It is difficult both to be good and to be powerful. This seems to be the common view among statesmen, sages, poets and thinkers. A core thesis among thinkers of the realist persuasion has been that in foreign affairs, being good may in the end be bad for the people you serve, and that moral ends may best be served by thinking in terms of power and how it should be preserved, instead of aiming to do directly what seems morally good. This lesson is repeated in the works of Machiavelli, Morgenthau, Kissinger and many others. Realism is about power, and though barren and inadequate as a description of the way international society functions, it is at least consistent. Likewise, liberal internationalism, though its proponents have sometimes mistaken aspiration for reality, is also consistent. But the attempt to combine the two, as Charles Krauthammer did in these pages ("In Defense of Democratic Realism", Fall 2004), presents difficulties in both theory and practice.

One difficulty with democratic realism is the problem of power in a democratic age. Once, we knew what power looked like. It possessed a big army and a big navy. You exercised power by beating someone else's army and taking their land, their money, their women. Sometimes you took over their territory and ruled it and them. In the last decades, these habits have died out among democracies: In an advanced industrial society, land is more a burden than an asset. We have left behind the static aristocratic society in which wealth appeared to be fixed, so that you could become rich only by robbing others. Today, peace and trade provide a better return than war and looting. From the point of view of wealth creation, war is a double negative. It destroys assets and does so at great expense.

In a democratic age, ruling others is problematic. The notion that all men are equal does not sit comfortably with empire. Nevertheless, the idea of spreading the democratic system of government has great attractions. We need an orderly world, and democracies are in the long run more stable than dictatorships. Besides, like it or not, our democratic values are universal. If all men are equal, then oppression anywhere is offensive; it may not threaten our security, but it threatens our self respect, for we are involved in mankind.

The theory that democracies do not fight each other attracts adherents as different as President Bush and Immanuel Kant. There are also skeptics like Alexander Hamilton, who pointed out that Rome and Athens were no less warlike for being republics. These were imperfect democracies, it is true, but so in one way or another are all democracies. Perhaps the fairest conclusion is that the no-war-between-democracies thesis needs more time to establish itself; up to the present, the sample of modern democracies has been toosmall. But for mature democracies, it does at least seem to have a plausible logic: Most people are cautious about voting for policies that may involve them risking their lives. And the evidence continues to accumulate.

The theory that well-governed societies will not produce terrorists is manifestly not true: Timothy McVeigh, British-born suicide bombers, and the Japanese terror cult Aum Shinrikyo are among the many counter-examples. But in a well-governed country, there will be a better chance of obtaining the support of the majority of the population against the terrorists. Legitimacy is usually one of the keys to success. Accountable police and intelligence services ought in the end to be more efficient. But here also evidence is too thin for much certainty.

Nevertheless, a world of well-governed countries with accountable executives, responsible assemblies and independent judiciaries seems instinctively preferable to any of the alternatives. We feel more comfortable with Japan mastering nuclear technology than with someone like Saddam Hussein doing the same. Europe at the end of the 20th century is safer than Europe at its start. East Asia looks to be on the way to being happier and safer than the Middle East.

The argument of the democratic realist thus has a compelling simplicity and logic: Democracy is desirable, perhaps even imperative, for our security, and America is now a dominant power in a way that is without precedent. So American power should be used to promote democracy. The problem with this argument is that American power--or at least the dimension of power in which America is most evidently dominant--is military power, and it is questionable how useful this is in creating democracies. There are several systemic reasons for thinking it may not be the best instrument for this goal.

Democratic systems and military systems are in many respects opposites. Democracy is bottom up; military is top down. In military systems, hierarchy and rank are fundamental; in democracy, the starting point is that all men are equal. Democracy is about due process, rights, limits to power; military systems work, necessarily, on the basis of obedience to orders. These differences become even more marked when an army has invaded someone else's country. An army gets its way by violence and by the threat of violence, very different from the processes of law embodied in democratic states. And whereas the principle of equality before the law is basic to democracies, there is nothing less equal than the invading soldier and the local civilian. Thus, although a foreign country may invade with the best of intentions and may bring with it professors of politics to explain democratic theory, what it does is fundamentally undemocratic. Its words may say the right things, but its actions tell exactly the opposite story.

Behind this lies an even more fundamental question. How much use is military power in a democratic age? What is the point of being the only superpower, the sole owner of the unipolar moment, if you cannot maintain control of a single medium-sized state or even a medium-sized town? Of course, the story in Iraq is far from over, and the United States, with Iraqi help, may well in the end establish order throughout the country. But--and this is the point--if it does so it will be with Iraqi help. No doubt even without Iraqi help, the United States could take full control by flooding the country with troops and using whatever degree of force was required. This would be in the logic of military power, which after all is about violence and threat. But it would not work. First of all, it would not work because the United States itself is democratic and its people would not permit it, and second, it would not work because Iraq, like every other part of the world, is infused in a primitive way with a democratic ideology. At the beginning of democracy is the idea of self-determination, the idea that you should be ruled by your own people and not by foreigners. In the violent culture of the Middle East, this may be expressed in insurgency or support for insurgents. In Central Europe, for years it was expressed in sullen resistance, underground movements and ironic humor. There too, forty years of rule, through surrogates but resting ultimately on military force, demonstrated the weakness of military power in a democratic age.

As an authoritarian state, the Soviet Union had less difficulty in being brutal, though even its willingness to use force declined over the forty-year occupation. For America it is not so easy. The United States may be the most powerful state since Rome, but unlike the Roman Empire, it is democratic, and its people will not tolerate Roman methods (solitudinem faciunt, pacem appellant). Military force still has its uses, but running or transforming other people's countries is not one of them. To see power only in military terms is a fundamental error in world politics.

Krauthammer might reply that while the theory might not work, the practice does. "We have succeeded in the monumental task of reconstructing Germany, Japan and South Korea." It is true that all of these countries have had an American force present over a long period and also that during this period they have become stable democracies. But the causation is not as simple as Krauthammer suggests. In fact, the three countries mentioned had very different histories, and the United States played a different role in each.

Germany under the Nazi Party was a tyranny, and its overthrow was a requirement for the re-establishment of democracy. Without the American contribution in World War II, this would probably never have happened. The same is true of the Soviet contribution. With the occupation, the difference between the American and Soviet approaches became clearer: In the Western zones--which were British and French as well as American--democracy and an open society were refounded; in the Soviet Zone a communist dictatorship was installed. But in the West, democratic institutions were not established but re-established. They were not new. Different parts of Germany had different histories: The state of Baden had universal suffrage from the early 19th century, while Prussia was still an autocracy. It is not clear whether the German state in the time of the Kaiser should be called a democracy or not. Everyone had a vote, but they were of unequal weight. By modern standards it would not qualify as democratic, but that is also true of most other countries of the period. The Weimar constitution, however, was undoubtedly democratic: For example, it gave votes to women a year before the 19th Amendment did the same in the United States and some decades before many other European states.