Iraq’s Dead Counted

The content on this webpage contains paid/affiliate links. When you click on any of our affiliate link, we/I may get a small compensation at no cost to you. See our affiliate disclosure for more info

Last updated on April 2nd, 2018 at 04:25 pm

Researchers surveyed 808 households for a study published last year by The Lancet which concluded that as many as 100,000 “excess deaths” had occurred in Iraq since liberation.

The UN has now released a survey of more than 21,600 households:

The invasion of Iraq and its aftermath caused the deaths of 24,000 Iraqis, including many children, according to the most detailed survey yet of postwar life in the country.

The UN report paints a picture of modern Iraq brought close to collapse despite its oil wealth. Successive wars, a decade of sanctions and the current violence have destroyed services, undermined health and education and made the lives of ordinary Iraqis dangerous and miserable.

The survey for the UN Development Programme, entitled Iraq Living Conditions Survey 2004, questioned more than 21,600 households this time last year. Its findings, released by the Ministry of Planning yesterday, could finally resolve the debate over how many Iraqis were killed in the war that overthrew the regime of Saddam Hussein in April 2003.

The 370-page report said that it was 95 per cent confident that the toll during the war and the first year of occupation was 24,000, but could have been between 18,000 and 29,000.

According to CNN, the UN survey was conducted throughout all of Iraq’s 18 provinces (the Lancet study examined 11). Also from CNN:

Iraq’s unemployment rate was 10.5 percent of a population of 27 million people, the report found.

That figure blows out to 18.4 percent when workers not looking for a job are included; the number of unemployed seeking work, however, compares reasonably well with data from France (unemployment: 9.4 percent).

While there has been progress since Saddam Hussein’s fall, “these data depict a very tragic picture of the quality of life,” Iraqi transitional Planning Minister Barham Salih said.

Salih said the mismanagement of Saddam’s government and his regime’s internal conflicts and those with its neighbors took a toll that spared no sector of the country’s infrastructure.

“Saddam Hussein has left us a wasteland,” Salih said. “This country could have been the economic powerhouse of the Middle East.”

And might well become so, in time, now that Saddam is gone and his sons are dead.

(Via Scott Campbell and Alan R.M. Jones)

UPDATE. Tim Worstall wonders why the new report isn’t getting much attention.

UPDATE II. Reader Tim Lambert writes: “The numbers are for different things. The 24,000 is for directly war-related deaths, while the 98,000 includes increases in disease, accidents and murder as well.”

Really? The UN report appears to claim that the two studies are directly comparable:

The ICLS data indicates 24,000 deaths, with a 95 per cent confidence interval from 18,000 to 29,000 deaths. The confidence level was estimated using a linearisation technique (using SPSS Complex Samples, version 12).

Another source (Roberts et al 2004) [the Lancet report] estimates the number to be 98,000, with a confidence interval of 8,000 to 194,000. The website “Iraq Body Count” estimates that between 14,619 and 16,804 deaths have occurred between the beginning of 2003 and 7 December 2004.

That’s from page 55 of the study’s analytical report (pdf file, at the above link).

UPDATE III. John Pilger hasn’t changed his script:

Published in the Lancet, the most highly regarded medical journal in the world, with the tightest peer-review procedures, the study found that “at least” 100,000 civilians had died violently …

He’s never been good with numbers.

UPDATE IV. Reader Michael in comments: “My local newspaper (The West Australian – which has a very consistent ‘bad news’ slant on Iraq) on Saturday quoted the report extensively but left out the 24,000 deaths figure …”

Posted by Tim B. on 05/13/2005 at 06:18 PM
    1. Germany was about 10.5% and climbing the last time I looked.  No, I don’t live and breath German economic figures-just ever so sorry.

      Posted by yojimbo on 2005 05 13 at 07:47 PM • permalink


    1. Plug in the number of otherwise able Algerians, Moroccans et al not looking for a job because the dole in Francistan is so good and the French unemployment rate goes well beyond 9.4%

      Posted by steve68 on 2005 05 13 at 08:03 PM • permalink


    1. I went to the EU sites and gave up on individual countries but the total EU unemployment is 8.9% for 03/05.
      They also have something called the Bank for the account of Companies “Harmonised”-or BACH.
      You simply can’t make this stuff up.

      Posted by yojimbo on 2005 05 13 at 08:08 PM • permalink


    1. 24,000 dead is a long, long way from 100,000, or 150,000, or whatever the latest dread figure being touted by anti-war moonbats.  I wonder, since this is the figure released by the Great Gilded UN, will the left accept it?

      Posted by RebeccaH on 2005 05 13 at 08:50 PM • permalink


    1. only 24,000 dead?  How can anyone be worried about that.  Now for Iran.  No.  Fucket, France.

      Posted by WomBatHed on 2005 05 13 at 08:54 PM • permalink


    1. Way to go, WomBat and Steve, I bet David Heidelberg is sporting wood over your comments now.  “Ha ha!  Steve thinks the only people on the dole in France are Moroccans and Algerians!  WomBat thinks 24,000 dead is no big deal!  I’m SO FUCKING SUPERIOR.”

      Stop making it easy for him.

      Posted by Sortelli on 2005 05 13 at 09:04 PM • permalink


    1. It’s a post unrelated to Iraq but, ironically, Tim ‘100,000’ Lambert has a post up just now called The Case of the Vanishing Wish List.

      Posted by C.L. on 2005 05 13 at 09:21 PM • permalink


    1. It would be nice if we knew how many of the 24,000 dead were Baathists and terrorists who were killed fighting coalition troops.

      Posted by TimShell on 2005 05 13 at 09:23 PM • permalink


    1. Good point TimShell, although expect to see the figure played as ‘24,000 innocent civilians killed by the Evil Amerikkkans’.

      Posted by jic on 2005 05 13 at 09:41 PM • permalink


    1. The invasion of Iraq and its aftermath caused the deaths of 24,000 Iraqis…

      How many of those were from “insurgents” blowing themselves up in a car-bomb?

      Honestly, are we morally responsible for those deaths?  The guys who blow up one car, then blow up another nearby as soon as people come running up to help?

      Remember the Iraqi-enlists who turned on coalition forces when the time came?  Are they in there too?

      Just asking.

      Posted by zeppenwolf on 2005 05 13 at 10:10 PM • permalink


    1. I staggered back when I saw these two grafs in the CNN piece:

      The information about deaths was “derived from a question posed to households concerning missing and dead persons during the two years prior to the survey. Although the date was not asked for, it is reasonable to suppose that the vast majority of deaths due to warfare occurred after the beginning of 2003.”

      One of the first casualties of the war was that Iraqi fellow at the border town, smacking the mural of Saddam with his sandals—he was killed by the baathists the next day. Meanwhile, the last few thousand of deaths have been from terrorist knobs blowing people up.

      Nice of the UN to give us all the credit, but honestly, let’s at least try to fake some kind of science in one of these. Please.

      Children under 18 accounted for 12 percent of the deaths, the report said, while the information on infant mortality and malnutrition shows that “the suffering of children due to war and conflict in Iraq is not limited to those directly wounded or killed by military activities.”

      The war machine gets the blame for Saddam’s decrepit system too huh? I seem to recall that the same UN was blaming the sanctions for thousands of dead children too.


      Posted by Ursus on 2005 05 13 at 10:22 PM • permalink


    1. I’m sorry for every innocent Iraqi killed in the war. I’m sorry if there prove to have been more such killed than were thought. I’m sorry for every Coalition serviceman and servicewoman killed.

      I’m sorry for all those many more innocents whom Saddam killed and for all innocents who have been killed because the world was unwilling to act or still is unwilling to act (Sudan), still living, as most of it does, in the 2nd Millennium.

      I’m glad for the lives that will be spared because we have not waited till it’s too late too act without resultant collateral mass horror. No Qusay will be in charge of a Ba’athist Iraq and its weapons programs as years and decades go by. The Middle East is on a better path now; the peoples of the Coalition to stand resolved and not to squander these hard-won victories.

      Posted by ForNow on 2005 05 13 at 10:28 PM • permalink


    1. The 370-page report said that it was 95 per cent confident that the toll during the war and the first year of occupation was 24,000, but could have been between 18,000 and 29,000.

      The use of “but” is misleading. The report was 95% confident that the toll was between 18,000 and 29,000. (Which is how confident that Crown Prince Abdullah was that zionists were behind a terrorist attack that happened last year).

      I’m somewhat suspicious of the figures of conditions in the 1980s. They seem to portray an improvement in conditions despite a war against Iran that bankrupted the country and killed hundreds of thousands of its people. Other dictators (eg Cuba) have been known to fiddle with their figures.

      Posted by Andjam on 2005 05 13 at 11:28 PM • permalink


    1. As I understood it the Lancet death toll included all who wouldn’t have died if there had not been a war. Thus they made assessments such as this person had a heart attack but if not for the war they would have survived. And insurgents of course. All deaths included on the grounds that the occupying power was to be held responsible
      I don’t know if it said, well Saddam killed this many per year directly or indirectly so we could subtract that number.
      While not clear which deaths this new survey includes it does appear more transparent and common sense based. Anyone know
      better I hope Walterplinge

      Posted by Ros on 2005 05 13 at 11:58 PM • permalink


    1. 24,000, eh?  Why, that’s the size of a small city, like, for instance, Hama.

      Posted by markwark on 2005 05 14 at 12:01 AM • permalink


    1. Ursus: One of the first casualties of the war was that Iraqi fellow at the border town, smacking the mural of Saddam with his sandals—he was killed by the baathists the next day.

      Is there a source for this?  I recall forwarding a request to help this fellow on to Chief Wiggles.  Don’t know what the resolution was, but it was weeks after the statue came down, and he was ostensibly alive then.

      Posted by Achillea on 2005 05 14 at 12:07 AM • permalink


    1. Statistics question:  Doesn’t the standard deviation vary as the inverse of the sqaure root of the sample size?  It seems like the sample increased roughly by a factor of 25 over the Lancet study so the confidence interval should have decreased by a factor of 5.  100,000 (roughly the C.I. from Lancet) divided by 5 is 20,000, much bigger than the +/- 5,000 reported here.

      Did the U.N. have a substantially different method for estimating the death toll? What else could justify such small error bars?

      Posted by jlyoder on 2005 05 14 at 12:26 AM • permalink


    1. I second TimShell.

      Neither study specified that the dead were all civilians. (although the Lancet study of course was a joke – extrapolating hundreds of thousands from something like a dozen actual battle-related deaths – even leaving aside their transparent political motives, I would be entirely unconvinced that their statistical methods were rigorous enough for that, or that they ever could be in a chaotic environment like Iraq).

      I would be very surprised if the forces of good haven’t destroyed at least 15-20 thousand terrorists and fascists. Probably many many more, although a large component were non-Iraqis who might not show up in the figures.

      The terrorists and fascists themselves must have murdered at least 10,000 of their countrymen. But the blame for that rests entirely with them and not a bit with the Coalition. Besides, from all accounts this is still a far lower figure than the number of innocents murdered during an equivalent period of Baathist rule.

      And despite their best efforts, the coalition may have unintentionally caused the deaths of some innocent Iraqis themselves. I would guess probably less than one hundred. If it had been more than that no doubt the fascist-aligned media would have squawked about every single one of them by now.

      To summarise:

      1) Approx 10+ thousand murdered innocents, not the good guys’ fault except that it shows an urgent need to kill more fascists.
      2) A very small number of innocents killed directly by the good guys. Lamentable but unavoidable, and vastly offset by lives saved by the good guys.
      3) 15-50 thousand dead fascists and terrorists. Excellent. A job well done.

      Posted by kipwatson on 2005 05 14 at 01:11 AM • permalink


    1. 24,000 dead (by any standard) is tragic.  But out of a nation of 25 million, that’s 0.096% of the population, over a two year period.

      Perhaps I’m just prejudiced, but that tells me that the Coalition forces have been especially careful to reduce their “collateral damages” (a lousy phrase).

      This does not include removing those deaths caused by terrorist actions (as someone above noted), which arguably could reduce the total attributed to “friendly fire” (a true oxymoron) on the part of the Coalition.

      However, the point of this sort of survey is not fix blame on either side, but to estimate the civilian casualties from the war; the focus is on Iraq, not the Coalition nor the terrorists.  Which is most appropriate.

      But I’ll bet Tim Lambert is having a cow about now.

      Posted by The_Real_JeffS on 2005 05 14 at 01:24 AM • permalink


    1. One correction to my post—terrorists should not be included in this tally, if they can clearly be identified as such.

      However, there’s a problem with that, identifying a person on a case-by-case basis, especially if you have relatives claiming otherwise.

      So the final tally is going to be under contention, if the UN didn’t bother to separate the terrorists killed in action from true innocent civilian bystanders (the article doesn’t say, and I haven’t read the UN report).

      Posted by The_Real_JeffS on 2005 05 14 at 01:29 AM • permalink


    1. I think Ros touched on a very important point. From what I remember the amount of people killed directly by Hussein averaged out to around 30,000 a year, add to that the 500,000 deaths that according to the Hussein supporters were caused by the sanctions (which turned out to be more specifically about the rorting of the oil for food scheme) and you have a figure that makes that 24,000 pale in comparison.
      24,000 deaths is horrifying, but its not even close to what was happening under Saddam Hussein. (But then I suppose he had a sovereign right to do what he wanted with his people – or at least thats what we are told by many lefties)

      Posted by Michael42 on 2005 05 14 at 01:48 AM • permalink


    1. The original Lancet study actually gave a death toll of about 200,000, not 100,000. The extra 100,000 was the toll they figured for Fallujah alone. It was so extreme and unbelievable that it was ommitted from the press releases, not for any scientific reason, but perhaps because even journalists might have had trouble believing it.

      The 100,000 deaths for Fallujah, about one-third of the population, implied a death rate far exceeding that of Dresden of Nagasaki, and rivalling Hiroshima. It was also much greater than Grozny, to note a modern parallel. Grozny suffered about 5-10% of its population dead in a sustained and brutal action that destroyed almost every single building in the city. It was a vastly more ferocious assault than in Fallujah.

      Posted by zscore on 2005 05 14 at 03:08 AM • permalink


    1. jlyoder, the reason where the std deviation for the UN study is so much smaller than the Lancet study is because of the superior statistical design of the UN survey, although it is still possibly a bit questionable as it was based on the last census of Iraq, which was conducted in 1997.  Nevertheless, this sampling methodology would be considerably more reliable than the Lancet study, which, let’s be frank, was never intended to be scientifically rigorous, but a political cudgel which subsequent events were to show was ineffective (Bush was re-elected).  Doesn’t stop moonbats from still trying to use it though.

      If anyone really wants to get into the methodology behind the survey, they can access the whole thing from here.  Its about 370 pages and is really about conditions in Iraq.  The death toll is an almost incidental question, hence the ambiguity of who killed who in the news reports of the survey (although you can pick your conspiracy theory as to why).
      The UN press release, BTW, is headlined:

      Daily living conditions in Iraq dismal, UN survey finds

      Typical, the scientific boffins put a lot of effort in to produce what is a ground breaking report, and the UNocrats politise it.

      Posted by entropy on 2005 05 14 at 05:27 AM • permalink


    1. I assume The Lancet will follow up the U.N. report by claiming the other 76,000 deaths were caused by Iraqis being forced to eat those American plastic turkeys.

      Posted by John57 on 2005 05 14 at 10:16 AM • permalink


    1. The UN report will keep Lambert busy obsessing for weeks. Hope he doesn’t hurt himself.

      Posted by J F Beck on 2005 05 14 at 10:46 AM • permalink


    1. The original Lancet study actually gave a death toll of about 200,000, not 100,000. The extra 100,000 was the toll they figured for Fallujah alone. It was so extreme and unbelievable that it was ommitted from the press releases, not for any scientific reason, but perhaps because even journalists might have had trouble believing it.

      Um, are you sure the Lancet was saying 100K died in Fallujah? I suspect that the extra 100K (or whatever) was what you got when you included Fallujah in sampling what Iraq was like.

      Posted by Andjam on 2005 05 14 at 10:51 AM • permalink


    1. Andjam,

      The official number in Fallujah was 200,000 dead, according to Tim Lambert, who’s spent more time analysing the Lancet study than most. There’s plenty of other links supporting the 200,000 number as well, just Google it.

      This would mean that if for every killed Iraqi there were two injured (two-to-one is a highly conservative figure, it’s more likely to be seven-to-one), the number of Fallujah residents injured or killed would be 600,000 – even though only 300,000-350,000 people actually live there.

      We’re talking a kill percentage that outstrips both Nagasaki and Hiroshima – comfortably.

      Posted by Leigh on 2005 05 14 at 11:09 AM • permalink


    1. Lemme get this straight: 30 years of brutal oppression, chewed on in two major wars, invaded, defeated, wracked by internal conflict from their holdout fascists and their imported buddies… and their unemployment rate is still no worse than Europe’s?  No wonder the Europeans don’t want them to get on their feet…

      Posted by richard mcenroe on 2005 05 14 at 11:21 AM • permalink


    1. It’s both hilarious and terrifying at the same time, Richard.

      Posted by Aging Gamer on 2005 05 14 at 11:51 AM • permalink


    1. My favorite headline for this study so far is one that said “Iraqis soldier on… without jobs, power, water, heat, houses…”

      A liberal’s ability to judge the rest of the world by our standards never ceases to amaze me, but it also goes a long way to show how naive they really are.

      As for Lambert, the guy came over to my blog to crap all over the comments, too. Guy’s out of his mind. If he can’t understand the difference between “100,000 civilians killed by Coalition forces” and “24,000 Iraqis of all types killed in war-related situations”, he’s beyond repair.

      Posted by Aaron – Freewill on 2005 05 14 at 12:01 PM • permalink


    1. … he’s beyond repair.

      Got that right, Aaron!  Asperger’s Syndrome describes the dude exactly.  Of special interest is this symptom:

      Preoccupation with restricted areas of interest (unusually narrow or unusually intense). Some examples are obsession with train schedules, phone books, stamp collecting, or other collections of objects.

      I suppose the Lancet Study might be considered a “collection of objects”.

      Posted by The_Real_JeffS on 2005 05 14 at 12:14 PM • permalink


    1. Yeah, he got a little terse after my rebuttal…

      Next he’ll be calling me at home. “The Lancet data isn’t junk, Aaron.”

      Posted by Aaron – Freewill on 2005 05 14 at 12:28 PM • permalink


    1. That Asperger’s Syndrome definition describes just about everyone I know, including me. Oops.

      Posted by Andrea Harris, Administrator on 2005 05 14 at 12:56 PM • permalink


    1. If Pilger’s still on his game he’ll inflate the Lancet figure to 100 000 000 000 000 dead Iraqis in a few months.

      Posted by James Waterton on 2005 05 14 at 01:03 PM • permalink


    1. Lambert’s been telling more conflicting stories about the nonsensical Lancet study than Kofi Annan has been telling about his son.  Credibility now headed for the Moho discontinuity.

      Posted by Roberts on 2005 05 14 at 01:38 PM • permalink


    1. Aaron, just tell Lambert that “The Lancet data are junk, Tim.”  The nitpicking on data are/datum is usually bugs people like that to no end.

      Posted by The_Real_JeffS on 2005 05 14 at 01:41 PM • permalink


    1. Common sense should tell people that 180 casualties a day for more than two years is just absurd.

      If we are to believe the Lancert report then the 300 people that have died in Iraq in the last couple of weeks are nothing new. I think the new report is probably much closer, but of course there is no perfect accounting in these kinds of situations.

      I think that in time it will get better for Iraqis, but the country is a wreck and only time, money and peace can fix it.

      Posted by terryelee on 2005 05 14 at 04:37 PM • permalink


    1. The wrecked Iraq is a result of Saddam’s manic government not the US invasion, according to the Times

      The UN report paints a picture of modern Iraq brought close to collapse despite its oil wealth.  Successive wars, a decade of sanctions and the current violence have destroyed services, undermined health and education and made the lives of ordinary Iraqis dangerous and miserable

      Posted by rog2 on 2005 05 14 at 06:26 PM • permalink


    1. Real Jeffs – Data is like water.  It’s not a plural, it’s quantitative.  You can have a little data, or a lot of data, or a shitload of data, or thirty-seven petabytes of data.  You can’t have two data.

      Well, that’s what I say, anyway, and I’m not listening to anyone who disagrees.  Pppppttt!

      Posted by Pixy Misa on 2005 05 14 at 07:52 PM • permalink


    1. Lambert is right about the studies measuring different things.

      Still, the Lancet study says little that is useful, just that at 95% confidence (the bar for statistical significance) at least 8,000 more people and no more than 194,000 more people had died than the period before the war.  That’s really not something we didn’t already know, which is probably why they threw out the 100,000 number, which is better than a guess but does not pass the bar of being statistically significant.

      What price in blood is liberty and democracy worth?  That’s the real question behind the numbers game.  Where you stand on the war tends to depend on how much you value those concepts.

      Posted by TallDave on 2005 05 14 at 09:21 PM • permalink


    1. Lambert is right about the studies measuring different things.

      The trouble for Lambert is that he has used the Lancet study to defend anyone (or at least muddy the waters) who used the 100,000 killed figure as a straight U.S kills figure. He could do this because it appeared to be the most comprehensive survey done, the glaring flaws (such as the fallujah figure) in the study didn’t matter, as long as people didn’t delve to deeply into the guts of it, it had the PERCEPTION of credibility, in fact many people didn’t even seem to realise that it wasn’t an actual body count.
      Now the shoe is on the other foot, whether Lambert likes it or not, anytime he or anyone else tries to bring up the 100,000 figure it will be easily dismissible because of the far higher credibility of the U.N study.

      Fair or not, this appears to make Lamberts spirited defence of the Lancet study look very foolish and gullible indeed.

      I noted that my local newspaper (The West Australian – which has a very consistent “bad news” slant on Iraq) on Saturday quoted the report extensively but left out the 24,000 deaths figure – it would seem that they are unhappy about this figure as Lambert is.

      Posted by Michael42 on 2005 05 14 at 11:45 PM • permalink


    1. It’s true that the Lancet figure has been misrepresented, but I debated the study extensively on Lambert’s blog and he’s been honest on the point of “kills” vs. “excess deaths” (though of course many others haven’t).  I do think far too many people who should know better have been guilty of treating the 100,000 number as statistically significant, which it isn’t.

      More study is always good.  Hopefully, this report will be more honestly represented—by both sides.

      Posted by TallDave on 2005 05 14 at 11:56 PM • permalink


    1. You can have two Datas, if you call one Lore.

      Posted by Aging Gamer on 2005 05 14 at 11:57 PM • permalink


    1. ‘Reader Tim Lambert’ haha! take that Lambo! This new U.N data looks pretty bad for your agenda T/L but hey, dont fret! – you can always go back to stalking John Lott.

      Posted by Lucky Nutsacks on 2005 05 15 at 01:22 AM • permalink


    1. I debated the study extensively on Lambert’s blog and he’s been honest on the point of “kills” vs. “excess deaths” (though of course many others haven’t).

      TallDave, I’m not saying that he was personally misrepresenting the figure as proven kills, clearly it was very important to him that the Lancet figure was seen as credible, and given the time and effort he has put into trying to counter the lack of credibility, there was obviously no way he would make the mistake of writing something that was easily provable as wrong.
      What he was doing was defending people who DID use the figure as a proven fact in columns and articles. He can’t do that now, at least not with any credibility.

      My favorite headline for this study so far is one that said “Iraqis soldier on… without jobs, power, water, heat, houses…”

      Aaron, try this one – from the article in the “West Australian” I mentioned before:
      “Invasion piles on misery for Iraqis”

      Posted by Michael42 on 2005 05 15 at 02:09 AM • permalink


    1. Pilger: “Published in the Lancet, the most highly regarded medical journal in the world, with the tightest peer-review procedures, the study found that `at least` 100,000 civilians had died violently …”
      Except the article in question was not submitted to peer review, it was deemed necessary by its authors to get it out before the US election – and Lancet editors went along. Then, its methodology was called OK but not its data. Then its method was questioned as well, but mostly in re Fallujah. Now…

      Posted by John Anderson on 2005 05 15 at 02:30 AM • permalink


    1. That Asperger’s Syndrome definition describes just about everyone I know, including me. Oops.

      I also reckoned a few years ago that I had it.

      I once wrote an article touching on Asperger Syndrome in Xena ‘zine Whoosh.

      (Gosh, how times change – I quoted an article by Alan Ramsey about Tim Fischer. I wouldn’t be likely to quote Ramsey about anything nowadays)

      Posted by Andjam on 2005 05 15 at 05:10 AM • permalink


    1. He’s not Aspergers, just everyday OCD.

      Posted by Aging Gamer on 2005 05 15 at 06:36 AM • permalink


    1. Just as a reminder, the original Johns Hopkins study, as reported in Lancet, concluded that the invasion had caused 98,000 additional deaths in the 17.8 months following the March 2003 invasion.

      That’s usually rounded off to 100,000 extra deaths, as compared to the number of deaths that would have been expected in a similar period just before the invasion.

      The Johns Hopkins (Lancet) team threw out its Fallujah figures because they looked far too high. The team’s conclusions were based on 89 deaths reported by 958 Iraqi households, and extrapolated to represent all of Iraq.

      Had the research team also included the 53 deaths (52 of them violent) reported in Fallujah, they would have extrapolated an additional 200,000 extra deaths, for a total post-invasion death toll of 300,000.

      But OK, stick with the 100,000 for the sake of analysis. That’s what the Johns Hopkins (Lancet) team did.

      The research team found that of the 100,000 additional deaths, 60.07 percent were due to violence. That is 60,070 extra deaths from violence alone.

      How does 60,070 compare to the new U.N. figure of 24,000?

      Some might call the numbers “different.” If you want to be judgmental, you might call one of the numbers “wrong.” (And the Lancet number might be farther off if it turns out the U.N.’s 24,000 included all causes of death – not just violence.)

      The new U.N. study interviewed 21,600 households. The old Lancet study interviewed only 958 (or 988, counting Fallujah).

      I’d give the edge to the new study.

      Posted by Free Frank Warner on 2005 05 15 at 06:46 AM • permalink


    1. One more note: Adjusting for the shorter time period covered in the U.N. survey (It appears to be 13 months), the Lancet figure for deaths by violence would be about 43,000—still higher than 24,000.

      Posted by Free Frank Warner on 2005 05 15 at 07:16 AM • permalink


    1. Pixy, think of the fun we could have tweaking Lambert on his grammar!  🙂

      Posted by The_Real_JeffS on 2005 05 15 at 08:16 AM • permalink


    1. 98,000*43,000/60,070=70,152 Lancet, roughly, time adjust.

      Now if we unscientifically assume that the other lancet numbers are also higher by 79% of actual, the lancet non-violent casualties over that period are 15,154.  15,154+24,000=39,154.

      Expand that out back over the lancet time period the lancet total becomes 54,698.

      Now, another problem with the lancet study is that, while there sampling methodoly may not have been biased, much of their sample participants may have been.  Also, I think, many medical deaths, such as heart attacks, simply happened earlier because of the war.

      Posted by aaron_ on 2005 05 15 at 07:28 PM • permalink


    1. I think the violent deaths are probably exaggerated more than the medical deaths, so the real total is probably between 71,457 and 54,698, and little closer to the 71,457 (using the lancet time period).

      Posted by aaron_ on 2005 05 15 at 07:37 PM • permalink


    1. That’s funny I thought that the majority of the Lancet 98,000 were caused mostly from allied air attacks. At the lowest level for “mostly” (51%) that means that the UN report comes out at less than half the Lancet study on an apples -v- apples basis.

      Posted by jpaulg on 2005 05 15 at 10:07 PM • permalink


    1. Aaron, I’m not sure your calculations work. Could you explain them a little?

      And JPaul G, I’ll break down the 98,000 deaths in the Johns Hopkins (Lancet) study for you.

      Of the 98,000, 60.07 percent were from violence (yes, principally air attacks), 24.67 percent from accidents, 13.93 percent from heart attacks and strokes, 11.48 percent from infectious diseases, and 8.15 percent infant deaths. That adds up to more than 100 percent because in two categories, chronic disorders and “other,” the death rate dropped.

      How many of the 98,000 died from violence? Multiply 60.07 percent times 98,000, and you get 58,869 the Lancet study says died violently in the 17.8 months after the March 2003 invasion.

      To adjust for the 13-month post-invasion period in the U.N. study, divide 13 by 17.8 and multiply times 58,869. That’s 42,994. (That’s the 43,000 I mentioned in the early post.)

      In other words, the Johns Hopkins (Lancet) study found about 42,994 deaths by violence in the 13-month period that later was studied for the U.N.

      In the U.N. study, the comparable number of war dead turned out to be 24,000.

      Posted by Free Frank Warner on 2005 05 16 at 12:37 AM • permalink


    1. Thanks FFW for doing to actual maths.

      Posted by jpaulg on 2005 05 16 at 12:44 AM • permalink


    1. Frank, I used your numbers from previous comment to adjust for time.

      I assumed that the other deaths were exaggerated as much as the extra deaths from violence and the 24,000 was comparable, (43000-24000)/24000=.7917 for the 79%.  Non-violent deaths, I just used (98,000-60,070)*Time Adjustment/1.79=15,154. Total increased deaths, 24,000+15,154=39,154.

      39,144*Time Adjustment=54,698. (Or just 98,000/1.79)

      Shannon Love pointed out that the UN study may not include crime deaths not directly related to conflict in the study, and also the Lancet number omits military deaths.  Post even has Tim Lambert spam, quite funny.

      Posted by aaron_ on 2005 05 16 at 06:36 AM • permalink


    1. Thanks, J Paul. But now I have to point out that Tim Lambert over at the Deltoid blog says we can count only 11 of the 21 violent deaths in the Johns Hopkins (Lancet) study, if we are to compare apples to apples.

      I’ve asked Tim Lambert where the “11” comes from. I’m sure he has more details of the Lancet study, so I’ll have to readjust some numbers when I get the more detailed breakdown.

      I assume the “11” subtracts non-war-related violent deaths (murders, for example) from the “21.” I’m a little surprised the number drops that far (“10”).

      Posted by Free Frank Warner on 2005 05 16 at 06:40 AM • permalink


    1. Oh, the 71,457 is 24,000*Time Adujustment + 98,000-60,070.  It’s just looking at the Lancet non-violents plus the UN violent time-adjusted. It assumes that the Lancet non-violents deaths were accurate.

      Should also point out that there was an increase in violence following the UN survey period.  However, I dont think it was any where near the 201% increase in average violent death rate that would be necessary to make the Lancet and UN numbers line up.

      Posted by aaron_ on 2005 05 16 at 06:46 AM • permalink


  1. The 201% is wrong I used 60,070 violent deaths and 12 months for the 24,000 time period and 6 for the post period.  Here’s what it is with corrections:

    (7,264-1,846)/1,846=293% increase in average violent death rate.

    Posted by aaron_ on 2005 05 16 at 07:46 AM • permalink