¹ The notion of an “imminent” attack is somewhat vague but is usually understood to include not only situations in which missiles or warplanes have been launched but have not yet struck their targets, but also situations in which forces have been mobilized with apparently aggressive intentGoogle Scholar.

² Walzer, Michael, Just and Unjust Wars: A Moral Argument with Historical Illustrations (New York: Basic Books, 1977Google Scholar).

³ In the absence of authoritative judicial determination, international law is subject to a variety of interpretations. Some might hold that under current international law preventive action can be legitimate because the Genocide Convention obligates states/parties to take action to prevent genocide. However, this claim is contestable because, according to Article 103 of the UN Charter, the prohibition against preventive use of force trumps all other treaties, including the Genocide Convention. It might also be held, on the contrary, that preventive force is not permissible under international law even with Security Council authorization, except in order to respond to a “threat to international peace and security.” For our purposes, it is not necessary to take a stand on these legal questions. We merely follow most authorities in treating the legal status quo as prohibiting preventive use of force without Security Council authorization, but permitting it if such authorization has been providedGoogle Scholar.

⁴ For the classic statement, see Morgenthau, Hans J., Politics among Nations: The Struggle for Power and Peace, 6th ed., revised by Kenneth W. Thompson (New York: Knopf, 1985), p. 12Google Scholar. Advocates of the National Interest view need not be moral skeptics; instead they may hold that state leaders have one supreme moral obligation—to serve the national interest—and that this obligation overrides all other principles of morality. Those who espouse the National Interest view may disagree among themselves, of course, as to whether a policy of recourse to preventive force, or any particular decision to use preventive force, is in fact likely to serve the national interest or be detrimental to it.

⁵ “National Security Strategy of the United States of America September 2002,” p. 6; available at http://www.whitehouse.gov/nsc/nss.pdf.

⁶ We explore the flaws of the Just War Blanket Prohibition in more detail belowGoogle Scholar.

⁷ The position we endorse is sometimes called Moderate Cosmopolitanism, which allows one to give a limited priority to the interests of one's own nation and does not require strict impartiality. It is a liberal form of cosmopolitanism, since it emphasizes the basic human rights of all persons. One could imagine other kinds of cosmopolitanism, such as Marxism, that forbid discrimination on grounds of nationality, but do not focus on human rights. For a philosophical justification for the assumption that there are human rights, see Buchanan, Allen, “Human Rights,” in Justice, Legitimacy, and Self-Determination: Moral foundations for International Law (forthcoming, Oxford: Oxford University Press, 2003CrossRefGoogle Scholar).

⁸ The humanitarian law of war can be seen as an imperfect institutionalization of some of the most important jus in bello principles of just war theory. However, given the distinctive risks of the preventive use of force, special institutional arrangements are required for responsible decisions to use force preventively, over and above better compliance with rules of humane warfare that are designed to apply to all uses of forceGoogle Scholar.

⁹ The prima facie argument seems to imply that preventive action to protect basic human rights is not only permissible but also obligatory. However, since we seek to advance a feasible proposal, our discussion of institutional arrangements does not assume such an obligation. The successful implementation of the accountability regime we propose might be an important step toward the more ambitious goal of institutionalizing an obligation to prevent massive violations of basic human rightsGoogle Scholar.

¹⁰ By beginning with the assumption that the use of force is sometimes permissible to stop massive violations of basic human rights already under way, we are of course implicitly rejecting the absolute pacifist view that the use of military force is never justified because it inevitably involves harm to innocent persons. We share this assumption with the four perspectives that we criticizeGoogle Scholar.

¹¹ See Loewy, Arnold H., “Conspiracy,” in Criminal Law in a Nutshell (St. Paul, Minn.: West Group, 2000), p. 260Google Scholar; and Dressier, Joshua, “Inchoate Offenses,” in Cases and Materials on Criminal Law, 2nd ed. (St. Paul, Minn.: West Group, 1999), p. 765Google Scholar.

¹² In one sense, this objection points out the obvious: that tradeoffs exist in policy between timely action and certainty about the necessity of action. The longer one waits, the more information one is likely to gain; but by the same token, one's ability to affect the situation may be reduced. Britain and France had better reason in 1939 than in 1936 to regard Germany as a threat to their security, but the cost of averting the threat was higher in 1939Google Scholar.

¹³ Gordon Lauren, Paul, “Coercive Diplomacy and Ultimata: Theory and Practice in History ,” in George, Alexander L. and Simons, William E., eds., The Limits of Coercive Diplomacy (Boulder, Colo.: Westview Press, 1994), p. 23Google Scholar.

¹⁴ Schelling, Thomas C., Arms and Influence (New Haven: Yale University Press, 1966), pp. 69 – 78Google Scholar.

¹⁵ Independent International Commission on Kosovo, Kosovo Report: Conflict, International Response, Lessons Learned (New York: Oxford University Press, 2001Google Scholar).

¹⁶ Schedler, Andreas, “Conceptualizing Accountability,” in Schedler, Andreas, Diamond, Larry, and Plattner, Marc F., eds., The Self-Restraining State: Power and Accountability in New Democracies (Boulder, Colo.: Lynne Rienner, 1999), p. 17Google Scholar. The word “punishment” could be interpreted in too formal and restrictive a way. We accept this definition only with the proviso that it refers more broadly to a penalty.

¹⁷ Waltz, Kenneth N., Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979Google Scholar).

¹⁸ Thucydides, , The Peloponnesian War, Richard Crawley, trans., bk. V, para. 84 (New York: Modern Library, 1982), P . 351Google Scholar.

¹⁹ For a more general argument, see Keohane, Robert O., “Governance in a Partially Globalized World,” American Political Science Review 95, no. I (March 2001), pp. 1–13Google Scholar. See also Brennan, Geoffrey and Buchanan, James M., The Reason of Rules: Constitutional Political Economy (New York: Cambridge University Press, 1985), p. 30Google Scholar; and Frohlich, Norman and Oppenheimer, Joe A., Choosing Justice: An Experimental Approach to Ethical Theory (Berkeley: University of California Press, 1992Google Scholar).

²⁰ In the World Trade Organization states behave in this way. They are reluctant to put forward legal arguments to defend their own practices that could be used in the future by other states to justify protectionist measuresGoogle Scholar.

²¹ The analogy of conscientious objection may be helpful. Conscientious objectors may avoid military service but are expected to subject themselves to at least equivalent burdens and risks as those who do carry arms, by serving as medics, for exampleGoogle Scholar.

²² The Fourth Geneva Convention relative to the Protection of Civilian Persons in Time of War of August 12, 1949, and other provisions of international law specify the obligations of occupying powers. These include ensuring respect for basic human rights, including restoration and maintenance of law and order and provision of food and medical care to the population. Under our proposal these obligations would not be diminished, but occupying powers whose actions had been judged by the ex post review to be unjustified would have their ability to make decisions severely constrainedGoogle Scholar.

²³ Here it might be objected that the strongest states might agree ex. ante to pay compensation if an ex. post review found their preventive actions to be unjustified, but simply fail to pay up. To avoid this problem, assets could be escrowed ex ante. Such an arrangement was used in the agreements that ended the U.S.-Iran hostage crisis in 1981 and that established the U.S.-Iran Claims Tribunal. See Carswell, Robert and Davis, Richard, “Crafting the Financial Settlement,” in Christopher, Warren et al. , eds., American Hostages in Iran: The Conduct of a Crisis (New Haven: Yale University Press, for the Council on Foreign Relations, 1985), pp. 21–34Google Scholar; document in ibid., Appendix D, pp. 405–22.

²⁴ See Akerlot, George “The Market for Lemons,” Quarterly Journal of Economics 84, no. 3 (August 1970), pp. 488–500CrossRefGoogle Scholar.

²⁵ I. L. Claude likens the UN Security Council veto to “a fuse in an electrical circuit,” reflecting the wise conviction that “in cases of sharp conflict among the great powers the Council ought, for safety's sake, to be incapacitated.”Claude, I. L., Power and International Relations (New York: Random House, 1962), p. 160Google Scholar.

²⁶ Article 51 of the UN Charter states that Members of the United Nations that respond with force to an armed attack shall “immediately” report their actions to the Security Council. Our institutional model 2 goes significantly beyond Article 51 in three ways: it expands the scope of a reporting requirement beyond the case of self-defense against an actual armed attack to cover preventive force; it specifies that the report must be made by an impartial body and must explicitly address issue of fact and of reasonableness in the justification of the use of force; and it attaches penalties in the case of an unfavorable report. Article 51, in contrast, allows the party using force to report on itseIf specifies no criteria of evaluation for the report, and mentions no consequences of an unfavorable reportGoogle Scholar.

²⁷ This necessity raises the problem of an ex post veto by the intervening state or states of the findings of the impartial commission or penalties implied by those findings. This problem could be solved by adapting a suggestion recently made by Thomas Franck, who points out that Article 27 provides that “decisions of the Security Council on procedural matters shall be made by an affirmative vote of nine members.” That is, on procedural matters, the veto is inapplicable. Ex ante the Council could decide that votes on the composition and report of the impartial commission, and on any recommendations for penalties, would be considered procedural. (Professor Franck's proposal concerns specification of what constitutes a “material breach” regarding UN resolutions concerning Iraq. See his “Inspections and Their Enforcement: A Modest Proposal,”American Journal of International Law 96 (October 2002), pp. 899–900Google Scholar.

²⁸ When the United States invaded Iraq in 2003 without Security Council authorization, the Bush administration's chief justification for doing so was to destroy weapons of mass destrucltion, which it said were present in that country. It can be argued that as of nine months after the invasion began, the United States has failed to show either that there were weapons of mass destruction in Iraq at the time of the invasion or that there was good reason to believe that they were present. If this still appears to be the case after all the evidence is in, the problem of achieving credibility for the preventive use of force will be even greater in the future. The case for the ex. ante and ex. post accountability mechanisms we propose will be all the strongerGoogle Scholar.

²⁹ The United States gained Security Council approval for the first Gulf War in 1991, but not for the invasion of Iraq in 2003. Other countries paid approximately 80 percent of the cost of the former conflict; so far, the United States is paying over 80 percent of the costs of the latterGoogle Scholar.

³⁰ We are indebted to Michael Doyle for the suggestion that led to this proposalGoogle Scholar.

³¹ Machiavelli, Niccolo, The Prince (1532), ch. VI.Google Scholar

³² We are indebted to Jeremy Waldron for this pointGoogle Scholar.