New study estimates 655,000 Iraqis dead

When a study was released by researchers at Johns Hopkins in 2003 that put the Iraqi death toll at more than 100,000 people, response was predictable. Progressives were appalled, and right-wingers were chanting "Lies, damn lies and statistics." They're back with a new report that was published in The Lancet. Which means that it was peer-reviewed and found to use reliable methodology. But it isn't a pile of skulls, and that makes it as good as a left-wing conspiracy for the right. And as several blogs writing about the report noted, it was just last year that Fearless Leader himself put the death toll at "more or less" 30,000, with all of his characteristic dignity.

The researchers didn't count 655,000 bodies or death certificates to get to their numbers. They used a method called "cluster sampling." They interviewed people in a sampling of 1,849 households across Iraq, and they extrapolated their results from information they gathered on the 629 Iraqi deaths about which they learned in those interviews. Critics will inevitably claim that you can't base such huge numbers on such small ones -- and we're sure to hear almost immediately (what time does Rush Limbaugh's show start?) that John Hopkins, MIT and the Lancet somehow conspired with Democrats to release this information less than a month before the midterm congressional elections. (. . .)

Moreover, the researchers say, their results conform to the estimate of 100,000 deaths they made after a similar study in 2003 and 2004 and with news reports and government tallies showing that the Iraqi death toll has increased dramatically over the course of the past year. Boston University researcher Paul Bolton, who has reviewed the study, tells the Wall Street Journal that the researchers' methodology was "excellent" and represents the standard procedure for such work. "You can't be sure of the exact number," Bolton says, "but you can be quite sure that you are in the right ballpark."