basic soldier skills”— underscoring the scope of the challenge for preparing good troops for such missions.' In addition, troops conducting peace and humanitarian interventions also must work with

nongovernmental organizations that provide relief and other services, adding further

8

complexities to any mission. The United States and other foreign militaries cannot be expected to build other countries' armed forces up from the ground level; nor would any such offers necessarily be well received. But months of training, as opposed to weeks, are needed. So are refresher courses every one to two years. At least a doubling in the intensity of training per unit is appropriate. Exercises are also needed to practice coordinating operations at higher and larger levels of effort-notably, for missions involving brigades and divisions. Most of these exercises can be headquarters and staff efforts, as opposed to full-scale field training, but they are critical.

To get a handle on the costs of serviceable equipment for such countries, two different approaches can be taken. One is to examine the costs of a country such as Turkey or South Korea-a country that has typically tried, and succeeded, to field strong ground forces with fairly low defense spending. This approach tends to produce cost estimates that are somewhat too low, perhaps, since such countries do not typically buy large amounts of strategic lift or deployable logistics support equipment.

6 For a good explanation of how hard the U.S. military needed to work to improve its own standards after Vietnam, see Robert H. Scales, Jr., Certain Victory: The U.S. Army in the Gulf War (Washington, D.C.: Brassey's, 1994), pp. 1-38.

'U.S. Army, Field Manual 100-23: Peace Operations (Washington, D.C., 1994), available at www.adtdl.army.mil/cgi-bin/atdl.dll/fm/100-23/fm100-23.htm, chapter 3, p. 8.

8 See for example, Chris Seiple, The U.S. Military/NGO Relationship in Humanitarian Interventions (Carlisle Barracks, Pa.: U.S. Army War College Peacekeeping Institute, 1996); Byman et.al., Strengthening the Partnership.