In his essay “The Ethics of America's Afghan War,” Professor Richard W. Miller vigorously condemns the United States' continued counterinsurgency efforts in Afghanistan. To him, the moral costs do not justify the prosecution of the Afghan war. I concede at once that an assessment of costs and benefits may well lead to that conclusion; indeed, this is true of any war. However, in my judgment Professor Miller has failed to make his case. Simply put, his two central theses rest on dubious predictions and, more important, are morally objectionable. He proposes, first, that the United States withdraw from the country after brokering a settlement under which the Taliban would be allowed to rule over part of the country. Writes Professor Miller: “The United States has . . . a moral duty . . . to achieve a political settlement, conceding control of the Pashtun countryside to the Taliban” (p. 103). Second, he calls on the United States to abandon its delusions of grandeur and humbly accept that it can no longer achieve its objectives by wielding hegemonic power.Footnote ¹ According to Miller, the United Stated should pursue instead a policy of “graceful decline” (p. 125). I address those claims in turn.

The Taliban: A Study in Evil

The United States would be perpetrating a major injustice if it enabled the Taliban to rule over any part of the territory and over any person. Professor Miller does not merely say that installing the Taliban in the Pashtun territory is something that the world should, regrettably, accept as unavoidable: he claims that the United States has a moral duty to install the Taliban in the Pashtun territory. Should the United States do this it would be an accomplice, this time by action rather than omission, in the crimes perpetrated by a despicable regime. Professor Miller concedes, as he must, that “Taliban control of the Pashtun South and East would impose much injustice” (p. 114). Even this concession sounds euphemistic, so let us briefly recall some facts.

The Taliban's human rights record during the period it wielded power in Afghanistan is one of the worst in post–World War II history. This is a regime that executed girls as young as eight years old for the crime of attending school.Footnote ² The Taliban routinely stoned adulterous women to death and executed, flogged, and mutilated people for minor infractions. This is not hearsay: the Taliban proudly conducted these punishments in public. Suffice it to say that for the United States to enable this regime to consolidate its rule over any population would be a truly obnoxious foreign policy decision. Those concerned for the plight of women should be particularly alarmed by the prospect of reinstalling the Taliban in power, even in a smaller territory.Footnote ³ One of the cardinal principles of a morally justified foreign policy is the obligation not to cooperate with tyranny.Footnote ⁴ Yet this is exactly what the United States would do, should it follow Professor Miller's advice.

Despots can be classified into one of two categories. Most of them are opportunistic despots. Their paramount concern is to remain in power, as well as acquire ever more wealth and influence. Saddam and Stalin belong, I believe, in this category: the terrible harm they inflicted was opportunistic, motivated by their desire to keep or increase their power or wealth. But there are despots who are principled. These tyrants do evil not out of selfish motives, but because they act out of evil principles. Adolf Hitler, Jorge Videla, and the Taliban belong in this category.Footnote ⁵ These persons are typically fanatic and immune to corruption or other temptations. Here, being principled is a vice, not a virtue, because the value of fidelity to principle is entirely parasitic on the value of the principle. Hitler's tenacious efforts to exterminate the Jews, even when it was obvious that the war was lost and that the extermination efforts were sapping the energy of his Reich, is an example of how far principled evildoers can go in implementing their maxims. Hitler's quasi-religious belief in the superiority of the Aryan race and his tenacity in following this belief to the end is what made him terrifying.

The Taliban's proud public display of power in pursuance of rigid Islamic maxims and their firm defense of their crimes in the face of world criticism put them in the same category. Their attachment to the principles in the name of which they commit these crimes makes them particularly odious. Like Hitler, these despots are not corruptible, and that makes them worse, not better. Some may think the opposite, that the fact that they do these things sincerely in the name of Islam makes them less open to criticism. Perhaps, the argument goes, that is part of a culture that we Westerners do not fully understand. Professor Miller does not say this openly, but he comes close when he argues that the “political distance” between the United States and Taliban injustice is a moral reason against intervening.Footnote ⁶ I suggest exactly the opposite: there is more reason to fight principled evildoers than opportunistic evildoers, because the former, unlike the latter, cannot be bribed. The only way to stop principled evildoing is by violence.

Professor Miller has several replies to the humanitarian objection. The first is that the Taliban cannot be defeated. He says that the effort to defeat the Taliban with a long, nationwide counterinsurgency “would reveal an Afghan array of forces, interests, and passions sufficiently favorable to the Taliban to make them an unappeasable force with enduring nationwide strength” (p. 115). Were this true, Miller would be right that the United States should settle for a second-best solution. Except in very rare circumstances, even a justified war should not be waged unless victory is at least probable. Should the Taliban be the awesome military force that Miller supposes, then the United States should make an effort to reach the best settlement it can to protect its own interests and the human rights of the hapless Afghans. In that case, prosecuting the war would be foolhardy and irresponsible.

Of course, there are no certainties in war, and no one can predict with complete assurance that NATO will be able to eradicate the Taliban. However, I do not believe that the Taliban is the fearsome power that Professor Miller depicts. While the insurgents have shown unexpected resilience, they are no match for the Western coalition, provided that NATO invests the necessary resources to win and does not try to fight this war on the cheap, and provided that the Afghan government and people eventually assume the responsibility for governance. Certainly the U.S. government and NATO believe that victory is possible, and I have no reason to give less credibility to these sources than to those upon which Professor Miller relies.Footnote ⁷ Most important, NATO's strategy for the reconstruction of Afghanistan has been endorsed by the international community at several junctures. All of these sources emphasize the establishment of human rights and the rule of law within Afghanistan, as well as workable arrangements for security and development.Footnote ⁸ They assume that eradicating the Taliban, while difficult, is possible.Footnote ⁹

Professor Miller does not need to hold the extreme view that the United States cannot defeat the Taliban. He can simply claim that victory has become harder to achieve than was initially thought, costlier than originally expected. Let us concede this for the sake of argument. Here my reply is different: the determination and ferocity of the enemy is not a reason to quit; rather, it is a reason to fight harder. Imagine that in 1942, facing the German counteroffensive in the Ardennes, the Allies would have sought a compromise with the enemy, analogous to the one Professor Miller is proposing here. Imagine an agreement that would have allowed the Third Reich to rule over the Netherlands and Belgium while providing for German withdrawal from France. Crucifying the Dutch and the Belgians would have been the price of peace. I take it that such a solution would have been morally unacceptable, even had the Allies reasonably believed that the Germans would observe the agreement. If the Western coalition has a just cause in Afghanistan, such as rescuing Afghans from a barbaric regime, then the fact that the enemy has redoubled its efforts is an insufficient reason to abandon the fight. In war as in life, perseverance in the face of adversity is a virtue, not a vice.

There is an additional reason to apply these considerations to Afghanistan. The Western coalition initially invaded the country to suppress the al-Qaeda threat. With al-Qaeda presumably weakened or defeated, the Western coalition has an obligation to help Afghans reconstruct their ravaged country. It is not as if the United States is considering whether to invade Afghanistan. The United States is already there, and the sole issue is whether to stay the course or to quit. Quitting would have terrible consequences for everyone except for the Taliban and its collaborators. In particular, the consequences are dire for those Afghans whom the United States, under Miller's proposal, would throw into the claws of the Taliban—much like the Allies fed the Czechs to Hitler in 1938 in an effort to cut their losses.

Professor Miller's argument appears initially plausible because of how he characterizes the U.S. involvement in Afghanistan. On his view, the United States is meddling in Afghanistan to pursue its own imperial interests without much regard for the wishes and needs of Afghan citizens. But suppose that we describe the military operations differently: the United States is helping Afghans to fight insurgents bent on reinstalling a brutal regime in that country. I take it that Professor Miller would not object to the Afghans themselves fighting a counterinsurgency war to eradicate the Taliban. If such action is morally justified (as I assume everyone thinks it would be), so is aiding the Afghans in this task. This is not intervention anymore, because the NATO troops are there at the behest of the Afghan people. The NATO troops are performing an essential state function that the Afghans cannot perform themselves: defeating vicious insurgents and restoring civic order as the condition for liberty. The troops are attempting, in other words, to restore and maintain the political order because the local government cannot do it properly. Viewed in that light, the NATO troops are hired by the Afghans to defend them against the Taliban; they are the agents of the Afghans. Of course, if this were just an attempt by a political faction to enlist the United States in defeating a political enemy, then Professor Miller would have a point. The U.S. armed forces should not be at the service of any foreign political party that cannot defeat its adversaries in the polls. But this is not the right way to view Western involvement in Afghanistan. We know that the Taliban is not just an ordinary political party: it is a villainous organization. This means that the moral equivalence presupposed in the noninterventionist argument (that the United States should not intervene in foreign political contests) is inapplicable here. U.S. action is in part humanitarian.Footnote ¹⁰ One of its purposes is to save millions of innocent persons, and in particular women and children, from the prospects of despotic rule, the horrific nature of which is not speculative but certain.

Professor Miller disagrees. He thinks that there is a decisive difference between the Afghans trying to defeat the Taliban and the United States trying to help them. He claims that the “political distance” between the United States and Taliban injustice makes it morally problematic for the United States to intervene (p. 116). I do not fully understand this argument. In a sense it is obviously true that there is political distance between the United States and the Afghans. The United States differs from the Afghans who request its aid in two respects. First, the United States does not have long-term “property” interests in Afghanistan. Second, the United States, as Professor Miller points out, owes a fiduciary duty to the American people. But none of these factors is decisive in judging the morality of intervention. The first factor is relevant to the morality of post-intervention: because of its lack of legitimate long-term “property” interests in Afghanistan, it would be wrong for the United States to exact unreasonable political or territorial concessions (beyond those related to legitimate U.S. security concerns, such as ensuring that al-Qaeda does not return) in exchange for its help. The second factor is more relevant, as popular support for the war at home may sometimes be a precondition for its legitimacy. But popular support does not seem to be a major issue in this instance; for while public support for the war fluctuates considerably, President Obama won the presidency in part by convincing the American public that the war in Afghanistan was justified. Where applicable, these two factors (U.S. “foreignness” and a lack of U.S. popular support) do not disqualify U.S. help to the Afghans in their quest for human dignity; they just establish some limits to that help. Political distance did not disqualify the United States from liberating its European allies in World War II, nor did it preclude India from stopping the ongoing genocide in Bangladesh.

Professor Miller's suggestion that political distance “increased the seriousness of Taliban deaths as a moral reason against continued American violence” is especially unconvincing (p. 116). Of course, all of us would prefer conflicts to be resolved with no one getting killed, but if the Taliban is an unjust enemy at war, it is a fair target of lawful killing, political distance or not. In fact, killing a Taliban insurgent is even more justified than killing an enemy soldier in a conventional war, because the Taliban insurgent, unlike the typical conventional soldier, fully endorses the unjust cause for which he is fighting. Miller's claim here presupposes that there is a communal relationship between the Taliban and their victims worth protecting. But, as I have tried to show elsewhere, this idea is false.Footnote ¹¹ It derives from a communitarian prejudice that refuses to die: the idea that victims of oppression are in part responsible for their suffering. The relationship between a tyrant and his victims is no closer than the relationship between any criminal and his victim.

Viewing NATO forces as a surrogate for the Afghan people allows us to evaluate the costs that worry Professor Miller. For the costs in terms of human life are those that any society, unfortunately, must incur in order to prevent criminals from taking over the country. Those costs would not be any lesser should the Afghans attempt to eradicate the Taliban on their own. Given the nature of the Taliban, the Afghans are lucky to have NATO forces at their disposal to free their country from this scourge. In any war, we want the costs in innocent lives to be as low as possible. But it does not seem to me that defeating the Taliban could be achieved at a lesser cost, nor does it seem to me, above all, that the costs in terms of human lives and human suffering could be any less than the costs the Afghans and their liberators are now incurring. On the contrary: the costs for the large number of people who would have to endure Taliban rule for the foreseeable future would be, I suspect, much higher than the costs of the current war. Moreover, these costs should not be calculated only in terms of human lives. One should add the cost of tyranny, the fear of living constantly in the knowledge that any minor deviation can bring the secret police, the torturers, or the executioners.

Crucially, the Taliban, not NATO, kill most of the innocents. In a report released as I write these lines, the United Nations Assistance Mission in Afghanistan (not a particularly pro-U.S. source) has announced that the Taliban are responsible for 75 percent of civilian deaths.Footnote ¹² This is a topic in the philosophy of war that calls for deeper analysis: how should a just warrior compute the different kinds of victims when making assessments of proportionality? The different categories are: deaths of innocent civilians; deaths of non-innocent civilians (collaborators and supporters of the enemy); deaths of the warrior's own soldiers; and deaths in the ranks of the enemy. Belligerents bear the highest responsibility for the deaths of innocent civilians, which is permitted in only narrow instances specified by an appropriate version of the doctrine of double effect. Killing enemies is permitted unless unduly treacherous, while the moral status of the other two categories (own soldiers and non-innocent civilians) is unclear, although I would certainly expect a just warrior to care more for his own soldiers than for culpable civilians (assuming those can be identified). In addition, there are two different categories that cut across the previous ones: deaths caused by the just warrior and deaths caused by the enemy. The just warrior bears a higher moral responsibility for the deaths he causes than for those caused by the unjust enemy, and any calculation of cost in terms of human lives must reflect this moral difference. In Afghanistan, then, the fact that many of the casualties are enemies, and the fact that a high percentage of casualties are caused by the Taliban, must be considered in the moral evaluation of the costs of the war. These distinctions, of course, do not relieve NATO forces from the moral responsibility for the killings they cause, especially of innocents, but they suggest a more nuanced method of appraising moral costs.

Professor Miller concedes the humanitarian rationale in principle. He writes:

As an aside, this is a strange concession, because if the Taliban cannot be defeated, as Professor Miller believes, why would the desire for rescue make a difference? The United States, on Miller's own view, cannot rescue these people because it cannot win the war.

Miller rejects the humanitarian argument in this case, however, on the grounds that the Afghans themselves prefer ceding control of a portion of their territory to the Taliban rather than enduring the presence of foreign troops. Relying on a 2010 poll, Miller claims that a majority of people in the regions that are more likely to be ceded to the Taliban favor the settlement. While I have misgivings about the poll (Who was asked, exactly? Precisely what were they asked? What about the view of all Afghans, and not just those in those regions? What about women?), I will assume, gratia argumentandi, that those results are accurate. According to Miller, “a substantial majority were willing to accept the concession of control [to the Taliban] (63 percent in Helmand, 58 percent in Kandahar)” (p. 116). Helmand and Kandahar, we are told, are the two regions most likely to be ceded to the Taliban. The United States, according to Miller, should hear the voice of the people. A majority of those interviewed would rather live under the Taliban than suffer the presence of American troops. The popular will, reflected in majority sentiment in the Pashtun territory, supports a settlement with the Taliban. The attempt to rescue Afghans from injustice is misguided because the Pashtuns do not want to be rescued. Apparently, Miller thinks that the costs would be acceptable if the eventual victims of the Taliban's murderous governance asked for help, since we would then be balancing those costs against real benefits. But, we are told, this is not the case. Defeating the Taliban is too costly because the supposed benefit, rescuing Afghans from injustice, would not be reaped (since the victims do not want to be rescued). Thus, there is no benefit against which to balance the costs of the war. I think Professor Miller has to make this argument; otherwise, his admission of the importance of consent (in the form of desire for rescue) would be incomprehensible, given that he does not believe the United States can prevail. Let us call this argument the Popular Will argument. Any humanitarian intervention—any rescue mission—requires the consent of the rescued as evidenced by the ascertained desires of the majority of the population. If referenda are not available, opinion polls are an acceptable substitute.

The Popular Will argument, however, is fatally flawed. To see why, I will concede at once that a liberal government should not try to rescue persons from tyranny if those persons do not want to be rescued.Footnote ¹³ But this principle identifies the victims, and the victims alone, as those whose consent matters. Professor Miller's numbers (again, taking them at face value) mean that 37 percent of persons in Helmand and 42 percent of persons in Kandahar have withheld their consent to be ruled by this despotic regime. The persons in the majority in those regions who would welcome the Taliban are not entitled to consent on behalf of those who do not want to be ruled by the Taliban. Those who support the Taliban are collaborating with a murderous regime, so they do not have a valid communal interest in the tyrannical governance of their community—no more than the Germans who voted for Hitler had a valid communal interest in the persecution of the Jews. The 63 percent of Afghans in Helmand and the 58 percent of Afghans in Kandahar, those whose opinion Professor Miller considers decisive, are accomplices of the Taliban in a criminal enterprise. If they take arms to defend the Taliban against NATO they will be fighting an unjust war. They would be as responsible as the Taliban for the atrocities that, we know, will ensue. Surely, in Afghanistan there are groups who benefit from the Taliban's persecution of others. Maybe many men think they benefit from the crimes that the Taliban perpetrates against women. These are the rent seekers of the worst kind, those who capture the Taliban's terror machine for their own purposes. It is simply unacceptable for Professor Miller to vindicate these persons' desires to oppress their fellow citizens. Under any plausible democratic theory, tyranny is not one of the things a majority can impose on a minority.

When we consider Afghanistan as a whole, there is ample evidence that a vast majority of Afghans do not want the Taliban to return to power. According to the same poll Professor Miller cites, only 4 percent of Afghans would rather have the Taliban in power today, while 82 percent support the current form of government.Footnote ¹⁴ This evidence from the entire Afghan territory, I take it, steers Professor Miller toward the partition solution. Again, preventing the Taliban from oppressing people cannot be decided by polls. If, say, 51 percent of Afghans, and not just 4 percent, wanted the Taliban to return to power, they would still not have a greater right to foist this criminal regime on their fellow citizens. But the fact that more than 80 percent reject the Taliban surely provides evidence that Afghans might appreciate freedom more than observers give them credit for.

Professor Miller might amend the Popular Will argument as follows. Those persons who reject U.S. help are not necessarily collaborators. Rather, these are Afghan citizens who sincerely oppose both the Taliban and the presence of foreign troops for moral reasons. They might concede that the regime is murderous, that they are willing to oppose it politically, even violently. At the same time, they might abhor the presence of foreign troops even more, no matter how benign. They may prefer, in short, to lose their fight against the Taliban rather than to win with the aid of the United States. These persons are not collaborators and their opinion must surely count.

While this amended version of the argument saves these persons from the infamous label of collaborators, it fails for similar reasons. These persons (the majorities that Miller invokes) lack standing to consent on behalf of their fellow Afghans who have to endure the yoke of the Taliban. The Afghan man who rejects the attempt by U.S. troops to rescue a woman about to be stoned alive does not have the moral power to veto the rescue. His consent does not count. Only that woman, the victim, could (conceivably) refuse to be saved.

There is a different reason why Professor Miller's proposal should be rejected. The proposal amounts to a de facto partition of the country, because even if it would treat the Pashtun territory as a decentralized region and not as a sovereign nation, the territory would be governed by the Taliban and the Taliban only. If so, quite apart from humanitarian considerations, it would be wrong for the United States to impose this kind of partition in Afghanistan. As far as anyone can tell, the regions of Helmand and Kandahar do not have a moral or legal right to secede from Afghanistan. The Pashtun territory is part of Afghanistan, which means that it belongs collectively to the Afghans, not just to the people who live in the territory. The appropriate unit to consult for purposes of self-determination is therefore the whole of the Afghan adult population. The Afghan people have long-term property interests over the whole territory of the country, and any attempt at partitioning the country should be the subject of collective, nationwide consent. Even less acceptable is for the United States, which does not have long-term property interests in Afghanistan, to enforce a partition against the wishes of the Afghan population. True, the literature has recognized the exception of remedial secession. When people in a territory are seriously mistreated by the central government or by rival ethnic groups, secession is a morally acceptable way to escape that persecution. But, of course, Professor Miller's proposal turns this exception on its head: the de facto partition he proposes will enable tyranny and persecution by the Taliban.

Managing U.S. Decline

Professor Miller's second argument is that the grand strategic goal apparently pursued by the United States in Afghanistan, the preservation of American hegemony, must be rejected, whether or not the war serves that goal. Professor Miller joins others who have recently predicted the inevitability of American decline.Footnote ¹⁵ We are told that this is due to various factors, including the ascent of such powers as China, India, and Brazil. Writes Professor Miller: “Rather than seeking to maintain U.S primacy as long as possible, a morally justified grand strategy would seek to manage America's decline in global power so that the transition is as tranquil and orderly as possible” (p. 123).

The debate over American unipolarity is large indeed, and I can only confine myself to a few comments in reply.Footnote ¹⁶ The argument that the United States should give up her dominant position may take two forms. The United States should accept that her decline is inevitable, or the United States should renounce hegemony because it is morally wrong. Let us address both versions of the argument.

If decline is inevitable, then of course the United States should manage it gracefully and in an orderly fashion, much as the United Kingdom did following the disintegration of the British Empire after World War II. But I doubt that observers can predict something as complex and multifaceted as the decline of the United States with the information they have. To be sure, people often make these predictions, partly because it is fun, partly because they want the United States to decline, and partly because the costs of being wrong are low. Grandiose predictions of this kind are no more than cheap talk. The public did not call to task those who predicted the demise of capitalism and the rise of socialism, nor were those who failed to predict the fall of communism ever held accountable. But the decline of a world superpower is an event determined not by the forces of nature, but by the choices of millions of persons in world markets, as well as by many big and small-decision makers around the globe. Let us not forget that in the 1970s many people predicted that the United States was headed to an inevitable decline, and that the Vietnam fiasco and the Watergate scandal, coupled with the rise of socialism in the developing world, marked the sure demise of the United States and the values that it espoused. It was a time, we were told, for compromise, for accepting that liberal democracy and capitalism as we knew them were things of the past. We all know what happened: communism deservedly collapsed and the United States presided over three decades of global prosperity and increased freedom. Free markets and democracy (with ups and downs) became fashionable again. So the record of anti-Western doomsayers has not been very good.

The second version of the argument opposes U.S. hegemony on moral grounds. The argument is familiar: any kind of domination is objectionable, and the relations among nations should not be an exception. When powerful nations dictate to weak nations they violate the international law principle of sovereign equality and the moral injunction that people should not impose their ends on others.

In his article Professor Miller is as harsh on the U.S. role in the world as he is in his recent book on global justice,Footnote ¹⁷ claiming that the United States does not act on moral considerations but on purely selfish ones (p. xx). He cites, for example, the provision of U.S. aid to the opponents of a new pro-Soviet regime “in order to draw the Russians into the Afghan trap” (quoting Zbigniew Brzezinski), U.S. efforts to prevent a decisive victory for either side in the Iraq-Iran War, and similar examples. These facts should be enough, he thinks, to disqualify any attempt to justify American hegemony, especially of the violent kind. They also invalidate ab initio any pretense to vindicate the Afghan war as morally justified.

As a preliminary matter, the U.S. government acts morally when it pursues American national interest by acceptable means. This is because the United States government has a fiduciary duty toward its citizens. It is an agency hired by the people to represent them in the international arena. This is an area where the domestic analogy (that nations are like individuals) does not work very well. When I act selfishly you may think I am not a very admirable person. But when the government implements the mandate that the people have given it—namely, to protect and defend their interests and their liberty in a world replete with competitors and threats—then the government is doing the right thing. The government would be acting immorally if it pursued its own interest internationally, but if it pursues the national interest it is not acting selfishly but honoring a contract, the vertical contract between government and people. So the fact that in the examples cited by Miller the U.S. government was justifying its behavior in a non-humanitarian way—that is, by invoking the national interest—does not in and of itself make the act amoral.

There is another reason why national-interest motives do not disqualify an act as a candidate for moral justification. Acts lacking in moral motivation may be substantively valuable. First, an act performed for selfish reasons may be beneficial. A U.S. military commander in Afghanistan who attacks an enemy position may believe that he is upholding American national interests, but, as it turns out, defeating the Taliban will advance the cause of freedom for many people. His action is objectively valuable regardless of motives. Some of the examples that Professor Miller marshals can be characterized in this way. U.S. officials said that they wanted to support anti-Soviet factions in Afghanistan in order to trick the Soviets into invading, but perhaps countering Soviet power was the right thing to do regardless. Second, an act may be well motivated as to its ultimate goal but intrinsically objectionable. In these cases, an evaluation of the act must rely on a moral analysis of ends and means. Any violent act must be evaluated under the doctrine of double effect or similar frameworks. I agree with Professor Miller that there are cases where the value of the ultimate end does not suffice to justify the badness of the act. So when the United States does something, say, prevents Iran and Iraq from defeating each other (as noted by Miller), we must ask ourselves if doing this was justified in the light of any valuable end the United States might pursue. Sometimes it will, sometimes it will not. (In this particular example, it is far from clear that the United States should have taken sides with the Ayatollah Khomeini or with Saddam Hussein.) In Professor Miller's cold war examples, the United States was trying to contain a criminal regime that threatened to achieve world hegemony and suppress freedom globally. The interest of the United States was self-regarding because the Soviet Union threatened the West, but it was also morally valuable because the United States stood for freedom, while the Soviet Union stood for the exact opposite. I am firmly convinced that by over-relying on the national-interest rationale, many politicians and academics understated the case against communism.

To be sure, some or even many of the acts that the U.S. government did to reach that final goal were morally objectionable. These acts were immoral because they violated the principle that we should not use immoral means even to achieve moral ends. But that does not mean that the efforts to contain the Soviet Union were morally worthless. To see why this objectionable behavior is not a wholesale indictment of American hegemony, one has only to compare those efforts with the anti-Western efforts by the Soviet Union: these were objectionable all the way through, because they employed immoral means to achieve immoral ends. The upshot of all this is that the acts of governments should be evaluated with the ordinary tools of morality. At best, Professor Miller has shown what most people believe, that valuable ends do not justify intrinsically immoral means. But he has not shown that a hegemon that pursues valuable ends always uses immoral means and, most important, he has not shown that a government that sometimes uses immoral means to achieve moral ends is as bad as a government that pursues immoral ends, especially when there is no reasonable alternative.Footnote ¹⁸

More generally, the assumption that hegemony is per se immoral is questionable. As many authors have argued, international hegemony arises in a semi-anarchical world, and its merits or demerits have to be judged by examining what the hegemon does, and by comparing the hegemonic arrangement with the available alternatives. The alternatives are the state of nature, a balance of power, or a world ruled by a supranational organization, such as the United Nations. The first, the complete absence of world order, is particularly undesirable for well-known Hobbesian reasons. The inordinate energy that nations would have to spend in defensive efforts would sap whatever productive endeavors they could undertake. A system of balance of power is unstable for the same reasons that oligopolies are. Sooner or later, the oligopolists will be irresistibly tempted to defect. The third possibility, where most crucial decisions are made by a supranational organization, is also undesirable in view of the present alignment of forces and values in the world. The legitimacy deficit and the agency costs of such a system are staggering (although I will not attempt to prove this here).Footnote ¹⁹ So we are left with hegemony. As many authors (and history) have shown, the presence of a hegemon may be highly beneficial because it may ensure the provision of global public goods.Footnote ²⁰ Whether the other nations have consented or not, they benefit from the stability, tranquility, and predictability that the hegemon ensures. The hegemon also benefits, of course, but we must reject the simplistic idea that hegemony is a one-way relationship where the hegemon gets to do what it wants and dictates to others what they should or should not do. Other nations free ride on the hegemon on a number of issues, from defense to trade to the environment.

Finally, the claim that international hegemony is wrong because it is a form of domination suffers from an overly statist bias. It is wrong for me to dominate you, to ignore your own choices and substitute my ends for yours. But this reasoning does not always carry over to international affairs. States are not persons. When we say that a state “dominates” another, what we mean is that the government of state A tells the government of state B what to do or not to do, under some threat. But it may be the case that the government of state A is telling the government of state B to stop unjustly dominating its citizens. In that case, the right description is that the government of state B is unjustly dominating individuals, its own citizens, and the government of state A is putting pressure on the dominator, the government of state B, to stop dominating others. As I write these lines, the United States and others are putting pressure on the government of Libya to stop attacking its own citizens. Muammar Qaddafi protests that the West is trying to dominate Libya, to tell Libya what to do. But this is false, of course: the West is rightly exercising pressure on Qaddafi to stop mistreating Libyans. So the domination argument is plausible only in a specific (though admittedly important) class of cases: when the hegemon puts pressure on a weaker government to do something that the latter has a moral right not to do. The upshot is that the hegemonic relationship has to be disaggregated in order to determine whether or not the pressure exerted by the hegemon is morally justified.Footnote ²¹ That cannot be settled in advance by condemning all domination.

This leads us to American hegemony. Once we accept, perhaps reluctantly, that having a hegemon is preferable to the alternatives, the question is who should be the hegemon. The main reason to support American hegemony is that any other plausible candidate is likely to be worse. Would critics of the United States really prefer China to be the hegemon? Would they have preferred the Soviet Union to become the hegemon? The United States is the least bad of all possible hegemons because American institutions and culture embody liberal values. The acts that Professor Miller condemns are misguided attempts to advance liberal values and protect liberal cultures against the onslaught of illiberal forces. So people should be very wary about a world in which the United States cedes its dominance to other actors. Perhaps new actors will emerge that are better providers of global public goods and better defenders of liberty than the United States. But absent that, if the nature of global anarchy means that a hegemon will emerge perforce, I would rather have the United States in that role than any of its current or past competitors.

Evidence of how less bad the United States is than others can be collected in Afghanistan itself. While of course one wishes that no innocent civilians would ever be hurt in war, NATO, led by the United States, has made considerable efforts to reduce civilian casualties.Footnote ²² The United States spends considerable energy helping to build a reasonable political system in Afghanistan, one that would, if things go well, result in the protection of basic human rights and the rule of law.Footnote ²³ True, in the path to achieving these worthy goals, U.S. forces occasionally perform objectionable acts, either by negligence or by design, and we should all be alert to condemn those and to put pressure on the U.S. government to do the right thing in the right way, to match worthy ends with moral means. But as far as hegemons go, the Americans are far, far above the armies of destruction of past imperial conquerors. In fact, were the United States to decide to nakedly pursue its national interest, as did such past hegemons as the Austro-Hungarians, the Huns, and the Soviets, it would achieve its goals much more efficiently (and terribly). True, the commitment to liberty and human dignity is often hard to reconcile with the dilemmas posed by foreign threats of various kinds, such as terrorism and tyranny. While this difficulty does not excuse immoral behavior, the United States is as good a hegemon as humanity is going to get. If the war in Afghanistan is in part an effort to prevent American decline—for example, by not allowing our enemies to claim victory (as they would if the Taliban survived)—then that does not invalidate those efforts, provided of course that there are enough moral reasons to wage the war. Rather, the goal of preserving global liberal dominance reinforces the legitimacy of the humanitarian mission in Afghanistan.