In November, the Bush administration announced that it would do something in Iraq it has resisted for four years: negotiate a long-term bilateral military commitment with the Iraqis. General Douglas Lute, the White House's so-called "war czar," immediately told reporters that force levels and basing rights would be on the table. When I asked a spokesman for the Iraqi government about permanent U.S. bases in Iraq, he didn't rule it out. And indeed, the U.S. Army has been preparing for this prospect for years. If it looks like U.S. troops will be in Iraq forever, that's because the Bush administration, before it leaves office, is preparing to ensure U.S. troops will be in Iraq forever.

In the Times, Thom Shanker and Steven Lee Myers lay out what's at stake in the upcoming negotations, which are set to conclude by July. It's a valuable piece, since it provides the blueprint for how the administration will spin the bilateral accord. First, they're saying -- in Bob Gates's words -- that the U.S. has "no interest in permanent bases." Second, they're saying that the Iraqis are "tough negotiators" and "not supplicants." And finally, they're saying that the U.S-Iraq deal won't "tie the hands of the next president." All three statements are literally true and substantively false.

Sure, the deal won't say, as with Guantanamo Bay, that the U.S. will have a 99-year lease on the Baghdad International Airport complex. That's because of the fear that any such specificity would require congressional approval that the Democrats will not grant. So it's not hard to see what will happen: the commitment will simply be open-ended, the text will call for review in X number of years. Hence the next president's hands are untied and the Iraqis will push for constraints, etc. But look at history. It took the Philippines nearly 100 years to get the U.S. out of Subic Bay and the Clark Air Base. That's because the fact of the U.S. presence creates additional, subordinate facts -- economic dependency in the area around the base, for one, and more fundamentally, a political dependency on the U.S. for a security guarantee, which is the whole point of the bilateral deal. In Iraq, a weak central government requires the U.S. to keep it alive against its multitudinous armed adversaries, a weakness that Iraq's sectarian quasi-democracy actually fuels. (Elections in Iraq tend to become sectarian census counts in a power struggle.) So while the Iraqis may push back, no Iraqi government that could actually take power -- one led by the Sadrists, for instance, or the harder-line Sunnis -- would actually kick the U.S. out. That in turn drives a divide between the fearful Iraqi government and the anti-occupation Iraqi populace, further entrenching the government's dependency. Nouri al-Maliki and his successors have to think: Without the U.S., will I be strung up on a lamppost?

The whole idea of the deal -- and its timing -- is to tie the hands of the next president. It's true that the president won't formally be constrained, particularly if the deal won't be subject to Senate approval. But diplomacy is funny thing. It contains its own social force -- what Jonathan Rauch insightfully terms "hidden law." Diplomats worldwide work hard to ensure the promises their governments make survive political transitions. All the incentives of geopolitics are toward stasis: Bush's assurances need to be kept by his successor, otherwise it's difficult for that successor to get other countries to trust his/her commitments to them. A mercurial United States is not something the rest of the world likes, regardless of the merits of the changed policy. At the very, very least, Bush's successor faces an uphill battle to undo the bilateral deal -- and that's before the Iraqis start griping about the U.S. not keeping its word and the domestic press runs with that storyline. And, fundamentally, that's exactly why the Bush administration is negotiating this deal before leaving office. But once again, the administration will tell the American people that an indefinite occupation in Iraq is a bug -- things aren't going so badly that we need to leave! or they're not going so well that we can afford to leave! -- when it's, in fact, a feature.