The easiest way to understand why we are where we are in Kosovo is to recall what Bill Clinton said in early 1991 after Congress had authorized President Bush to attack Iraq. Just before American fliers and soldiers went into action, the then-governor of Arkansas reflected: "I guess I would have voted with the majority if it was a close vote. But I agree with the arguments the minority made." There you have the reasoning of a principled man who knows his own mind and is willing to run huge political risks to defend deep convictions.

Eight years later, the result is a war unleashed with a continuous barrage of high-minded rhetoric about the obvious evil of Slobodan Milosevic and his lieutenants -- and yet pursued with so many self-imposed limitations that its chief victims are the very Kosovars whom we claim to be helping. These limits flow from Clinton's reluctance to endanger his standing in the polls or Democratic prospects in the next election.

This contrasts unfavorably with Bush's behavior in the Persian Gulf War. Although Bush must have known the polls, his actions defied them. In November 1990 -- as the troop buildup proceeded -- a Gallup Poll asked respondents to choose between a ground attack and an air war on Iraq. By 74 to 17 percent, they preferred an air attack. In January 1991, as war loomed, a CBS/New York Times survey asked whether "the United States should start action" or wait "to see if the trade embargo and other economic sanctions work." It was a draw: 47 to 46 percent in favor of waiting.

Clinton's timid pursuit of the war unintentionally makes the case of many critics -- most prominently, ex-secretary of state Henry Kissinger in Newsweek and Owen Harries, editor of the National Interest, in the New York Times. They argue that Kosovo never engaged important U.S. interests. This is surely true. If it weren't, the war would be conducted with fewer restraints. But intervention (they say) has damaged U.S. foreign policy. It has hurt relations with Russia and China, threatening cooperation on collective interests (from nuclear proliferation to North Korea). And the inept use of U.S. force will make it harder in the future to influence or deter other nations with the threat of force.

These questions will linger, even if Milosevic is ultimately forced to withdraw his troops. But there is another issue: the role of morality in foreign policy. Kosovo may represent the first war in U.S. history that has been undertaken mostly for moral reasons. Clinton has defended his policies mainly in moral terms, describing Serb atrocities and Kosovar suffering. Writing in the New York Times, he put it this way: "The Balkans are not fated to be the heart of European darkness, a region of bombed mosques, men and boys shot in the back, young women raped, all traces of group and individual history rewritten or erased."

This revulsion reflects the feelings of most Americans, especially after regularly seeing TV images of exhausted refugees and Kosovar corpses. And Clinton is correct that America's inability to prevent all atrocities doesn't mean that it can't limit some. In Kosovo, the result is a war of good intentions. The problem is that Clinton ignores the hard question: If you start a war for moral reasons and then wage it in a halfhearted and hypocritical manner, is your behavior still moral?

By hedging his bets, Clinton has managed to make a moral crusade immoral. At the outset, he faced two equally distasteful choices. One was to stay out, because (as Kissinger and Harries argue) a war in Kosovo would be more trouble than it was worth. This would have subjected Clinton to harsh criticism that he was complicit in the slaughter of innocent Kosovars. The other choice was to intervene with ample force -- including low-level bombing and ground troops -- to improve the odds of helping the Kosovars. The dangers here were high U.S. casualties and hostile public opinion.

So Clinton rejected these choices and selected a policy to lessen his political risks. He would intervene but in a way that minimized U.S. casualties. No doubt this was a nifty political solution that, if the war were a public works bill, might be defensible. But here, the consequences have been indefensible both on foreign policy and moral grounds.

Morally, Clinton shifted almost all the risk of American policy onto the Kosovars. Their plight could hardly be worse -- and, arguably, might be better -- if Clinton had done nothing. Once NATO started bombing, there were no restraints on Milosevic. Clinton's early and repeated denials that ground troops would ever be used gave Milosevic no reason to stop. This was reckless; in its effects, it was almost criminal.

Clinton's defenders will say that he was (and is) limited by practical problems: divisions within NATO; public opinion; the Balkans' harsh terrain. All this is true. It's also true in every crisis. Bush faced huge difficulties in the Persian Gulf: reluctant allies (except for Britain); the delicacy of bringing Arab nations into the anti-Saddam coalition; vast distances to the gulf. Margaret Thatcher faced problems in retaking the Falklands. Leadership consists of overcoming problems.

Up to a point, this is Clinton's failure. (Query: If he had been president in 1991, would Saddam Hussein still be in Kuwait?) But because Clinton accurately gauges public opinion, the failure also highlights a foreign policy dilemma: How to remain faithful to U.S. values while recognizing the limits of U.S. power?

Americans want to project U.S. beliefs onto the rest of the world. But this faith in the superiority of their values also makes Americans suspicious of foreigners. This creates cycles of internationalism and isolationism. Clinton has wanted to stand for everything that is good and right. But he hasn't staked much to achieve what is good and right. Probably most Americans feel the same way. What Kosovo teaches is that good intentions are not enough. Carelessly deployed, they lead to bad results.