Lancet study, one year on

Les Roberts
comments on the shoddy reporting of his study:

I thought the press saw their job as reporting information. Most of the pieces discussing our report were written to control or influence society, not to relay what our report had documented. For example, the day after the article came out, Fred Kaplan, a defense correspondent for Slate magazine reported that, "Yet a close look at the actual study, published online today by the British medical journal the Lancet, reveals that this number is so loose as to be meaningless. But read the passage that cites the calculation more fully: We estimate there were 98,000 extra deaths (95% CI 8000-194 000) during the post-war period." He concluded that, "This isn't an estimate. It's a dart board."

First of all, we reported three things: violence was up 58-fold, becoming the main cause of death; the one neighborhood visited in Anbar Province had 1/4th of the population dead, statistically suggesting almost 200,000 deaths; and in the other 32 neighborhoods we estimated 98,000 deaths. When the three things were taken together, the likelihood was far greater that there were over 100,000 deaths, but Kaplan chose to only focus on the third finding. Secondly, the line which Kaplan quotes above did not appear in our paper. The line which does appear reads, "We estimate there were 98,000 extra deaths (95% CI 8000-194 000) during the post-war period in the 97% of Iraq represented in all the clusters except Falluja." Note that not only did Kaplan cut off the sentence but he inserted a period into his quotation. Finally, once he ignores the violence issue and the Falluja data, he implies that the result from the other 32 neighborhoods could be anywhere between 8,000 and 194,000. In fact, this normal distribution indicates that we are 97.5% confident that more than 8,000 died, 90% confident more than 44,000 died and that the most likely death toll would be around 98,000.

I am not so surprised that Fred Kaplan, a former military employee, now defense writer, wants to temper potentially critical findings about his colleagues. When Human Rights Watch weapons analyst Mark Garlasco was asked by the Washington Post what he thought about the report estimating 100,000 civilian deaths in Iraq, he said that he had not yet seen the report but 100,000 deaths seemed too high to him. His statement was abbreviated in the Post to state 100,000 deaths seemed too high. After reading the report and speaking with colleagues, his skepticism waned and he never repeated his doubt.

He does report some good news---the study may have partly achieved its primary goal of getting the coalition to change their tactics:

When the session was over, one officer asked me, what did I expect, we have dropped perhaps 50,000 bombs on insurgents hiding behind their own people. In the conversations that followed, I sensed great frustration on their part at the style of violent air-power based response to the insurgency designed to minimize US troop losses. While we will not know for years and maybe ever, colleagues in Iraq and pundits in DC tell me that the US has greatly curbed the use of airpower in Iraq, perhaps independent of or perhaps in part due to The Lancet report. As public health professionals, we collect data on problems to stop those problems from occurring. Thus if true, this change in occupation dynamics on the ground is most welcome and came about independently of the US public's attention or interest.

Also I should have mentioned this update on the study at Media Lens. As well as some more comments from Roberts, it has a retraction from John Allen Paulos, one of the very very few numerate critics of the study. Paulos originally wrote:

Given the conditions in Iraq, the sample clusters were not only small, but sometimes not random either... So what's the real number? My personal assessment, and it's only that, is that the number is somewhat more than the IBC's confirmed total, but considerably less than the Lancet figure of 100,000.

Paulos now says:

I regret making the comment in my Guardian piece that you cite: ... I still have a few questions about the study (moot now), but mentioning a largely baseless 'personal assessment' was cavalier.

and

A suggestion: use Katrina as a news hook and have LR write an OpEd for the NY Times (or Newsweek or some publication with a huge circulation) explaining sampling, clusters, and the problems associated with counting the dead. Next explain that this was done in the Congo and finally revisit the Lancet findings. The sympathy that Katrina arouses might enable him to get past the political resistance to the Lancet findings (resistance that probably won't disappear until this abominable war ends). He certainly has the standing to write such a piece, and the issue is still very important. I understand now the situation surrounding the study's original publication. I also understand LR's anger, but he should lose the vitriol to get such an OpEd published. Good luck. Best, JAP

Tags
Categories

More like this

The sad truth of the Lancet study is if the US was a bit more of an open Democracy, the civilian casulties would be an important part of the calculus on which the decision to wage, or continue to wage, war is debated.

As Lambert has now decively proven (probably for the 50th time), the Lancet study should have been discussed on what the numbers meant to us, not if the numbers were accurate.

In other news, a spokesman for the US military has said that the IBC numbers are credible.
The story goes on to make the same point that an exasperated Lambert, Davies, etc. have made (how many times now?) -- that the IBC numbers are almost certainly a lower bound:

One U.S. military spokesman said it is possible the figure for the entire war could be 30,000 Iraqis, which many experts see as a credible estimate. Others suspect the number is far higher, since the chaos in Iraq leaves the potential for many killings to go unreported.

As this sandstorm subsides, the Lancet study is looking better and better. After it is finally accepted that Roberts' methods were sound and his results realistic, we may look to Michael Fumento to apologize for having failed to read the report before commenting on it -- that is, when jetpack-equipped swine are making scheduled lunar shuttle flights.

For me the most valuable part of Roberts piece was his statement of the value of statistics and for more quantification of what is happening in Iraq, rather than the anecdotes much debate revolves around. I've criticised the study, but at the time it was best and only look at mortality in Iraq and though flawed it was worth doing and publishing just because of that.

For me the wonderful think about the net it that it short circuits the mass media. I don't mean that blogs are vitally important, at some people think, but that we're not reliant on the media reporting information any more. If I want to read the Iraq study, it's much easier with the internet to go to the source. It's fantastic that our dependence on newspapers for information has been gotten rid of.

LOL.

OK, so first Les Roberts complains about the media, then he pulls out virtually the ONLY major media source/figure that criticized his study. Gee, how misleading. Roberts was pissed that his study didn't get more media attention, even though almost every single major media news source reported on the study the day it came out or the day afterwards. What more did they want him to do? Pound it into the US public every day for a week? Give me a break.

Secondly, Kaplan didn't "quote" him, as I don't see any quotation marks. What Kaplan said didn't change the truth of what he was saying, either. You know, the dartboard statement. Whether he left out the part about Fallujah does not change that the confidence interval was gargantuan, which is what he was saying.

Then, Roberts tries to pat himself on the back for "making a difference". Oh geez, give me a break. They used an air strike just the other day, and they will do so whenever it becomes necessary. Roberts can dream about "making a difference" in this world, but that is just being a delirious egomaniac.

Also, Paulos never retracted saying that the clusters weren't selected randomly. He retracted another part of the statement he made, but not that. Again, Roberts is being very selective here, very misleading. That's understandable though, since he is on the defensive.

Yes, please do revisit the Congo study in tandem with the Iraq one. Please, please do. Then it will be obvious to anyone with a clue how different the methodogies were. Here's an even better suggestion: have Les Roberts use his Iraq study methodology in the areas affected by Katrina to calculate the "excess" deaths caused by it. Let's see what he comes up with, because we already have records to show us what the real number is.

Burying the Lancet? Yeah. Sure. That's why it was in every major media source on one of the busiest news days in the United States: the Friday before a presidential election.

Are you guys just delirious or insane? I'm starting to wonder. (Roberts included)

I guess it goes with the liberal idea that the freedom of speech means that you have the right to have your opinions heard across the world through the media megaphone, and anyone not receiving this right, or not receiving it enough, means that you are being "buried", "oppressed", "supressed", or "censored".

The UNDP study was barely mentioned by the media. The UNDP. Let me emphasize, the UN. A study by the UN. A dinky little survey done by 6 Iraqis in Iraq, led by Les Roberts, got more coverage than an exhaustive, enormous and detailed study about every facet of Iraqi life.

And Roberts, and his enablers, are complaining about being "buried".

Give me a friggin break guys.

Seixon, the UN study wasn't exactly favorable to the occupation. The numbers weren't that out of line with the Lancet study and Roberts says the UN study apparently didn't elicit all the deaths that occurred, so the difference would narrow.

The point is that civilian casualties should be front and center in the debate, with the media talking about all the studies and surveys, speaking to many different experts, and pressuring our government to tell what it knows. (And if it doesn't, making it clear that if there are whistleblowers who want to speak anonymously will be given a hearing.) Instead, both the UN study and the Lancet study go with little mention, and the Iraqi group which claimed over 100,000 deaths is something that only got passing mention--I found a piece at the Washington Times about it online.
Iraq Body Count gets treated with respect--the one methodology which is almost guaranteed to give an undercount is treated as though it were the gold standard in accuracy, precisely when estimates which go even higher start getting a tiny bit of attention.

As far as that air strike is concerned, from what I read the US military claimed all the deaths were insurgents, while Iraqis said many or most were civilians. Kinda makes you wonder how many other reported insurgents were actually civilians. Roberts didn't say the US had stopped using air strikes--he claimed they may have cut back. One would think everyone would be glad to hear it--even if one sets aside the morality, air strikes against an insurgency are a great way to turn even more of the population against the US military. There was that recent poll reported in the British paper, in which over 40 percent of Iraqis thought attacks on coalition forces were justified--air strikes are one way of raising that number.

By Donald Johnson (not verified) on 27 Oct 2005 #permalink

Donald,

I'm not saying that the UNDP study was favorable to the occupation, although it was more favorable on the issue of civilian casualties than other studies.

My major point was that the UNDP study gave a broad story about how Iraqis were doing, and the study went largely unmentioned in the media. The claim in the study about 24,000 civilian deaths went almost completely unmentioned by the press.

Why should a study carried out by 6 Iraqis led by one guy, sampling about 900 households and coming to a singular conclusion have more coverage than a study sampled by the UN, sampling almost 22,000 households, coming to many conclusions about Iraqis?

There is no sane response to this question other than that it shouldn't have had more coverage.

You said that the Lancet study went with little mention. What on Earth are you talking about? Every major media source covered it. How is that "little mention"? Can you explain this to me? Roberts said it, and Lambert and MediaLens obviously agree, but on it's a farce on its face. All of them just point to Fred Kaplan as if he was the only one who covered it. What are all of you smoking?

All the major news sources covered it. Little mention?

IBC, I haven't seen a single news article based solely on their count. I have seen virtually every major media sources write an article solely based on the Lancet study. Can you please tell me why in the hell you guys keep claiming that the Lancet study was "buried"?

It's almost like we are living in completely different worlds.

US airstrikes have gone down since the initial invasion. Wow. Has there been huge problems with Fallujah, or another place, recently? No? So they haven't needed airstrikes that much recently? Oh... The fact that Roberts even tries to attribute that to himself is just pathetic and indications of illusions of grandeur.

The USA said that they killed only insurgents in an airstrike, and unknown Iraqi sources say they were all civilians. I'd tend to think that the truth lies somewhere in the middle. Right? It's easy to pass off insurgents as civilians anyways. I mean, at what point does a civilian become an insurgent? When he is actively shooting at the coalition, or bombing them? Or is he still an insurgent when he is holding arms, planning attacks, etc? I think the distinction of "civilian" is being misused a lot here just because of that. "Oh my dad, he was an innocent civilian! Why did they kill him!?" See, a son may not want to acknowledge that his father was an insurgent, proclaiming him as completely innocent.

So as I say, I think the truth is somewhere in the middle.

As for the poll conducted on behalf of the British Ministry of Defence, check out my blog post on this.

Basic summary: the poll was conducted by an Iraqi university research team (unbiased? ehm...) and their findings on other factors than Iraqi sentiments towards the coalition, such as electricity, sewage, etc, are wildly off the mark compared with the UNDP study...

And as usual, I think it is a bit abominable to go around citing polls for which you don't know anything about the methodology, and obviously, their correlation with other studies.

Seixon's rants fail time and again to address the real issue here: aggressor nations have no interest in tallying up the total of civilians they kill. The US/UK authorities don't give a hoot for civilians killed by coalition bombs with the sole exception of how it might affect the second great superpower: public opinion. The elites are terrified of it, hence why there was such a mendacious attempt at softening up opposition to this crime beforehand in the US and UK via mass propoganda efforts through the state-corporate media apparatus. Moreover, had more than 80% of humanity not expressed complete opposition to US aggression, its quite likely that the US war party and its pathetic proxies (UK, Italy, Spain) would have employed carpet bombing techniques used in Viet Nam and Cambodia which killed hundreds of thousands of civilians.

We now know that Nixon more-or-less ordrered Cambodia to be reduced to rubble, as released transcripts of old war criminal scum Henry Kissinger to the US senate attest: "anything that flies on anything that moves". This is probably the most explicit example of a command for mass murder in history, by a commander in chief (the US president) and one of his senior staff, yet the coverage of it in US media circles and via the punditocracy was undetectable.

Seixon, you are defending the indefensible. As I have said countless times on Tim's blog, the super powerful and super violent answer to no one. The crimes of the current US regime and its prdecessors don't even have time to go down the Orwellian 'memory hole' in the US as they are effaced from public scrutiny as they happen.

By Jeff Harvey (not verified) on 28 Oct 2005 #permalink

Alex Blumberg will be doing a detailed pro-Lancet piece on the popular American public radio show This American Life this weekend. It should discuss the findings and why they were so resisted and ignored, apparently including interviews with Les Roberts and other major players. It should be available in RealAudio on the website Monday or Tuesday.

Seixon---

Roberts "egomania"--Irrelevant. It'd be interesting to know if the US dropped 50,000 bombs in areas where civilians were likely to be killed, and if they've greatly cut back on the use of aerial bombing and why.

The recent Iraqi poll--Some of the numbers as you report them were much different from the UNDP number--unemployment and water quality. Could that change in a year? Maybe. Others (sewage and electrical use) seemed consistent, especially given the vague wording. As for hatred of American troops, it could be that many Iraqis have had bad experiences with American troops-that combined with the normal human hatred of foreign occupation might explain the number, if accurate.

The UNDP vs. Lancet numbers--Not terribly different to begin with and if the UNDP survey was as cursory on violent deaths as Roberts claims, and if more interviewing brought out more deaths, maybe the UNDP survey isn't the gold standard.

Press attention to the Lancet report--It got one story around page A10 in the NYT when it came out and their religion editor mentioned it. The basic problem, I'll repeat, is that the issue of the civilian death toll is greatly underreported--if the press did its job the UN report, the Lancet report and every other estimate would be front page news and there would be periodic stories updating the controversy and the fact that we know so little. The NYT did another story on Wednesday on civilian casualties--page A12. Fine as far as it goes, but it doesn't go nearly far enough. They cite Iraq Body Count and Anthony Cordesman who uses US figures (it doesn't explain what those figures are). Cordesman calls IBC the best guestimate in town, but at least adds that its figures are far from complete. When it discusses who causes the casualties, the NYT in a typical display of cowardice only mentions insurgents. The press, incidentally, is perfectly comfortable giving wide ranges of estimates for killings when some enemy is responsible, and they don't worry overmuch about the accuracy so far as I can tell.

On innocent civilian vs. insurgent--The line is always blurry in guerilla wars. That's why guerilla wars are very hard to morally justify.

By Donald Johnson (not verified) on 28 Oct 2005 #permalink

Jeff,

Way to bring the discussion way off into the woods, once again. I already explained to you why each side doesn't count each others dead. I guess you will never get it. Everyone counts their own, for practical and logical reasons. I know you'd like US soldiers to go around counting dead people while getting shot at, but I am not so sure that they would like to do that. You know, that would get even more people killed, and I thought we were trying to avoid that. Eh?

I'm defending the indefensible? So you'd rather want Saddam Hussein in power, is that it? Or what? Vietnam was very horrible, you are absolutely correct. You should look up from your history book and see that the Iraq was is nothing like the Vietnam war. If you don't, you'll miss out, as you obviously have been for the last 2 years.

Donald,

Yes, it would be interesting to know how many bombs the USA dropped, when, and why. The blatant fact is that we don't know the answers to any of these questions, so I find it quite amusing that Roberts sits there and pats himself on the back in his own fantasy land.

I'm sorry, but I don't think unemployment has gone from around 10-20% to 40% in southern Iraq in just one year. The Iraqi GDP grew an estimated 52.3% in 2004. You really think that translated into a 20-30% spike in unemployment? Riiight.

Why are we even talking about the MoD poll? We have no clue about the methodology, how many were polled, not a damn thing except for its conclusions, that a journalist cherry-picked none the less. It is obvious that its results are way off on all the factors cited.

I have seen the Lancet cited in news articles again and again. I saw it splashed on the front pages of news web sites. I saw it everywhere for that period right after it came out. If you guys call that "burying" it, then I just cannot help you out of your delirious state. It might be your opinion that every paper in the USA should have run it on page A1, but you might want to consider the fact that the Friday before the presidential election is one of the heaviest news days in existence for the US.

I'll say it again, just because you feel that you should have been heard on the frontpage, doesn't mean that it should be so. The UNDP study was barely even mentioned at all, a study done by the world's largest bureaucracy with member nations from most of the world. So you can whine all you want, but it got coverage, and at least on the internet, it was allll over the place.

*[Paragraph deleted. Seixon, I will not permit you to make false claims about statistics on this thread. You already have a thread where you can do that. Tim]*

This again comes back to Roberts' egomania. Aw, his study didn't get on page A1 on the New York Times on one of the busiest news days there are. Boohoo, cry me a river. Next time do a robust study and maybe someone will put you on page A1.

> Quote [Paragraph deleted. Seixon, I will not permit you to make false claims about statistics on this thread. You already have a thread where you can do that. Tim]

Tim your remedy is worse than the disease. By all means censor personal invective and abuse but leave "false claims about statistics" to be cured in the usual manner: falsification.

If the error bothers you enough to delete but not enough to correct, then point to where it's been made and corrected before. Expecting visitors to hunt through all possible candidate posts on your blog isn't the way to go about that either.

Statistical science is strong enough to withstand impugnment on a blog -- in any case censorship is no way to defend her.

hmmmm Seixon, I would very much like to know the methodology behind the following claim on your weblog:

Are we to believe that Iraqi infrastructure has gone so much backwards over the last year?

I think that if you're criticising other people's methodology so much, you ought to refrain from making so many completely unsupported statements. Furthermore ...

The Iraqi GDP grew an estimated 52.3% in 2004. You really think that translated into a 20-30% spike in unemployment?

It is 2005, not 2004.

Roberts says the military mentioned the 50,000 bomb figure. It'd be interesting to know exactly how many were dropped in urban areas, but unless the US chooses to release that figure and many others, we won't know.

You keep missing the point about our "whining". As citizens of the country involved in a guerilla war, it's our duty to whine about our government not keeping us fully informed on who we might be killing, and it's also our duty to whine about the press if it doesn't try to fill the information gap. I can't think of too many stories more deserving of front page treatment--certainly the press had no problems spreading US government BS about WMD's on page 1.

By Donald Johnson (not verified) on 28 Oct 2005 #permalink

>The Iraqi GDP grew an estimated 52.3% in 2004.

Supposedly - if true it partially offset the fall in 2003.

I find it amusing that given your ferocious attakc on the Lancet Journal you're happy to accept as accurate to the first decimal place an estimate of 2004 economic growth that published two months before the year in question even ended.

Additionally, a major part of that growth in economic activity was attibutable to security spending by the US military and private contractors (which regularly exceed 50% of the total cost of a project.)

By Ian Gould (not verified) on 28 Oct 2005 #permalink

the most likely death toll would be around 98,000

Is there any excuse for someone claiming to be statistically literate to make a deliberatly misleading statement like that?

There is always a 'most likely' possibility, there is a 'most likely' set of numbers to come up in next weeks lottery. What justifies picking that number out when it actually rather unlikely to be right?

Also, why is he talking of 'death toll', when the survey, as it regularly pointed out here, doesn't measure that?

Above all, why does he say 'post-war'? Doesn't the survey cover the period of the artillery firing/tanks-shooting unambiguous actual war? Is he really trying to attribute deaths caused _during_ the war to the current situation?

If so, isn't that a bit arbitary? If you took the before/after periods to be 1 year before and after Bush landing on the carrier, how different would the figures be?

soru

> Is there any excuse for someone claiming to be statistically literate to make a deliberatly misleading statement like that?

Whom do you think you're fooling?

There is absolutely nothing misleading about the statement that 98,000 is the most likely result. We're discussing statistics. Robert's statement is mathematically correct given the data in the Lancet report. Your comment does not make sense to me. Stating that 98,000 is the most likely result is just as "misleading" as stating that 50 heads and 50 tails is the most likely result if I flip an unbiased coin 100 times.

If you disagree, what is your analysis of the data? Please try to be numerate in your response.

By John Palkovic (not verified) on 29 Oct 2005 #permalink

It certainly is astonishing to see Roberts claim that he and his team have changed the characteristics of the use of air power in the conflict.

This man clearly has no understanding at all of the dynamics of the war and its changing circumstances. The war has changed significantly from what was almost a conventional conflict, to its present form as a middle eastern equivalent of low level operations like those seen in Northern Ireland. A transition from conventional war to low level ops will quite naturally see a reduction in the use of aerial bombing.

Certainly, amongst my circle of military colleagues, none of them have even *heard* of the study by Roberts et al. How does he think he is changing military tactics- through mental telepathy?

Wilbur, since Roberts says that the reported curbing of the use of airpower in Iraq was "perhaps independent of...The Lancet report", it is wrong to attribute to him the view that he is changing military tactics. It is highly misleading to suggest that what is going on in Iraq is the equivalent of low level operations like those seen in Northern Ireland. The British army never even fired tank rounds in Northern Ireland, never mind dropping bombs.

By Kevin Donoghue (not verified) on 30 Oct 2005 #permalink

I don't know where he got his information, Wilbur, but the reporter Seymour Hersh said that the use of airpower in the second part of 2004 was rising dramatically, probably because of Fallujah, which was heavily bombed in the summer and fall before the final assault. The US does have a tendency to fight guerilla wars with airpower--there was Vietnam, of course, and even in El Salvador there was a phase where the US client state cut down on outright massacres by its forces and switched over to using airpower against villages thought to be pro-guerilla.

Roberts might be paying the Pentagon an undeserved compliment in suggesting that the possibility of high civilian casualties has caused it to change its tactics, assuming that it has done so. It doesn't take a genius to know that the destruction of Fallujah probably didn't endear the US to the Sunnis.

By Donald Johnson (not verified) on 30 Oct 2005 #permalink

>a middle eastern equivalent of low level operations like those seen in Northern Ireland.

So how many airstrikes did the British use on IRA positions in an average year?

From memory, the total death toll from "The Troubles" was around 10,000 over a thirty year period. Even allowing for the Iraq population being 8-10 times larger than that of Northern Ireland (again, from memory, I believe the relevant figures are circa 3 million for NI and 25 million + for Iraq), the level of civilian deaths in Iraq is several times higher than in Northern Ireland.

As to whether Roberts has "no understanding at all of the dynamics of the war", he has visited Iraq, spoken with US officers serving there and has a history of carrying out studies of war-related fatalities.

What are your comparable qualifications?

By Ian Gould (not verified) on 30 Oct 2005 #permalink

The report's "primary aim" was to get "the coalition to change their (sic) tactics"? I thought it was supposed to be an impartial and scientific survey.

Roberts' air-power not-quite-a-thesis is laughable and, from someone so bright, must be regarded as deliberately tendentious.

No surprise there.

His conclusion about the interest and attention of the US public is just sour grapes, of course. He had hoped to make a difference to the outcome of the US election.

As far as I'm concerned the exact number of dead is not so important.

- Dr Les Roberts.

Ian noted; *"What are your comparable qualifications?"*

Well, I'd like to think that a few years of soldiering on my part, and four friends who have served in Iraq, would give me some insight (I'll hasten to add that I have *not* been there personally, but then I doubt you have either).

Somehow, reading your comment, I don't think you are familiar with the difference between conventional ops and low level ones. Claiming that the RAF and Army Air Corps have never used force in Northern Ireland is wrong for starters. Lack of GPS guided weapons has prevented the use of high tech airstrikes in Northern Ireland, as most of the troubles were over before they were invented. However, Army Air Corps aircraft have been used to provide suppressing fire using small arms in the past, and have frequently been exposed to small arms fire and an infra-red missile threat from the ground. Indeed a work colleague of mine has actually flown on some of these missions back in the late eighties.

Other than scale, there is little difference between a low level operation in one conflict and the other. Many of the tactics used by troops and insurgent sides are similar. And, if Northern Ireland had a population of 25 million, with the violence scaled up to match, air strikes from high speed jets might have been used, always bearing in mind that different military cultures could have seen any number of different directions in tactics.

Saying that Northern Ireland wasn't on exactly the same *level* of violence doesn't really cut it either. Iwo Jima and San Carlos bay were not on the same level of violence either, and yet they were both conventional ops. Northern Ireland and, say, either Algeria or Kenya in the '50s were not on comparable levels of violence, and yet they are all low level ops.

Reading Roberts in the article is a good insight into someone who has a clear and passionate political bias. I have no problem with that- except it was not explained by him when he released the study (If he wishes to accuse Kaplan of a bias, for instance, this is relevant). His statement that *"All sources that distinguish the assailants agree that the Coalition forces have killed many times more than the anti-coalition forces"* conveniently leaves out the sources that say no such thing. And I really can't suppress a giggle at his suggestion, without any evidence, that training Iraqi troops may be fueling the insurgency because some might be loyal to it. Well, he is getting inventive there. Certainly from what I've heard there have been plenty of defections in Iraq- they all seem to be going the *other* way.

And Kevin, I noticed you stated that the British Army had never used tanks in Northern Ireland in combat.

I believe they have actually used them in other low level ops (the former Yugoslavia being one). And there has been casualties from the use of armoured vehicles in Northern Ireland. Indeed an eight year old boy was accidently killed in the early seventies when 0.50' calibre rounds fired from an armoured car penetrated a wall during a contact, killing him in his bedroom. Granted, this was an armoured car, not a tank, but then Northern Ireland is a lot smaller than Iraq. If Belfast was ten times as big, and had large open streets, who knows what the british army might have used.

Heavy use of armoured vehicles has always been a constant in Northern Ireland. But then use, or the lack thereof, of armour is not a reliable indicator that something is a conventional op or a low level one.

CL claims

>He had hoped to make a difference to the outcome of the US election.

This is not true. Roberts has specifically denied this.

There was no 'Irish Republic' with a military apparatus, tens of thousands of troops, special forces, an air force, armoured divisions etc that the British had to overcome before engaging a remnant insurgency. The IRA was an insurgency within a state ab initio. For comparative purposes, one wouldn't merely need to assume Northern Ireland "had a population of 25 million, with the violence scaled up to match." More importantly, one would need to assume Northern Ireland was the bastion state of the IRA - as Iraq was controlled by the Ba'athists. Insofar as previous analogies have overlooked this obvious point, they're of very little relevance.

It follows that airpower was vital in the state-on-state war in Iraq, less so when the Iraqi regular army and militias, as well as its military superstructure, were either destroyed or scattered by the disarray. That is the explanation for the reduction in airpower usage.

Very little explains the 'Mission Accomplished' slogan that came between the two phases - one of the most foolish PR stunts in recent presidential history.

One other thing:

There is absolutely nothing misleading about the statement that 98,000 is the most likely result. We're discussing statistics.

- John Palkovic (above).

As far as I'm concerned the exact number of dead is not so important. It is many tens of thousands. Whether it's 80,000 or 140,000 dead, it's just not acceptable.

- Les Roberts (quoted in the Socialist Worker).

I would argue that Dr Roberts has no idea how many civilians died in Iraq. He is interested in politics and morality. I think that's great. Good for him.

CL says

>I would argue that Dr Roberts has no idea how many civilians died in Iraq

And after that he will argue that black is white and up is down.

dsquared,

hmmmm Seixon, I would very much like to know the methodology behind the following claim on your weblog

It looks to me as if I am posing a question, not making a claim. Those question marks can be a bit hard to parse sometimes, I know.

I find it hilarious that I have to even make a claim to rebut the findings of a study we know nothing about other than its (cherry-picked) conclusions. I provided facts, and asked a logical question.

Once again it shows that certain people will take the results of a poll as the truth, without knowing anything about how it was conducted. The fact that its conclusions showed results that were way off the UNDP study results, even though they were about a year apart, begs a question or two.

Yes, the GDP figure was for 2004, which I said. There is no GDP figure for 2005 yet since 2005, for us innumerates at least, has not ended as of today. Surely you know that they release GDP figures for a year after the year has concluded? So when the GDP figure for 2005 is released sometime in 2006, will you apologize to me if that figure shows an increase in the GDP for 2005?

In other words, do you think, in all seriousness, that after the 52% increase in the GDP in Iraq in 2004, that, in order to facilitate the supposed doubling of unemployment the Iraqi research poll claims to have occurred over the span of one year, the GDP figure for 2005 in Iraq was -52% or something like that?

Tim,

Thanks for censoring my comments. Shows that instead of debating something, you would much rather just delete and cover up. You have not responded at all to the list of false statements I have compiled on you, and you say I have my own thread to use my false statistics claims, but you never comment there.

*[I will not have you take over this thread with more of your bogus claims about statistics. I don't need to comment further in the other thread. Your claims have been refuted and you are the only person who doesn't know that. Tim]*

Oh, and Lambert, you seem to take the position that Mr. Roberts cannot be disingenuous. You say that he denied wanting to influence the election. Really? That's one thing I addressed in my first "debunking" of the Lancet study, which was riddled with some errors as I have acknowledged. Well, the statistics part was, not the part about Roberts' motivations and statements. He said:

"On the 25th of September my focus was about how to get out of the country," he recalls. "My second focus was to get this information out before the U.S. election."

If he did not want to influence the US election, why was his secondary focus, aside from being alive, to put it out before the election?

Then he says:

"This was going to do more good in terms of changing policy if it came out in October than if it came out in November," he says. "But we never had any delusions that this might affect the U.S. election."

He has also acknowledged that Kerry and Bush would pursue very similar policies regarding this regardless of the outcome. So in other words, he says he wanted to put the information before the election to influence policy, although policy did not hinge upon the elections, even according to him.

Does that make any sense?? He never had any delusions about it making a difference, yet he still rushed it to publication before the election? Why? To influence policy, even though he has said himself that it wouldn't hinge upon the election??

See, the plausibility of all of this just looks hilarious on its face. So Roberts had to invent a new excuse for publishing before the election:

Mr. Roberts insists that his primary motive for rushing the paper to press was not political. He says he is glad the paper appeared before the election because he was concerned for his Iraqi colleagues' safety. Had the paper come out after the election, he argues, it would have looked like a cover-up. Dr. Lafta, he says, "would have been killed -- there is just no doubt."

There we go. Now Roberts no longer has to explain why he published before the election, he did it to save the lives of his colleagues! What a noble guy. Only one problem:

Dr. Lafta, in an e-mail message to The Chronicle, disagrees: "My personal opinion is that this was an unjustified fear."

Ooops.

Yes, Lambert, Roberts has denied he wanted to influence the elections. The problem is that the denial isn't plausible, and you are once again uncovered as an uncritical Roberts-apologist.

>Surely you know that they release GDP figures for a year after the year has concluded?

Actually Seixon, the ESTIMATE of Iraqi GDP growth for the year ended December '04 was released by the IMF in October 2004.

As regards the unemployment figure, you are more-or-less correct for a change (by the laws of probability it had to happen eventually).

Different sources for Iraq's unemployment rate show it at anything from 30% to 70%.

There's a huge unexplained variance in the figures which may, for example, be caused by people lying to government researchers about their employment status to protect their eligibiltiy for food rations.

By Ian Gould (not verified) on 31 Oct 2005 #permalink

Seixon, if the issue of Iraqi civilian casualties had become front page news before the election, it probably wouldn't have made a difference in the election, but it might have forced the candidates to address the issue. Or so Roberts thought. That's how it's supposed to work in a democracy--bad things are brought to light and politicians are forced to address the problem and the problem is solved, either by someone new coming into office or by the incumbent who wants to avoid looking bad. n practice, Roberts clearly wasn't paying close attention to American politics in 2004. Both candidates avoided the torture issue--Bush for obvious reasons and Kerry because he was afraid concern over that would be used to make him look weak. (Kerry--concerned more for the rights of al Qaeda than for those of Americans. I could write the attack ads myself.) So it was incredibly naive of Roberts to think that bringing up Iraqi civilian casualties during the election campaign would make any difference. I think his motivation was entirely justified--the issue deserves much more press, but boy, as he said he sure didn't understand his own country very well.

By Donald Johnson (not verified) on 31 Oct 2005 #permalink

I listened to the This American Life episode mentioned in comment 8 and found it very interesting. Particularly poignant was the fact that Garlasco, who questioned the accuracy of the study without having read it, had been just a few weeks before one of the people in charge of selecting the targets for the bombs. You can't make this stuff up.

Regarding Roberts' interest in seeing the study come out before the election (which from the TAL interview did seem clear enough), I find it not at all amazing for him to want to see that happen once the study was complete. As a consumer of US media, the great extent to which civilian deaths are discounted is excruciatingly apparent to me. Why would Roberts conduct these studies at all if he had no interest in seeing the information brought to public attention to the greatest extent possible? Apparently some of the commenters on this post would have preferred it if Dr. Mengele had been around to do the study instead. It seems that the worst excesses of imperialism have never lacked defenders.

By Steve Bloom (not verified) on 31 Oct 2005 #permalink

Ian,

I see that the State Department estimated a 16% GDP growth rate for 2005 back in August. They also have an unemployment rate at 28% for 2004. Not sure how they derive their numbers, hopefully not from some simple poll. If that is the case, then it seems as though the unemployment went up from the time the UNDP did their survey, until the end of the year. That doesn't make much sense though, seeing as how the GDP growth was 52% in 2004, and estimated to 16% in 2005. Could also be explained by varying definitions of unemployment (the UNDP used two different definitions and came up with either 10 or 18% on a national level back in May 2004). In any case, the rate of what, 40%, in southern Iraq from the MoD poll seems a bit high.

Donald,

Just look at Roberts claiming that his Iraqi colleagues were in danger if he didn't publish right away, and watch them deny it. This was his response when he was asked if his timing of publication was political. Hmmmmm.

Steve,

You seem to be attributing a point of view to me that I do not have. It is not my opinion that he should never have published his findings. I think it would have been more appropriate for him to have spent more time conducting it, and getting a better sample. Beyond that, I think that if the security situation didn't warrant doing a good enough survey, then he should have just postponed it.

So of course, none of that was an option because he had a presidential election to influence. Which basically shows that he was politically motivated in conducting the survey. Him making up other reasons he had to Fast-Track the survey at the Lancet are also painfully obvious.

Roberts: I had to publish soon for the safety of my Iraqi colleagues

Iraqi colleague: wtf are you talking about man?

That pretty much sums it up.

Oh and Tim, is there a reason why you have suddenly decided to start censoring my statistics? I mean damn, you went for a long time without doing it, but now you are. Why? I sure hope it isn't because I have finally stumbled onto the elephant I was looking for earlier...

You never refuted what I have been saying recently, in fact, you ceased to comment after I came to my new conclusion.

The commenters on my blog are side-stepping my new conclusion left and right. By gosh, I think I struck gold!

You also have not addressed the numerous patently false statements you have made in the process of the previous debate. Kevin wanted to see them, I gave them, and he never commented. Ouch.

Censorship is not typically a good sign. Just know that over at my blog, you can comment all you like and I will not censor you as long as you don't spam. You can even go and call me a bunch of bad names, I don't really care.

Even your adorers are becoming alarmed with your new practice of censoring me. You should welcome my comments, as they seem to draw in some traffic...

Seixon, Roberts has conducted several similar studies using essentially the same methodology. These studies were received with approval. Can you point to substantial differences between the Iraq study and the prior ones? Please be specific.

By Steve Bloom (not verified) on 31 Oct 2005 #permalink

Yes, Steve, I can. I already did. The methodology used for Iraq was completely different from what was used in the DRC study, for example. I already posted a link to the DRC study somewhere around here, and anyone interested can read it and see how different it is from the Iraq one. The initial sampling methods were different, and also, the DRC one did not use the "grouping process" that has become so infamous in the Iraq study.

I don't think Roberts should be defensive if asked about the timing of the paper publication. Obviously the study was meant to have an effect on the real world--it showed that the chief cause of increased mortality was violence, much of it by coalition forces, and Roberts thought it would have more impact if published before the election. His political judgment was way, way off, of course. You couldn't even get the Presidential candidates to talk about the torture scandal, let alone some statistical study on Iraqi mortality.

By Donald Johnson (not verified) on 01 Nov 2005 #permalink

Donald,

Was it really a revelation that violence accounted for an increase of deaths in a war zone? Was there anyone who doubted this would be the case? Newsflash: people die during a war.

Roberts' media blitz was basically a "cat drinks milk" story.

You can't deny that Roberts was fudging in trying to explain his timing for the release, as talking about the safety of his colleagues. Complete BS and just trying to row away from the fact that the timing was entirely political.

I don't know if Roberts's colleagues were in danger or not, or whether Roberts thought they were in danger and they really weren't. You might or might not have heard, Seixon, but a lot of members of the Iraqi professional class have been murdered, often by insurgents for the crime of working with Americans, and I knew that independently of what Roberts or his Iraqi colleagues have said about the subject. If the insurgents got wind of what the paper concluded (that Americans were causing the bulk of the killing) and thought that the study was being suppressed, it wouldn't be unreasonable for Roberts to fear for their lives. Maybe his Iraqi colleagues weren't afraid--it would still be natural for Roberts to feel some responsibility for them.

As for violence causing deaths in war, yeah, Seixon, most people, including world-class epidemologists, have heard about this. Duh. But what Roberts and others have often found that in wartime the biggest cause of death isn't violence--it's often disease and/or famine. In the American civil war, for instance, I remember reading that only about 200,000 of the 600,000 deaths were from violence. A lot of soldiers got sick and died. Roberts thought that, for instance, increased sewage in the streets because of destroyed infrastructure might cause increased mortality. To his surprise, violence turned out to be the main cause of increased mortality in Iraq. That wasn't obvious ahead of time, and btw, if his figures for violence are overstated as you suspect it's quite possible he's wrong. You ought to think of these things yourself, instead of going for the cheap shot.

By Donald Johnson (not verified) on 01 Nov 2005 #permalink

Donald,

Last I checked, warfare has changed since the civil war. Has it not? Also, last I checked, we have developed many things to fight disease and many of the problems that led people back in the old times to die. You know, like medicines, vaccines, and so on, and so forth.

Iraq is also not like Africa, where you would expect lots of disease and the like. Iraq is a semi-modern society and cannot be compared with the DRC.

Thus, to say that it was unexpected to find out that perhaps the largest cause of death in the Iraq war was violence... That would be suppressing just about all information available on Iraq and the state of warfare in the 21st century.

People in Africa kill each other with blades, knives, and guns. People in the Middle East kill each other with bombs, planes, rockets, RPGs, assault rifles, you get the picture.

Gee, I wonder where more people will die from violence compared to disease....... (Here I am not even mentioning many other vast differences between Iraq and African nations...)

If you would have asked me in 2002 whether more Iraqis would die from violence rather than other causes in a prospective war, I'd have said yes without a doubt.

Doesn't take a rocket scientists, much less an epidemologist, to figure that out.

PS. As far as his excuse about his colleagues getting attacked: Dr. Lafta, only one of two Iraqis named in the study, did not think this excuse was justified. I'd say that the Iraqi "FREEDOM FORCE" has other things on their minds than hunting down mostly unnamed scientists who failed to publish a study that mostly helps their cause quick enough. If no one ever made it clear that they had taken more time to do the study, how would anyone even get the idea that it had been "suppressed"?

Here's a tip: don't aid tailors fabricating stories out of whole cloth, you're bound to get snipped.

Golly, Seixon, without a doubt you'd have known violence would top accidents, disease (with sewage in the streets), malnutrition (with malnutrition up). Why even bother to have surveys when people could consult you? I note that in the Lancet survey the split was 60/40, so violence wasn't that far ahead.

By Donald Johnson (not verified) on 02 Nov 2005 #permalink

I do believe surveys are done to find out some figures, not always to find out if such and such has happened. For Iraq, it just makes sense. Immense bombing, ensuing guerilla war, moderately modern healthcare structure... Seems to me like violence would be the obvious choice for top killer. You're entitled to your opinion though, as should I be.