Back to Messageboard | Home | Site Map  

Ignorance on Iraq death estimates  
Posted by ALP on March 25 2008, 20:05 » Uploaded 25/03/08 20:29  

I keep seeing evidence of ignorance on basics - from people who should know better.

The Guardian's Jonathan Steele wrote that the WHO study (which estimated 151,000 violent deaths) was a "civilian death toll". It wasn't - it included civilians and combatants, as did the two Lancet studies.

Robert Fisk, John Pilger and others have written (incorrectly) that the higher estimates (eg of Lancet 2006 and ORB) are "civilian" death estimates.

Meanwhile, Dav, over at Medialens, has been busily emailing journalists "correcting" their mistakes, or so he believes. Unfortunately, Dav has got it wrong as well. (He writes, incorrectly, that the WHO estimate of 151,000 was for "civilians"). And, it's not a good idea to rely on the mistaken claim from the unreliable WSWS that "Including violent deaths, the total mortality figure would then rise to roughly 558,000—much closer to the results of previous studies." Better to stick with what the IFHS team say about their own study than to try to make it look more like the Lancet study (with extrapolated wishful - or disingenuous - thinking).

Let's hope that Dav has enough integrity to correct his own errors - perhaps by emailing those journalists and admitting he got the basics wrong too.

"Correcting for the distorted vision of the corporate media - with a different set of distortions".

COMMENTS Post comment


Comment 01 – dav March 25 2008, 23:23

Fair cop...

Dear SC,

I mistakenly wrote that the Iraqi government and WHO study does not include combatant violent deaths, this is incorrect. The study excludes those deaths due to causes other than violence.



Comment 02 – Several Famous Epidemiologists March 26 2008, 08:30

Well, at least some people are prepared to admit to some of their own errors, unlike the two guys who write the Medialens mailings.

The latter quoted a figure of "600,000" which, like the figure of 558,000 mentioned here, they associated with the IFHS.

I think the idea is to exploit people's ignorance. Not very scientific, and not very honest.

Comment 03 – ALP March 26 2008, 12:34

The WHO/IFHS estimate is much closer to IBC's figures than to Lancet 2006's (and is explicit in its support of IBC data and rejection of Lancet data). Consider, then, this (from Medialens):

The new [WHO] study is identical to both Lancet studies in one key respect - it suggests that an appalling humanitarian catastrophe has taken place in Iraq under US-UK occupation. This, in the end, is the point that matters.

In that case, one must assume that IBC's work also suggests an "appalling humanitarian catastrophe" (and, of course, it does). So, is this "in the end" the "point that matters" to Medialens? Well, you wouldn't think so from Medialens's two-year smear campaign against IBC.

Comment 04 – ALP March 26 2008, 12:34

I've just noticed the following posted by someone at Zamparini's blog. I think this is probably typical of the ignorance surrounding this issue. The confusion evident here probably arises from the likes of Medialens and John Tirman trying to make it look as if WHO and Lancet 2006 "agree":

The more recent studies that I know of are Lancet 2006 which estimated about 600,000 excess deaths since the occupation. Its finding may be different from WHO's because my understanding is that the WHO study estimated that 151000 Iraqi's had died violently as a result of the occupation whereas the Hopkins study found that around 600000 more Iraqi's had died since the occupation than would have been expected to die under Sadam Husseins government which I think would also include non violent deaths.

In fact, Lancet 2006 estimated 601,000 violent deaths*, which should be compared with WHO's 151,000 violent deaths - a difference of 450,000 violent deaths.

*655,000 excess deaths from all causes.

Comment 05 – dav March 26 2008, 15:22

Most people have predictably resorted to short hand in referencing the study. It should of course be referred to as:

'Iraqi government in collaboration with WHO' study

Comment 06 – ALP March 26 2008, 15:59

You should let John Tirman and other Lancet defenders know, dav. They seem to refer to it as the "Iraqi government estimate". Remember, it's peer-reviewed epidemiology - therefore we should get down on our knees before its divinely awesome credibility, and never criticise it. Or at least that's what I've been programmed to think from reading Medialens alerts.

And be sure not to refer to the "highly respected British polling firm, ORB", as that survey was conducted by an Iraqi firm, IIACSS, that didn’t exist before 2003, founded by an Iraqi with apparently only limited formal training in survey methodology.

I'm sure you'll clarify that point whenever you mention the ORB estimate in future emails to journalists.

Comment 07 – Rixey March 26 2008, 16:14

Never waste an opportunity to lampoon Ml, do you ALP. For your information the JHU guys have been generous to the WHO survey. Les Roberts praised it in a professional manner and didn't try to dismiss it.

I don't appreciate, or understand, the jibe at ORB. It's their poll, published exclusively on their site, for which they take full responsibility. They're brought all their polling experience and expertise (a lot of it, I see from the ORB bios) to the survey. They conducted a second survey in response to feedback on the first. That indicates thoroughnes and professionalism to me. If critics find serious problems with the WHO study, will they be as committed enough to do further work? I've not seen any responses from them yet (although admittedly I've not seen major criticsms raised yet either).

You should ease of on the smart-arse act ALP. As for the errors you mentioned, so what. Even Chomsky got it wrong on ORB. Big f*cking deal. The bigger picture is what is important, not these little details.

Comment 08 – dav March 26 2008, 16:17

While there is apparently no standardised code for smoke signals, I'll choose to interpret the steam rising off your head as conveying agreement.

Comment 09 – ALP March 26 2008, 17:09

Another gem from dav's book of 'General-Purpose Replies For When You're Stumped'. Still it's better than Zamparini's "if in doubt, say 'genocide denier'."

No, I don't agree that the study should be called the 'Iraqi government in collaboration with WHO' study. If we're all being consistent, I think it should probably be called the Iraq Family Health Survey, just as the "Lancet" study should be called, "Mortality after the 2003 invasion of Iraq: a cross-sectional cluster sample survey".

But since it's called the "Lancet" study, perhaps we should refer to the IFHS as the "NEJM" study?

"Lancet" and "JHU" are pretty inaccurate as descriptive terms for a study. I guess people will call studies by whatever names suit them. Tirman will continue to call IFHS the "Iraq government figures", as that suits his purpose (which is clear enough), and I'll continue to call it the WHO study (or IFHS), as that suits my purpose of conveying that it's peer-reviewed science.

Comment 10 – Rixey March 26 2008, 19:29

Incidentally, not all epidemiology is "gold standard". It may be that the Lancet studies were conducted to a much higher standard than the WHO study. Who is to know?

Comment 11 – Russ Bridger March 26 2008, 19:40

Whenever I turn on my Bullshit Detector, I inevitably find that dav's posts are off the scale, and this thread is no exception.

Those responsible for the NEJM study have chosen to call it: "The Iraq Family Health Survey". See the website:

Or you could call it the COSIT-WHO-MoH IFHS NEJM-published study, but it's a bit of a mouthful, especially for Medialens clones who struggle with basic facts.

Comment 12 – Ged March 26 2008, 19:57

Can anyone explain the 558,000 figure? I've been trying to follow this as best as I can, but it gets more confusing over time. Confusion over the numbers doesn't serve the antiwar cause.

Comment 13 – dav March 26 2008, 20:02

Then you don't agree with the people who conducted the study. You're prerogative I suppose.

"A large national household survey conducted by the Iraqi government and WHO"

Implementing agencies:
Ministry of Health / Iraq
Central Organization for Statistics & Information Technology
Ministry of Health/Kurdistan
Kurdistan Regional Statistics Office
In collaboration with WHO/Iraq[...]

Comment 14 – Russ Bridger March 27 2008, 09:47

dav spoke: "A large national household survey conducted by the Iraqi government and WHO".

That's not the name they give the study, you twit. The name they give it is "Iraq Family Health Survey".

Comment 15 – dav March 27 2008, 11:21

By labelling it the WHO study, one is referring to who it was conducted by. Much the same as referring to IBC figures, or the Johns Hopkins study. Those are the terms in which we are considering it. Not the title of the published document.

Comment 16 – Several Famous Epidemiologists March 27 2007, 15:20

Not really, dav. For one thing, WHO are publishing virtually all the main material on the study (except for the material published in NEJM) - it's hosted on the WHO website, with WHO taking public responsibility for educating people about it, FAQs, etc. On that basis alone, one is justified in referring to it as the "WHO study".

Another thing - "John Hopkins" doesn't cover all the parties involved in conducting the "Lancet" study. "John Hopkins" is a far from ideal label for the study.


Post comment

Display email
  Don't display    Display
Lifespan of comment   Delete after 3 weeks    Keep permanent if post is permanent  


Optional link URL:
eg ""
Optional link text:
eg "Media Hell"


  Messageboard Back to top