Ethics & International Affairs Volume 17.1 (Spring 2003): Roundtable: Evaluating the Preemptive Use of Force: Self-Defense in an Imperfect World [Full Text]

Mar 2, 2003

In his address at West Point on June 1, 2002, President George W. Bush appeared to be signaling America’s willingness to regard the mere possession of weapons of mass destruction (WMD) by potential enemies as grounds for an anticipatory war.1 Historically, however, a clear distinction has been drawn between preemptive and preventive, or anticipatory, war, with the latter regarded as illegitimate. The National Security Strategy announced by the president on September 20, 2002, was more conventional in its approach to preemption, but doubts remain as to whether the old distinction can be preserved. And this discussion is taking place in the context of a specific problem, namely the apparent desire of Iraq to obtain WMD and the determination of the United States, and, less clearly expressed, the UN Security Council, to prevent this from happening.2 At the time of writing (November 2002) this matter remains unresolved, but preemption has not received the attention it deserves. It is not only the United States that may engage in preemptive action; the possibility exists in other parts of the world and with respect to other conflicts, most noticeably on the Indo-Pakistani border.

The distinction between preemption and prevention is made by reference to the notion of an immediate threat.3 States, it is presumed, do not have an unqualified right to use force in international relations, but they do have the right to defend themselves—a right established in customary international law and reaffirmed in the UN Charter—and, crucially, are under no obligation to allow an aggressor to strike first. The right to preempt is thus an extension of the right of self-defense, if, and only if, it is indisputably the case that there is an imminent threat of an unprovoked aggression.

Prevention, on the other hand, involves action in response to some putative future rather than immediate threat and is not legitimate. The assumption is that states have other ways of responding to such potential threats, and the presumption against being the first to use force, which is enshrined in the UN Charter and based on agreements such as the 1928 Kellogg-Briand Pact and the 1945 Charter of the International Military Tribunal at Nuremberg, applies. Less often expressed, but equally compelling, is that in a world environment that is by its very nature insecure, a strategy of meeting all possible future threats is simply not viable; sufficient unto the day is the evil thereof.4


To illustrate the distinction between preemption and prevention, Israel’s strike against the Egyptian air force that began the Six Day War of 1967 was accepted by many to have been a preemptive act because, it is said, the Israelis had good reason to believe they were about to be subjected to an attack that might have been fatal had the enemy been allowed to strike first. Alternatively, it might be argued that the closure of the Straits of Tiran might have been seen as tantamount to a declaration of war against Israel.5 On the other hand, the Israeli strike on the Iraqi reactor at Osirak in 1981 was widely seen as not meeting the criteria for a legitimate act of preemption. The reactor was in the process of construction and could not be seen as an immediate threat.6 Both these judgments could, of course, be contested, but, in any case, the distinction aimed at is clearly visible.

From this distinction some obvious problems arise of both a prudential and a normative nature. How can states really know that an attack is imminent? How certain do they have to be before they may properly act? Are they obliged to justify their risk assessment, and if so, to whom? These are good questions, but they rest upon some unspoken assumptions about the way the world works. They assume that a quite strong domestic analogy holds, and that a state’s exercise of its right to self-defense can be judged in much the same way that the violent behavior of an individual might be judged in a domestic dispute, in which self-defense is usually qualified in terms of the nature of the threat and appropriateness of the response. This domestic analogy in turn rests upon a particular and contestable reading of the current state of international law and organization.

Suppose we lived in a world with a fully developed legal framework governing the use of force in international relations and outlawing its use as an instrument of foreign policy, and in which there were an international body charged with the maintenance of peace and security that possessed legitimate and effective decision-making procedures and the capacity to enforce its judgments. In such a world, the distinction between preemption and prevention would be clear-cut and of critical importance. Preventive war would be totally illegitimate because if one state believed that developments in another might threaten its security, it could refer the matter to the international body, where it would be investigated and dealt with appropriately. On the other hand, the possibility of the legitimate preemptive use of force remains as a theoretical concomitant of the residual right of self-defense. If a surprise attack were genuinely imminent and certain, and if there were insufficient time to refer the matter to the world body, since states are entitled to defend themselves, a state would be under no obligation to allow the aggressor to strike first. Given the assumptions made above, this is an unlikely scenario, but it would not be impossible.

The problem is that we do not live in this world. There is indeed an international legal order governing the use of force and a body charged with the maintenance of international peace and security, but the legal order is radically imperfect, as is the United Nations. Moreover, within the actual contemporary system states differ radically in terms of power, their commitment to the rule of law domestically and internationally, and in the responsibility and accountability of their governments.

International law condemns acts of aggression, but there is no court or other body with the power to rule on all disputes, and no international police force to preserve law and order. The UN Security Council is charged with the maintenance of international peace and security, but the veto power of the five permanent members means that even a wide consensus on a particular problem may not lead to action. The structure of the UN General Assembly is equally undemocratic, although in a somewhat different direction. The one state, one vote rule in the General Assembly allows a two-thirds majority to be constructed out of well under 10 percent of the world’s population. Moreover, many, perhaps a majority of, UN members do not have democratic internal decision-making structures, and the elites who control their UN votes may well be motivated by concerns that have nothing to do with the issue at hand. Usually the rich and powerful are able to buy votes without too much difficulty. Sadly, there is no reason to think that anyone—the P-5, the nonpermanent members of the Security Council, or the members of the General Assembly—is likely to approach matters of peace and security in a fair-minded way. Partly for this reason, when resolutions are actually passed they are quite frequently ignored, especially if they go against the interests of the rich and powerful.

What all this means is that whereas there is a legal order that makes a clear distinction between preemption and prevention, unfortunately there can be no guarantee that the ideal substitute for preventive war—effective action by a legitimate international body—can be achieved. Here we approach the heart of the problem: In an imperfect world, how should states act with respect to potential threats to their security?


In the absence of an ideal international political framework that could restrain states in all but those few cases in which a preemptive attack was truly necessary, three propositions should guide the preemptive use of force. First, it might be assumed that it would be good to live in a world with an effective legal order and effective and legitimate global institutions, where the kind of existential insecurity of the present world no longer existed. Second, since this is not the case, it would be unwise and imprudent for states to behave as if it were. Given that there are some predatory states, and that it cannot be assumed that the existing international institutional structure is capable of handling the problems they pose, it may sometimes be necessary for nonpredatory states to act unilaterally or outside the official institutional structure. Third, states should, as far as possible, try to act in such a way that they encourage the transformation of the world into one in which effective institutions do exist, or at a minimum do not make such a transformation more difficult.

States have a strong self-interest in preserving a clear distinction between preemption and prevention and in making it clear that the latter is rejected as a national strategy. Based on the above propositions, states should not allow themselves to pursue the chimera of absolute security. The fact that a state might pose a threat in the foreseeable future should be seen as a reason to construct defenses against it, to pursue a containment strategy, because to attempt instead to eliminate all such threats is to commit to an endless series of wars to end all wars. States should as far as possible involve the United Nations and other institutional structures in the process of containment, although it must be acknowledged that sometimes this will be impossible. In the same vein, states must be aware of the possibility of a sudden act of aggression on the part of predatory states and must be prepared to react in such circumstances. Reaction in this case could involve preemption if the information that an attack is imminent is sufficiently reliable. States cannot, in the last resort, delegate the task of deciding on the reliability of this information to some other body. If a state has reasonable grounds for believing itself to be in danger of being the victim of aggression, it has the right to act in self-defense. This right would be strengthened if the state in question had rejected prevention as a general strategy because cases in which preemptive action was genuinely necessary would then be less likely to be misinterpreted.

In the extreme, states cannot be expected to delegate to others the decision of whether sufficient grounds exist to justify preemption, but they should be obliged to explain their reasoning in public and to seek wide approval for their actions. However imperfect the existing institutional structure may be, it is the only one there is, and there is no possibility that it will become more effective if it is ignored in a crisis or treated with contempt by those who have the power to act alone. What is important in this context is not the formal legal implication of particular votes, but the search for approval and consent. For example, the fact that NATO was unable to achieve a UN Security Council resolution endorsing its action in Kosovo in 1999 is perhaps less significant than the fact that a motion condemning its actions was defeated by twelve votes to three, the twelve including countries such as Malaysia and Argentina, on whose support NATO cannot always rely. The veto power of Russia and China prevented a positive vote, but NATO was able, nonetheless, to claim a degree of legitimacy for its action.7 A technically positive vote by the UN Security Council may be no more important than the political legitimacy that stems from a more general sense of approval for one’s actions. Legitimacy should be an issue. It is best if states behave prudently, and in extreme circumstances they will act as judges in their own cause. But they must justify themselves to the wider world through all available channels.


How does this translate into a consideration of potential action against Iraq? By the time this article appears it is likely to have been overtaken by events, but the principles that apply can still be set down. As of now, the United States and the United Kingdom have presented ample evidence that the current Iraqi regime poses a serious threat to the security of the region, its own people, and thus, indirectly, the world. Twice in the last quarter century Saddam Hussein has launched unprovoked acts of aggression against his neighbors, and well over a million people have died as a result. Should Iraq come into possession of serious stocks of WMD the possibility of another such act of aggression is quite high, as is the possibility that such stocks could be transferred to third parties. The world should respond to this potential danger, as the UN Security Council has recognized in Resolution 1441. What is less clear is that the present strategy of containment has failed. If it is accepted that insecurity is an inherent feature of the current international order, and that the search for total security is chimerical, then good reasons are required to reject the policy of containment in favor of immediate action. So far such reasons have not been provided.

There is, however, one argument that is not usually employed by those who wish to see preemptive action but which is quite compelling. Even though it is probably not the case that the often-quoted figure of half a million Iraqi children killed by sanctions is anywhere near accurate—all such figures come ultimately from the Iraqi regime, even if they are given credence by other bodies—and although the responsibility for civilian suffering can be laid at the door of the Iraqi regime itself rather than at that of the United Nations, the main victims of containment have been innocent Iraqis and not the regime itself.8 Nonmilitary sanctions almost always involve attacking the interests and sometimes the lives of “soft” targets—women, children, civilians—and policies of containment that rely on them are morally suspect. The fact that such sanctions are so often regarded as morally superior to acts of violence that target the regime itself and its military defenders is a striking commentary on the inability of many progressively minded Westerners to think clearly on this subject. That the innocent are suffering is usually presented by the antiwar movement as a reason for abandoning the sanctions regime on which containment is based, but it can equally, and much more sensibly, be seen to provide the best available justification for moving away from containment and employing force to bring about a change in the Iraqi regime.

An overthrow in this case would involve a particular kind of humanitarian intervention, and the politics and ethics of humanitarian intervention are rarely seen in the same context as discussions of preemptive war. In both cases preemption is involved, but on a different moral basis. Normally, preemption is discussed in the context of a potential act of aggression, that is, an immediate threat to state sovereignty, the assumption being that the territorial integrity and political independence of the sovereign state are the supreme values protected by any international order. However, for the last fifty years such sovereignty norms have been accompanied by a growing body of norms that challenges the right of sovereign states to do whatever they like to their own people, summarized by the idea of universal human rights.9 This development should change the nature of the discussion. Just as preemptive action may sometimes be necessary to protect state sovereignty, so military action is sometimes required to protect a people from their sovereign. If a legitimate policy of containment designed to prevent aggression has the unintended consequence of imposing great suffering on the innocent Iraqi people—suffering the regime has done nothing to alleviate and may be promoting for propaganda reasons—that may provide a reason for moving to a more active policy.

There are prudential issues involved here and the costs of carrying out a forcible change of the regime may rule out action in this or similar cases, in the same way that a policy of anticipatory war to prevent the development of WMD might prove unviable. The point of raising a forcible change of regime in this context is not to advocate a humanitarian intervention but to widen the range of considerations that ought to be addressed and to challenge the compartmentalization that leads to the isolation and privileging of the rights and interests of sovereign states. In any practical case in which preemptive or anticipatory action is proposed, the rights and interests of individuals ought to enter into the equation, and it cannot be assumed that these rights are subsumed by the rights of states.


* An earlier version of this article was presented at a roundtable at the Carnegie Council on Ethics and International Affairs, New York, N.Y., November 14, 2002. I am grateful to the other participants for their input, and to Joel Rosenthal and Tony Lang of the Carnegie Council for the opportunity to contribute to this roundtable. I am also grateful to Michael Donelan, Christopher Hill, and Nicholas J. Wheeler for comments; none of the above, of course, is responsible for the views expressed in this article.
1 “Remarks by the President at the 2002 Graduation Exercise of the United States Military Academy,” June 1, 2002; available at
2 “The National Security Strategy of the United States of America September 2002”; available at www.white
3 Michael Walzer, Just and Unjust Wars, 2nd ed. (New York: Basic Books, 1992), presents this distinction with great clarity.
4 See David C. Hendrickson, “Towards Universal Empire: The Dangerous Quest for Absolute Security,” World Policy Journal (Fall 2002), pp. 1–10.
5 For a judicious recent discussion, see Michael Oren, Six Days of War: June 1967 and the Making of the Modern Middle East (New York: Oxford University Press, 2002).
6 See the strong condemnation contained in UN Security Council Resolution 487, June 19, 1981.
7 There is a large literature on the Kosovo case. See, e.g., Ken Booth, ed., The Kosovo Tragedy: The Human Rights Dimensions (London: Frank Cass, 2000).
8 See Robert Fisk, “The Dishonesty of This So-Called Dossier,” Independent, September 25, 2002, for an uncritical presentation of the charge that “we” have killed half a million Iraqi children; available at The British government has vigorously contested this interpretation. See, e.g., Tony Blair, “Prime Minister’s Iraq Statement to Parliament,” September 24, 2002; available at
9 For a general discussion of this issue, see Chris Brown, Sovereignty, Rights and Justice (Cambridge: Polity Press, 2002).

You may also like

AUG 9, 2023 Podcast

Nuclear Ethics for this Moment

This panel explores ethical questions surrounding nuclear weapons and builds upon a symposium published in the most recent issue of "Ethics & International Affairs."

MAY 10, 2023 Journal

Ethics & International Affairs Volume 37.1 (Spring 2023)

The editors of "Ethics & International Affairs" are pleased to present the Spring 2023 issue of the journal! The highlight of this issue is a symposium organized ...

MAY 4, 2023 Article

A New Era for "Ethics & International Affairs"

The editors of Carnegie Council's quarterly journal "Ethics & International Affairs" are proud to announce the beginning of a new era in our publishing history. Starting ...

Not translated

This content has not yet been translated into your language. You can request a translation by clicking the button below.

Request Translation