There is nothing much wrong with Iraq that can't be improved by having fewer American troops there. So contend outgoing Secretary of Defense Donald Rumsfeld, the anti-war stalwart Rep. John Murtha, and, apparently, the James Baker-Lee Hamilton-led Iraq Study Group.

The ISG's report won't be released until Wednesday. At first it seemed it would recommend a steady drawdown of the American combat role throughout the next year, to end entirely by 2008. Now, it seems the recommendation will be vaguer, suggesting that President Bush kinda maybe, if conditions are right, based on the judgment of U.S. commanders acting with all due diligence, should reduce the U.S. combat role sometime.

The logic behind this proposal has already been set out by Rumsfeld and Murtha. In a classified Nov. 6 memo, Rumsfeld favorably mentions the idea of modest reductions in U.S. troops "so Iraqis know they have to pull up their socks." In support of his proposal for a withdrawal, Murtha says: "We cannot expect the Iraqi people to take over unless we give them incentive. ... I'm convinced there'll be more stability, less chaos." Obviously, Rumsfeld and Murtha disagree on much, but their preferred approach shares enough that it roughly can be called the Rumsfeld-Murtha Option. Would it work?

If American troops are contributing to instability in Iraq, then the Anbar province -- where American troops have always been too thin on the ground -- should be an oasis of calm. Cities unpatrolled by Americans should be orderly, with well-functioning local governments. No droopy socks should be in sight. Of course, the opposite is the case. It is cities like Tal Afar and Fallujah -- inundated by American troops -- that are relatively stable, while the rest of the province is being taken over by al-Qaida.

The Rumsfeld-Murtha Option is based on the theory that the Iraqis are hanging back, enjoying the luxury of having their country occupied by 140,000 foreign infidels. The simple reason that the Iraqi government isn't stepping up, however, is that it doesn't have enough functioning troops. Rumsfeld and Murtha are correct, in one sense, that Iraqis will step forward to take control of their country when we begin to leave, except those Iraqis will be the likes of Muqtada al-Sadr -- extremists who have built private armies that are more effective than any fighting force in the country, outside of the U.S. military.

That is why U.S. troops are the only hope for stability in Iraq. Newsweek magazine reports on an exchange between two generals about the undermanned, failed Baghdad security plan. A four-star general asked Lt. Gen. Peter Chiarelli, who was running day-to-day ground operations: "Do you have enough forces? Enough to clear an area and stay there to secure it 24/7?" Chiarelli replied, "Of course not." The four-star then predicted, "It's going to fail, it's absolutely going to fail."

It used to be that liberals understood this dynamic better than many conservatives. Once, they touted Gen. Eric Shinseki's recommendation -- blown off by Rumsfeld -- that it would take hundreds of thousands of U.S. troops to pacify Iraq. But now it's liberals who call for a Rumsfeldian policy of fewer boots on the ground. This shift makes sense if liberals think the war is irretrievably lost. There is evidence for that proposition, but none for the Rumsfeld-Murtha argument that Iraq will be a better place, with a stronger central government, if we begin to leave.

Since Bush is not ready to quit in Iraq, he was right to fire Rumsfeld and has been right to reject Murtha's call for a pullout. Bush also must rebuff any finessed version of the Rumsfeld-Murtha Option offered by the ISG. As retired Gen. Barry McCaffrey has noted, if we reduce our combat power in Iraq beyond a certain point, even the minimal presence of American logistics troops needed to support the Iraqi army will be unsustainable -- "we'll end up with 5,000 U.S. troops hostage in that country."

The Rumsfeld-Murtha Option is wishful thinking at its worst.