The Bush administration has proclaimed a doctrine of unilateral preemption as a core part of its National Security Strategy. The limits of this approach are demonstrated daily in Iraq, where the United States is bearing the burden for security, reconstruction, and reform essentially on its own. Yet the world cannot afford to look the other way when faced with the prospect, as in Iraq, of a brutal ruler acquiring nuclear weapons or other weapons of mass destruction (WMD). Addressing this danger requires a different strategy, one that maximizes the chances of early and effective collective action. In this regard, and in comparison to the changes that are taking place in the area of intervention for the purposes of humanitarian protection, the biggest problem with the Bush preemption strategy may be that it does not go far enough.

In the name of protecting state sovereignty, international law traditionally prohibited states from intervening in one another's affairs, with military force or otherwise. But members of the human rights and humanitarian protection communities came to realize that, in light of the humanitarian catastrophes of the 1990s, from famine to genocide to ethnic cleansing, those principles will not do. The world could no longer sit and wait, reacting only when a crisis caused massive human suffering or spilled across borders, posing more conventional threats to international peace and security. As a result, in late 2001, an international commission of legal practitioners and scholars, responding to a challenge from the UN secretary-general, proposed a new doctrine, which they called "The Responsibility to Protect." This far-reaching principle holds that today UN member states have a responsibility to protect the lives, liberty, and basic human rights of their citizens, and that if they fail or are unable to carry it out, the international community has a responsibility to step in.

We propose a corollary principle in the field of global security: a collective "duty to prevent" nations run by rulers without internal checks on their power from acquiring or using WMD. For many years, a small but determined group of regimes has pursued proliferation in spite of—and, to a certain extent, without breaking—the international rules barring such activity. Some of these nations cooperate with one another, trading missile technology for uranium-enrichment know-how, for example. Their cooperation, dangerous in itself, also creates incentives for others to develop a nuclear capacity in response. These regimes can also provide a ready source of weapons and technology to individuals and terrorists. The threat is gravest when the states pursuing WMD are closed societies headed by rulers who menace their own citizens as much as they do their neighbors and potential adversaries.

Such threats demand a global response. Like the responsibility to protect, the duty to prevent begins from the premise that the rules now governing the use of force, devised in 1945 and embedded in the UN Charter, are inadequate. Both new principles respond to a growing recognition, born of logic and experience, that in the twenty-first century maintaining global peace and security requires states to be proactive rather than reactive. And both recognize that UN members have responsibilities as well as rights.

The duty to prevent has three critical features. First, it seeks to control not only the proliferation of WMD but also people who possess them. Second, it emphasizes prevention, calling on the international community to act early in order to be effective and develop a menu of potential measures aimed at particular governments—especially measures that can be taken well short of any use of force. Third, the duty to prevent should be exercised collectively, through a global or regional organization.


We live in a world of old rules and new threats. This period did not begin on September 11, 2001. Before then, politicians and public figures were already lacing their millennium speeches with calls for a new global financial architecture, new definitions of national self-interest and humanitarian intervention, and new ways of organizing international institutions. They recognized that the existing rules and institutions created to address the economic, political, and security problems of the last century were inadequate for solving a new generation of threats to world order: failed states; regional economic crises; sovereign bankruptcies; the spread of HIV/AIDS and other new viruses; global warming; the rise of global criminal networks; and trafficking in arms, money, women, workers, and drugs.

Although the worst threats to the international order in the 1990s arose from internal conflicts—civil wars, ethnic bloodletting, and resurgent nationalism—the cardinal doctrines of the post-1945 order apply to wars between nations, not within them. The UN Charter binds states only to refrain from the use or threat of force in "their international relations" and explicitly protects their "domestic jurisdiction" from outside interference. And a broad doctrine prohibiting intervention in a state's internal affairs is well established in customary law.

Granted, under the charter, the UN Security Council may take action when it determines the existence of a threat to international peace and security. And nothing prevents it from identifying a government with no internal checks on its power that possesses or seeks to acquire WMD as a threat to the peace and taking measures against it. But articulating and acknowledging a specific duty to prevent such governments from even acquiring WMD will shift the burden of proof from suspicious nations to suspected nations and create the presumption of a need for early and, therefore, more effective action.

Consider, for instance, how recognizing a duty to prevent could have changed the debate over the war in Iraq. Under existing law, the Bush administration could justify intervention only by arguing that Iraq held WMD in violation of Security Council resolutions. Even though Saddam Hussein's Iraq was subject to special Security Council restrictions precisely because of its earlier illegal nuclear program and use of chemical weapons, the United States could not argue that Saddam posed a threat warranting intervention simply because of his absolute power, his past behavior, and his expressed intentions. Now suppose that last March, the United States and the United Kingdom had accepted a proposal by France, Germany, and Russia to blanket Iraq with inspectors instead of attacking it. Presumably those inspectors would have found what U.S. forces seem to be finding today—evidence of Iraq's intention and capacity to build WMD, but no existing stocks. Would the appropriate response then have been to send the inspectors home and leave Saddam's regime intact? The better answer would have been to recognize from the beginning the combined threat posed by the nature of his regime and his determination to acquire and use WMD. Invoking the duty to prevent, the Security Council could have identified Iraq as a subject of special concern and, as it was blanketing the country with inspectors, sought to prosecute Saddam for crimes against humanity committed back in the 1980s.

The inability to prevent WMD proliferation by dangerous regimes is a concern that has confounded at least the last three U.S. administrations. President George H.W. Bush defined the issue in terms of "outlaw" states, to distinguish regimes that followed international rules from those that defied them. President Bill Clinton used the term "rogue states" until 2000, when his administration began referring to "states of concern" to signal that the goal of U.S. policy was eventually to reintegrate states, if not their dictatorial rulers, into the international system. The present administration's use of the term "axis of evil" suggests a sterner version of the first Bush administration's approach. It leaves little room for diplomacy, forcing the United States to either advocate regime change or do nothing.

All these approaches, moreover, miss a key point. It is not states that are the danger, but their rulers—a relatively small group of identifiable individuals who seek absolute power at home or sponsor terrorism abroad. These rulers and their regimes can be identified by evaluating their behavior according to criteria already documented in the UN system: the rule of law and human rights; rights of association and organization; freedom of expression and belief; and personal autonomy and economic rights. The international system remains uncomfortable distinguishing one country from another, but such distinctions are already embedded in the UN system and they should be emphasized as the basis for effective international action to deal with the dangers we now face.


In the wake of Somalia, Haiti, Rwanda, Bosnia, and Kosovo, a halting process of revising old rules to meet today's threats has begun. In the fall of 2002, Secretary-General Kofi Annan repeated a challenge he first made to UN members in 1999, urging the Security Council to discuss "the best way to respond to threats of genocide or other comparable massive violations of human rights." Although the Security Council has yet to heed Annan's call, the Canadian government did, appointing former Australian Foreign Minister Gareth Evans and Annan's Special Adviser Mohamed Sahnoun to head a distinguished global commission of diplomats, politicians, scholars, and nongovernmental activists. In December 2001, the commission issued a report, titled "The Responsibility to Protect," that took on nothing less than the redefinition of sovereignty itself. The Evans-Sahnoun Commission argued that the controversy over using force for humanitarian purposes stemmed from a "critical gap" between the unavoidable reality of mass human suffering and the existing rules and mechanisms for managing world order. To fill this gap, the commission identified an emerging international obligation—the "responsibility to protect"—which requires states to intervene in the affairs of other states to avert or stop humanitarian crises.

This concept challenges the traditional understanding of sovereignty by suggesting that it implies responsibilities as well as rights. According to the commission, sovereignty means that "the state authorities are responsible for the functions of protecting the safety and lives of citizens and promotion of their welfare;" that "the national political authorities are responsible to the citizens internally and to the international community through the UN;" and that "the agents of state are responsible for their actions; that is to say they are accountable for their acts of commission and omission."

The commission's boldest contribution, however, was to argue that the responsibility to protect binds both the individual states and the international community as a whole. The commission insists that an individual state has the primary responsibility to protect the individuals within it. But where the state fails to carry it out, a secondary responsibility to protect falls on the international community acting through the UN, even if enforcing it requires infringing on state sovereignty. Thus, "where a population is suffering serious harm, as a result of internal war, insurgency, repression or state failure, and the state in question is unwilling or unable to halt or avert it, the principle of nonintervention yields to the international responsibility to protect."


By the time the Evans-Sahnoun Commission issued its report, in December 2001, much of the world was focused on the most dramatic of the world's new threats: an emerging breed of catastrophic terrorism. The prospect that al Qaeda or a comparable group might gain access to WMD drove the Bush administration in the fall of 2002 to announce a doctrine of preemption in its National Security Strategy. In the ensuing controversy, humanitarian concerns took a back seat to the imperatives of national security, narrowly defined. But today the links between the two sets of issues, especially the need to tackle them with proactive strategies, are becoming more evident.

The commission's effort to redefine basic concepts of sovereignty and international community in the context of humanitarian law are highly relevant to international security, in particular to efforts to counter governments that both possess WMD and systematically abuse their own citizens. After all, the danger posed by WMD in the hands of governments with no internal checks on their power is the prospect of mass, indiscriminate murder. Whether individuals are targeted for execution over time or vaporized in a single instant, the result is the same: a massive and senseless loss of life. We argue, therefore, that a new international obligation arises to address the unique dangers of proliferation that have grown in parallel with the humanitarian catastrophes of the 1990s.

The duty to prevent is the responsibility of states to work in concert to prevent governments that lack internal checks on their power from acquiring WMD or the means to deliver them. In cases where such regimes already possess such weapons, the first responsibility is to halt these programs and prevent the regimes from transferring WMD capabilities or actual weapons. The duty to prevent would also apply to states that sponsor terrorism and are seeking to obtain WMD.

This responsibility would apply to cases where the underlying set of agreements restricting WMD programs—the Nonproliferation Treaty (NPT), the Biological Weapons Convention, and the Chemical Weapons Convention—has not prevented a regime without internal checks from pursuing dangerous weapons, or when such a state withdraws from its obligations or cheats on them, or when a gap in existing rules needs to be filled to prevent such a regime from acquiring WMD or the means to deliver them.

Why emphasize the absence of internal checks on a government's power? We are not trying to distinguish "good" governments from "bad" governments, much less democracies from nondemocracies. Nor are we arguing that governments that have internal checks on their behavior always obey international law; they are bound by the same international norms restricting the development and use of weapons of WMD as are other states, and their compliance must be monitored too. But the behavior of open societies is subject to scrutiny, criticism, and countermeasures by opponents, at home and abroad. Also, existing nonproliferation agreements can circumscribe these states' behavior or, if political circumstances change dramatically, as they did, say, in South Africa in 1989 and in Argentina and Brazil in the 1990s, they can provide a path for states to give up their nuclear ambitions or, in the case of Pretoria, even their weapons.

On the other hand, the international community may only discover the danger posed by a closed society with no opposition when it is too late. In such cases, standard diplomatic tools are simply not up to the job. The greatest potential danger to the international community is posed by rulers whose power over their own people and territory is so absolute that no matter how brutal, aggressive, or irrational they become, no force within their own society can stop them. Their rule is absolute precisely because they have terrified, brainwashed, and isolated their populations and have either destroyed internal opposition or subdued it by "closing" their societies, restricting information as much as possible. Such leaders may simply seek to consolidate their power and to be left alone. But if they choose to menace other countries or support terrorist groups, it is far more difficult to find out what they are doing and take effective measures to stop them.

Just as the responsibility to protect cannot apply to all regimes that abuse their citizens' human rights, the duty to prevent cannot apply to all closed societies with WMD programs. To be practical, the duty has to be limited and applied to cases when it can produce beneficial results. It applies to Kim Jong Il's North Korea, but not to Hu Jintao's (or even Mao's) China. Existing nonproliferation tools, updated to close loopholes, would continue to apply to most countries, and the effectiveness of these rules would be reinforced by the perception of greater determination to deal firmly with the most serious cases.


The main international nonproliferation agreements stigmatize weapons or certain categories of weapons rather than regimes or leaders. Aiming at the weapons themselves rather than the states or regimes that develop or acquire them has been judged to be a more objective basis for international action. The problem with this approach is that its opening proposition is to treat North Korea as if it were Norway. This flaw has exposed the nonproliferation regime to abuse by determined and defiant regimes, especially those headed by dictatorial rulers. It is also the weakness that makes the NPT and, more broadly, the nonproliferation system vulnerable to charges that the only ones restrained by nonproliferation agreements are those nations that do not need restraining.

In truth, the NPT—the cornerstone of international efforts to prevent the spread of WMD—has helped stanch nuclear proliferation in the overwhelming majority of cases. It has also provided a pathway for states seeking to terminate their nuclear programs. But the NPT has not prevented a small group of determined states, including Iran, Iraq, and North Korea, from traveling down the nuclear path. These states, sometimes operating within the scope of the treaty, managed to develop advanced nuclear programs and, in the case of North Korea, the material for actually producing nuclear weapons.

How did this happen? In the name of fairness and due process, the NPT does not make it possible to meaningfully distinguish parties to the treaty that are in good standing from parties with clear nuclear designs. Parties may take action against a state that breaches the treaty only when clear evidence of the breach emerges—but by then their options may be limited and it may already be too late. Mohammed ElBaradei, director-general of the International Atomic Energy Agency, summarized the treaty's approach when he said recently of planned inspections into a clandestine uranium enrichment program in Tehran, "Let me point out here that what we do in Iran is what we do everywhere else. We treat Iran exactly as we treat all other member states." Of course, the agreement Iran struck with France, Germany, and the United Kingdom last October to halt its uranium-enrichment activities is a welcome development. But it comes too late. At this stage, international pressure might succeed in freezing Iran's existing program, but it is unlikely to reverse it.

Just as effective gun control in the United States requires both outlawing the most dangerous weapons and ammunition and applying more stringent controls on citizens with criminal records and other risk factors, an effective international nonproliferation campaign must target both WMD and international actors with suspect intentions. It must be based, in other words, on the recognition that leaders without internal checks on their power, or who are sponsors of terror, and who seek to acquire WMD are a unique threat. An international duty to prevent such regimes from acquiring WMD capabilities would allow preventive actions against them, such as bars on their participation in civilian nuclear programs, which have provided cover for illegal weapons programs in Iraq and Iran.

The recent agreement with Iran, though overdue, indicates a growing recognition that the one-size-fits-all approach articulated by ElBaradei is limited and that the legal rules on nonproliferation are evolving in the direction of a duty to prevent. The provisional agreement treats Iran very differently from "all other ... states." It recognizes that regimes such as Iran's, because they sponsor terrorism, repress democracy, and have clear nuclear designs, are not entitled to the same rights as other NPT members. It also demonstrates the range of preventive options available to deal with proliferation dangers.


Like intervention for humanitarian purposes, international action to counter WMD proliferation can take the form of diplomatic pressure or incentives, economic measures, or coercive action, often in combination. It can also incorporate new strategies, such as indicting individual leaders before the International Criminal Court or a special court for crimes against humanity, grave war crimes, or genocide when such charges apply, as they certainly would have with Saddam Hussein and possibly with Kim Jong Il. Still another alternative could be support for nonviolent resistance movements that are dedicated to democratizing their governments.

To be effective, incentives must be tailored to a state's particular needs. Where a state seeks WMD for their perceived deterrent value, security assurances by a nation or group of nations, formally organized or not, may make adequate alternatives. Where a state trades in sensitive technologies in exchange for hard currency, economic incentives—including assistance from international financial institutions, direct bilateral aid, and trade incentives—may be more appropriate.

Coercive action may take the form of economic penalties, including measures targeted at the state's rulers, their close associates, and their families. Curbs on financial flows or on sensitive trade that provides financial support for a state's weapons programs, including a crackdown on black-market trade, can be a very effective brake. (Counterfeiting and the illegal drug trade are believed to support North Korea's WMD programs.) Coercive action can also include embargoes, informal or otherwise, to block the transfer of weapons or relevant technologies and material. The Bush administration's Proliferation Security Initiative, an 11-nation effort to stop the shipment of WMD, their delivery systems, and related materials at sea, by air, or on land, is a step in the right direction. The initiative is intended to prevent the transfer of nuclear weapons, weapons materials, and missiles, as well as trade in contraband that supports these weapons programs. France and Germany are participating, despite their opposition to the Iraq war, but not China and Russia, whose cooperation is critical to making it an effective system.

A jugular issue is how to monitor compliance with any pledges to freeze or reverse nuclear programs. The Iraq experience suggests that UN inspections stopped being effective when Baghdad succeeded in dividing the Security Council and international support for them broke down. When UN Security Council Resolution 1441 revived the inspections, with the unanimous backing of the Security Council, Baghdad grudgingly cooperated with inspectors. Intrusive inspections endorsed by a united Security Council, backed up by the threat of force, may have worked better than they have been given credit for. Although it is easy to dismiss the effectiveness of inspections in closed societies, we need to review systematically the experience of international inspections for lessons learned. It may be that intense international pressure can make a system of rigorous inspections effective enough.

The Bush administration's announcement of a preemption doctrine set off alarm bells in the United States and abroad, chiefly because of the precedent it would set in terms of a unilateral determination that another state poses a sufficient threat to justify a preemptive strike. In truth, the use of force to preempt an imminent threat has always been part of international law, and it has been an option that the United States has held in quiet reserve and occasionally used. In cases in which terrorists appear poised to strike, preemption is clearly the preferred course of action.

Unfortunately, the preemptive use of force is often difficult to justify because clear evidence that a threat is imminent is rare. The U.S. strike on a pharmaceutical plant in Sudan in 1998 was intended as a preemptive strike against a facility suspected of producing chemical weapons, but evidence that activities there were illicit remains thin. Furthermore, preemption is usually impractical because suspected facilities are often difficult to spot or hit. States have taken precautions in recent years in response to the Israeli bombing of Iraq's Osiraq reactor in 1981 and to the NATO and U.S. bombing campaigns in the Balkans and the Middle East. Many facilities are buried in bunkers deep underground and dispersed over wide areas. They are especially difficult to locate in closed societies. This is not to suggest that the use of force should be discounted as ineffective but to highlight that the most effective action is preventive, because undoing a nuclear program is orders of magnitude more difficult than preventing one in the first place.

Nevertheless, as in the Iraq case, keeping force on the table is often a critical ingredient in making diplomacy work. It may be especially necessary for effective inspections and monitoring of WMD programs in closed societies. Force may be considered as part of an interdiction effort, may be targeted at specific dangerous facilities, or may be part of broader military action as a last resort.

The utility of force in dealing with the most serious proliferation dangers is not a controversial proposition. In a little-noticed statement last June, the EU announced a "strategy against proliferation," identifying "coercive measures, including as a last resort the use of force in accordance with the UN Charter" as one of its "key elements." Later that month, the G-8 group of leading industrialized countries, which includes Russia, approached the subject more gingerly but nonetheless agreed that WMD and the spread of international terrorism were "the preeminent threat to international security," and that force ("other measures in accordance with international law") may be needed to deal with them. And, as noted earlier, Kofi Annan himself called on the Security Council to develop criteria for the early authorization of coercive measures.


The contentious issue is who decides when and how to use force. No one nation can or should shoulder alone the obligation to prevent a repressive regime from acquiring WMD. Although the Security Council, still reeling from the Iraq crisis last March, now seems more interested in papering over its differences than in tackling these questions, it remains the preferred enforcer of collective measures. The unmatched legitimacy that the UN lends to Security Council actions makes it easier for member states to carry them out and harder for targeted governments to evade them by playing political games. On the other hand, rifts within the council allow states to pursue WMD to advance their programs, leaving individual nations to take matters into their own hands, which further erodes the stature and credibility of the United Nations.

Given the Security Council's propensity for paralysis, alternative means of enforcement must be considered. The second most legitimate enforcer is the regional organization that is most likely to be affected by the emerging threat. After that, the next best option would be another regional organization, such as NATO, with a less direct connection to the targeted state but with a sufficiently broad membership to permit serious deliberation over the exercise of a collective duty. It is only after these options are tried in good faith that unilateral action or coalitions of the willing should be considered.

In any event, the resort to force is subject to certain "precautionary principles." All nonmilitary alternatives that could achieve the same ends must be tried before force may be used, unless they can reasonably be said to be futile. Force must be exerted on the smallest scale, for the shortest time, and at the lowest intensity necessary to achieve its objective; the objective itself must be reasonably attainable when measured against the likelihood of making matters worse. Finally, force should be governed by fundamental principles of the laws of war: it must be a measure of last resort, used in proportion to the harm or the threat of the harm it targets, and with due care to spare civilians.


Humanitarian protection is emerging as a guiding principle for the international community. In the same vein, we propose a duty to prevent, as a principle that would guide not only the Security Council in its decision-making but also national governments in shaping their foreign policy priorities. Accepting this principle would require the United States to accept that specific criteria be met before preventive action of various types would be authorized. At the same time, the principle addresses many of the problems raised by the approach being advanced by other nations to deal with WMD.

The international legal rules governing nonproliferation, as well as those determining sovereign rights over a given population and territory, are evolving. Nations are interpreting old rules in new ways and trying out new practices in response to new threats. It is impossible to predict when and how a new international consensus will emerge, but now is the time to elaborate new principles that could structure a broad legal regime.

Ours is not a radical proposal. It simply extrapolates from recent developments in the law of intervention for humanitarian purposes—an area in which over the course of the 1990s old rules proved counter-productive at best, murderous at worst. The responsibility to protect is based on a collective obligation to avoid the needless slaughter or severe mistreatment of human beings anywhere—an obligation that stems from both moral principle and national interest. The corollary duty to prevent governments without internal checks from developing WMD capacity addresses the same threat from another source: the prospect of mass murder through the use of WMD, which have a destructive potential far beyond the control of any attacker.

In a world in which such governments can get access to the most devastating weapons and make them available to terrorists, we must take action. We are operating under a set of rules governing the use of force that were framed for a very different world, one of sovereign states, conventional armies, and noninterference in a government's treatment of its own citizens. These rules can continue to serve us well only if they are revised and updated to meet a new set of threats. Accepting a collective duty to prevent is the first step toward sustained self-protection in a new and dangerous era.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Lee Feinstein is Acting Director of the Washington Program of the Council on Foreign Relations. Anne-Marie Slaughter is Dean of the Woodrow Wilson School of Public and International Affairs at Princeton University and President of the American Society of International Law.
  • More By Lee Feinstein
  • More By Anne-Marie Slaughter