The president's plan to send a "surge" of troops into Iraq seems a clear instance of too little, too late. Two years ago, or better still from the start of the war, an open-ended commitment of many extra soldiers—certainly many more than envisaged for the new surge—would have done some good. (Whether it would have been enough to deliver success is another question.) But the situation has worsened dramatically since then. Supposing this modest and apparently temporary increase in forces happens, any improvement in security is likely, under current circumstances, to be brief.

To criticize the plan in this way may be too kind. It takes the proposal at face value. The new policy may, in fact, be an instance of something more squalid—a further expense of American lives in an effort not to secure "victory" (who any longer believes that that is possible?) but merely to buy political cover. Perhaps the White House is calculating that Democrats in Congress will find a way to block the new deployment, allowing President Bush, if he has the gall, to blame them for the steadily worsening failure on the ground. Perhaps the surge is intended to make it easier for the White House to blame the Iraqis if the situation fails to improve: "We tried everything, but they refused to be helped." Or perhaps, as some reports indicate, the new plan was simply a matter of coming up with something, anything, that was different from the recommendations of the Iraq Study Group, which reported in December. If so, additional lives stand to be sacrificed to little more than vanity.

It would be difficult to exaggerate the damage done by this war—and I say that with deepest regret, as somebody who advocated it. There is no easy exit. Much as one may wish that the war had never been started, it was started and that decision cannot be undone. The fact that the skeptics of 2003 have been vindicated does not mean that they are right to argue, as many of them do, that the best course now is the speediest possible withdrawal of American forces. That too would be a mistake. A too-rapid withdrawal, according to a preannounced schedule, with or without an intervening "surge," would be terribly dangerous.

Bad as the situation in Iraq may be, a precipitate retreat would make things worse, and maybe much worse. To be sure, it is an appalling thing to put American soldiers' lives at risk in such an uncertain and unpromising venture. But an unrestrained sectarian conflagration in Iraq—something that could still happen—would be a far worse outcome than anything seen so far, would gravely set back American interests, and would rest heavily on the country's conscience for years. It might even (as the Iraq Study Group pointed out) suck America back in.

So what is the answer? What is the least of the available evils? On the key question, the Iraq Study Group says that America's combat forces "could be out of Iraq" by early next year—without explicitly saying that they should be—leaving only 20,000 soldiers or so attached to Iraqi units as trainers, plus a further number of Special Forces (to be deployed against Al Qaeda and other American enemies, not as part of the effort to stabilize the country). The group said that the Iraqi government needs to make faster progress in approaching a variety of reconciliation "milestones"—in other words, that it should try harder to convince the minority Sunnis that their interests will not be trampled. If it fails, American military and economic support should be rolled back. The group's report also called for the U.S. to undertake new diplomatic initiatives directed at Syria and Iran, which have big stakes in the Iraqi conflict, and on Israel-Palestine.

The Iraq Study Group's report rightly expresses concern about the dire consequences of a too-rapid withdrawal. But isn't an almost-total withdrawal of combat forces within barely a year "precipitate"? And the report seems to envisage this outcome regardless of what the Iraqis do. If good progress is made on reconciliation, the violence will subside and American troops can be pulled out. Otherwise, the violence will not subside—and American troops will be pulled out. Where is the incentive in this for the majority Shiites to come to an accommodation with the Sunnis? And where, if America is leaving come what may, is its leverage over Iraq's internal politics?

Even now, the best course, in my view, would be a much bigger commitment of extra forces, of the kind that Sen. John McCain has advocated from the beginning, together with an undertaking that they would remain until security had been re-established. I believe this would offer the best chance of retrieving something—less than "success," to be sure, but more than the abject, disgraceful, and debilitating failure in prospect—from this mess. But I readily concede that, as the Iraq Study Group report said, this option is politically unsustainable. Americans voted against the war in November. The administration is discredited and commands no trust. The electorate regards it, with reason, as dishonest and incompetent. To put it mildly, that narrows the options.

The least-bad feasible option is to follow, by and large, the recommendations of the Iraq Study Group, minus the indication of a prompt withdrawal, whether or not preceded by one last push. Pursue the diplomatic possibilities, strive for progress on Israel-Palestine, apply all pressure on the Iraqi government to foster reconciliation with the Sunnis—and hang in there. To be sure, this course is a horrible thing to contemplate. It leaves American forces in danger with no clear prospect of success. The only thing to be said for it is that it is on balance less likely to fail, in a way that gravely damages American interests, than a policy of "announce a schedule and withdraw."

Reflecting on the course of this conflict, what lessons should advocates of the war such as myself draw from what has happened?

I cannot find refuge in saying that the Bush White House screwed up a basically good idea. That is too easy. There is no question that the administration made huge mistakes in the war's prosecution—above all, in committing too few resources to the endeavor, and in seeming entirely unprepared for the bad-case scenario that subsequently unfolded. But so much of what war advocates believed to be true has turned out to be wrong that I doubt that even a well-prosecuted war would have secured America's aims.

The whole misadventure began, of course, with the egregious intelligence failure over Iraqi weapons of mass destruction. I don't believe that the administration outright lied about this; they spun the evidence to make a stronger case, but I think they believed that the WMD were there. However, I take no comfort in this: I find it scarier than an outright lie. What use are the intelligence agencies if they can get such a thing wrong? How on earth did that happen? Democrats' desire to indict the administration for dishonesty—though amply justified in one way—is a pity in this respect. It has distracted attention from this astonishing intelligence failure. There should be a continuing national outcry over the error, not over the supposed lie.

The other false assumption was that the prospect of post-invasion civil war—a risk that many skeptics emphasized—could be contained. I vastly underestimated the shocking ferocity of sectarian animosity in Iraq. (After what happened in post-Yugoslavia, that is a difficult mistake to forgive.) If far more resources had been committed from the beginning to restoring internal security, the outcome might have been different.

But the rage that Iraq's rival Muslim communities are directing at each other makes one doubt it.

Having said that, and though I wish I had never supported the war, I still cannot regard it as completely misconceived. That view supposes that everything would have been fine if only we had left well alone. Key assumptions underlying the case for invasion turned out to be wrong—but not every assumption. Saddam Hussein was a sworn enemy of the United States; he was intent on acquiring WMD; and the sanctions regime that had imperfectly contained him after the Gulf War was breaking down. Had there been no war, a nuclear-armed Iraq led by Saddam would by now be a not-so-distant prospect. And a successful war would have empowered the United States to face down Iran, now also on the point of acquiring nuclear weapons. Those were eminently worthy objectives. In Iran's case, unfortunately, we may yet discover just how worthy.

Things have not worked out that way. On present accounting, just as the skeptics said, the direct and indirect costs of the war—including, not least, the costs to America's reputation and security—have greatly exceeded the benefits. And the world is a far more dangerous place as a result.