The Lancet study says: "we estimate there were 98,000 extra deaths (95% CI 8000-194000) during the post-war period". As Fred Kaplan points out in his critique of the survey in 'Slate', the bracketed bit "(95% CI 8000-194000)" is not explained by Lancet but actually means that the study's authors are 95% confident that the war-caused deaths are somewhere between 8,000 and 194,000. Kaplan rightly says: "This isn't an estimate. It's a dart board". So, there is a huge level of uncertainty in the Lancet's finding.

The method used was to interview households - 988 in all - and ask them about deaths before and after the invasion. Corroboration was not sought in most cases - indeed only in 78 households were death certificates requested. Only 63 households actually provided them.

The main problem relates to 'cluster sampling'. Several 'clusters' (sample areas) fell within the Sunni triangle. Figures are exaggerated when the distribution of deaths is skewed.

Another methodological flaw, not covered by Kaplan, relates to  the fact that, in comparing pre- and post-invasion data, the Lancet study did not adjust figures to account for migratory movement.

In light of recent discussion about what it means to be on the Left, I think it worth adding that it includes questioning of official and seemingly authoritative sources.

Here's a link to the Kaplan article, which appeared in 'Slate': http://slate.msn.com/id/2108887/fr/rss/#ContinueArticle  This link also leads to the Lancet study.

The anti-war leaders have used the very flawed Lancet survey in two dishonest ways: firstly, to suggest it is authoritative and, secondly, to suggest that the (flawed figure of) 100,000 casualties are wholly the result of the US invasion.

The death toll in this war has been caused by both sides - one side wishing to restore Ba'athism (Arab fascism) , the other seeking to bury it once and for all through the establishment of a democratic, federal, system of governance based on a constitution that guarantees minority rights.

In such a situation, I blame the 'resistance' of the former for any loss of life.

The 'Iraq Body Count' group offers a more reliable methodology for ascertaining the toll in Iraq. Have a look at the site: http://www.iraqbodycount.net/    

Barry

Hi Barry a couple of things why should we take Iraq Body Count as authorative it has clear methodological problems which underestimate the toll why should we consider migration as affecting the Lancet survey with 150,000 US troops entering Iraq people have been created a positive balance of 50,000 I am more impressed with Alberts comments prior to the war when he said that if 100,000 died in the liberation of Iraq so be it. The only people who have a clear idea of the death toll is the US government and for some reason they wont tell. BTW Im very upset by the treatment of Lindy England compaired to the real criminals Bush and Saddam not to mention that soldier that was filmed murdering that helpless wounded Iraqi Lindy has been treated very harshly indeed PS isnt it good that the Iraqis have forced the occupiers to hold elections

The Lancet study's authors are 95% confident that the war-caused deaths are somewhere between 8,000 and 194,000. This has now become a much-cited definite figure of 100,000 - and, moreover, attributed by opponents of the war entirely to the Coalition forces. I think the point is worth making that the survey is neither authoritative nor reliable.

As for the real figure, I have no desire to under-estimate it. But I remain highly skeptical of the way the Lancet report has been used in public debate.

Is there a point at which the death toll would reach a level where it would render the objectives of the war not worthwhile? In my opinion, it's not just about estimating how many might have died anyway under the old regime but also of those who would have been killed/tortured/oppressed by it in the future. In the American Civil War, more than 600,000 people died. The US population was about 34 million at the time. Was the abolition of slavery worth it? I think so, definitely, yes. Because it's not just the immediate present that is at issue - but the entire future.

On Abu Graihb, I think the US administration missed an opportunity to try for mutiny the "higher-ups" who knew what was happening and who effectively condoned it, violating their own explicit rules of conduct. I've never come across evidence that George W Bush fits into that category. Where is the evidence that he knew what was happening and condoned it? It did not serve Coalition strategic objectives to have such war crimes occur. And they have lost an oportunity to show any resolve to really punish the senior culprits.

The US has consistently wanted elections in Iraq - that's hardly news - but at one point Bremer started talking about something other than direct elections. It's to the credit of Sistani in particular that he would have none of that, and insisted on the type of elections that did in fact take place. It's also good, I think, that Bush's preferred candidate was not successful at the polls. These factors suggest the elections were indeed free and fair. Full marks to the Iraqis indeed - but also full marks to the US and Coalition for overthrowing Ba'athism and making elections a possibility and a reality.

And nought-out-of-ten to the anti-war movement that would have kept the Iraqi people in chains.

Barry