THE POLL NUMBER--80 percent of Iraqis want Americans out of their country--has become a staple of Democratic antiwar rhetoric. Representative John Murtha cited it in his proposed House resolution calling for the immediate withdrawal of American troops from Iraq. Democratic national chairman Howard Dean mentioned it in the radio interview in which he said the war in Iraq is unwinnable. Senator Barbara Boxer put it this way on Fox News Sunday: "Eighty percent of Iraqis want us gone." Senator John Kerry made reference to the poll number in responding to a speech on Iraq by Vice President Cheney last month.

Democrats may be oblivious, but there are reasons to be skeptical of the 80 percent figure. Does it actually reflect support among Iraqis for immediate withdrawal? Probably not. Or have Democrats misconstrued its meaning to justify their demand for the removal of American forces from Iraq, in Murtha's words, "as soon as possible"? It sure looks that way.

There's no doubt about the authenticity of the opinion survey from which this particular finding emerged. That poll was conducted secretly by an Iraqi university for the British military in August. And the poll's results were disclosed in October by Sean Rayment, the defense correspondent for the Sunday Telegraph in London.

The problem with the 80 percent figure--actually, 82 percent were said to be "strongly opposed" to "coalition forces in Iraq"--is threefold. One, it clashes with the demographics of Iraq. Two, it's at odds with other polls in Iraq. And three, the question that elicited the

The population of Iraq is made up of Sunnis (20 percent), Shia (60 percent), and Kurds (20 percent). The Sunnis in the poll would no doubt say they strongly oppose "the presence" of the American-led coalition in Iraq and favor instant withdrawal. It was the Sunnis and their leader, Saddam Hussein, who ruled Iraq for decades before being overthrown by coalition forces in 2003. The Shia, on the other hand, were persecuted by Saddam and thus were happy to see him and his colleagues ousted. The Kurds were even more viciously persecuted and are downright pro-American. So the obvious conclusion, based on Iraq's demographics and simple math, is that 82 percent of Iraqis probably do want Americans to leave eventually, but that they don't favor immediate withdrawal.

If we knew the "internals" of the poll's sample, we could say for sure whether 82 percent of a representative sample of Iraqis said they favored immediate withdrawal. I contacted Rayment, who broke the poll story, and learned the sample size (1,264 Iraqis), but not the breakdown of Sunnis, Shia, and Kurds. That remains unknown, at least publicly. It matters, though. If the sample consisted disproportionately of Sunnis, for example, that would explain a high number of respondents who want U.S. forces to withdraw immediately. However, it wouldn't be a faithful reflection of overall Iraqi opinion.

Earlier polls tell a different (and clearer) story, though still not one that's favorable to keeping American troops in Iraq indefinitely. In March 2004, a BBC poll of 2,500 Iraqis found that 51 percent opposed the continued presence of coalition troops in their country. And in May 2004, a poll in six Iraqi cities, including ones with significant Sunni populations, put the percentage of Iraqis who want coalition forces to "leave immediately" at 41 percent. And 55 percent said they would feel safer if those forces left.