According to the BBC, a maximum of 2,000 people were dying per month at the height of the war. That’s 24,000 per year [even though less have died recently, we’ll ignore that for now], and 110,000 since the war began. No small figure, but nearly a million off the estimate in your link and precisely 10 times off the mark as quoted in your article.

It’s one thing to be wrong, but one of us is a million corpses wrong.

Well, instead of offering an explanation myself, perhaps an updated version the latest greatest body count would suffice:

You remember the Lancet study, don’t you? Researchers from the British medical journal conducted a survey of 1,849 randomly selected households across Iraq, inquiring about family fatalities immediately before and after the March 2003 start of the war. They then tallied the responses, gauged the differentials, multiplied out their results by the entire Iraqi population, and concluded that there had been roughly 600,000 “excess deaths” as a result of violence associated with the war. But how, you might ask, did the researchers know their respondents weren’t exaggerating the death toll? Because, they assured us, over 90% of the respondents who claimed that a family member had died were able to produce a death certificate.

Except if you multiply out the death certificate numbers, that means over 90% of those 600,000 violent deaths would have accompanying death certificates — i.e., more than 540,000 death certificates.

Who issued them?

[…]The actual civilian death toll in Iraq may never be known with great accuracy. For what it’s worth, the left-of-center researchers at the Iraq Body Count website, who shun the dubious survey and projection methodology employed by both ORB and Lancet, put the current tally between 74,000 and 82,000 — a range which feels more realistic and seems to jibe with high-end estimates by responsible media sources.

[…]The rest of us can grieve, quietly, for the civilian casualties, whatever their true number, give thanks that they’re trending down, and hope for a resolution in Iraq that ends them.

Fair enough, balbulican. Here is the link, and my site has a new post on this topic as well which deals with the various sources from which the BBC arrives at their figures.

Nope, there’s nothing in the queue, and nothing waiting in moderation. Try sending again (and copy so you don’t lose it if Akismet goes wonky again).

To summarize the key points of your response:

a) You note that there is a wide range of estimates of casualties, and that the stat cited by Stageleft is at the high end of the continuum.

That is correct. What you don’t note, however, is that those states are the product of TWO Lancet studies, not one: and that the Lancet studies were pretty much echoed by the findings of two separate studies by John Hopkins University.

2) You cite another blogger who dismisses the Lancet study because, basically, he doesn’t like their methodology. Ummm…okay. I didn’t find his critique very compelling or rigorous, to be honest - I don’t think he’s a researcher or analyst. There are much more challenging discussions of the Lancet methodology, principally in the pages of Lancet itself.

3) You note that care should be used in citing statistics, and that folks with a political point to make tend to use statistics that favour that point.

Yup.

(Note: Your earlier message just popped out of the ether, above.)

Since I have no basis upon which to judge your feeling, I can’t comment on either that or your estimate.

I don’t know. I’d have to review the various estimates, determine the degree to which they’re measuring the same thing (which the samples on your site don’t), find out what I can about the methodologies, read some opinions critical of those methodologies, and make a judgment.

Although in looking at your sources (SL) I found nothing to nitpick, I’m still mistrustful of all statistics. I use and interpret statistics in what I do. I’m well aware of how stats can be manipulated.