Tim Lambert (deltoidblog AT gmail.com) is a computer scientist at the University of New South Wales.
I think it is worthwhile to update James Wimberly's comparison of surveys of deaths in Iraq. In the table below death tolls have been extrapolated to give a number of deaths due to the war so far.
| Survey | Violent deaths | Excess deaths |
| ILCS | 150,000 | |
| Lancet 1 | 290,000 | 420,000 |
| IFHS | 280,000 | 700,000 |
| Lancet 2 | 1,100,000 | 1,200,000 |
| ORB | 1,300,000 |
It is interesting to see that the IFHS ends up right in the middle, between the two Lancet studies. If you think that the IFHS study is reasonable then you must conclude that Lancet 1 has been confirmed and the critics of Lancet 1 were wrong.
What about the comparison between Lancet 2 and the IFHS? Some folks are arguing that the larger sample size of the IFHS meaning that its estimate trumps Lancet 2, but the larger sample size just reduces the sampling error. The top of the IFHS confidence interval for violent deaths (220,000) is much smaller than the bottom end of the Lancet 2 interval (420,000) so the difference can only explained by non-sampling error, and here the IFHS isn't necessarily better. It was too dangerous to visit 11% of their clusters and in the ones they could visit they say that as many of 50% of the violent deaths may have gone unreported. They have attempted to correct for these problems, but the corrections may be not be enough. For example, they use the IBC numbers to estimate the violence in the unvisited areas relative to those they visited. But the most violent areas are going to be too dangerous for reporters as well, so the IBC will tend to undercount deaths in those areas relative to less violent areas. This doesn't mean that the Lancet 2 estimate doesn't have non-sampling errors as well -- Iraq at present is just a really difficult place to survey. I think that the best we can do is guess that the real number of violent deaths in Iraq to date is somewhere in the range 300,000 to 1,0000,000. Even the lowest number in the range is a horrendous death toll.
The numbers for excess deaths are for Lancet 2 and the IFHS are closer to each other, which suggests that some of the difference in violent death rates could be deaths being misclassified (either as violent by Lancet 2, or non-violent by IFHS). Since the Lancet usually checked death certificates, it seems that they would be less likely to misclassify deaths.
Notes: I did the calculations a little differently from Wimberly. He used the published totals on a given date from the IBC to scale the figures -- I used the actual total number of deaths in the current IBC data base for a given time period. For example, the ILCS covered the first 13 months, so it is scaled by (IBC current total)/(IBC total for first 13 months). In addition, I only used IBC scaling for violent deaths. For excess non-violent deaths I assumed that the death rate was constant.
The IFHS did not publish an estimate for excess deaths, so I computed it from the rates in the paper. The IFHS FAQ states:
Further analysis would be needed to calculate an estimate of the number of such deaths and to assess how large the mortality increase due to non-violent causes is, after taking into account that reporting of deaths longer ago is less complete.
Taking this factor into account would reduce the IFHS excess deaths estimate in the table above.
Politics
Comments
I'm going to forward a link to this post to the public editor at the NYT. I don't see why the press can't do analytical pieces like this now and then. Or alternatively, I wonder if there's any chance they'd publish it on the op-ed page. Of course it would have to be submitted. (Hint.)
Posted by: Donald Johnson | January 15, 2008 12:32 PM
Your Lancet 1 violent death estimate seems to exlude Falluja. Is that correct? What would the violent death estimate be if you included Falluja, as the L1 authors did in every discussion that they give in the paper of violent deaths? I think that it would match more closely to L2 (and be much higher than IFHS).
For example, here are some statements about violent deaths from L1.
There is no doubt that the L1 authors feel that Falluja is representative of the devestation in many parts of Iraq and that, were another survey done at the time, it would have included at least similarly horrific cluster. From that, it seems like any estimate of the number of violent deaths from L1 ought to include Falluja. (This is not true for excess deaths, since the authors explicitly provide an estimate for that. They provide no estimate for violent deahts.)
Posted by: David Kane | January 15, 2008 12:42 PM
I'll second what Donald Johnson said. This would make an excellent "Op-Chart" as the Times calls them. They run an Op-Chart every couple of months by Iraq War booster (yet self-described critic) O'Hanlon. If he doesn't have an exclusive license, this would make a good change.
Posted by: David Kane's friend | January 15, 2008 12:57 PM
I see the Washington Post still has their Iraq casualties widget using IBC only. They show a "maximum count" of 88,004 "Iraq Casualties (estimates)" as of Jan. 14.
Posted by: David Kane's friend | January 15, 2008 1:01 PM
PS on the WaPo widget: In fairness if you click on the (tiny) "about these figures" link it will take you to an explanation that does mention Lancet 2. But what fraction of readers are going to do that? Since Donald hit up Times Public Editor Hoyt, I'll hit up WaPo Ombud Howell.
Posted by: David Kane's friend | January 15, 2008 1:07 PM
Given that ORB admitted to problems with their survey I'm surprised that you would try to throw it in with the others.
Posted by: James | January 15, 2008 1:42 PM
You really should have the date ranges in that grid. The various studies covered different time periods which is a big deal if you're comparing raw total numbers.
Posted by: dbomp | January 15, 2008 1:48 PM
Given that ORB admitted to problems with their survey I'm surprised that you would try to throw it in with the others.
Which of these surveys didn't admit to problems?
Posted by: Crust | January 15, 2008 1:50 PM
So did the new IFHS survey, problems in execution so extensive they didn't even survey baghdad or anbar (due to danger), and problems in analysis because they had to fill those holes with estimates.
Posted by: dhogaza | January 15, 2008 1:56 PM
"There is no doubt that the L1 authors feel that Falluja is representative of the devestation in many parts of Iraq and that, were another survey done at the time, it would have included at least similarly horrific cluster."
They thought it was likely that any sample of 33 clusters might contain a horrific cluster, but to say they thought there was no doubt is an overstatement.
And people who criticized L1 at the time criticized its estimates with and without Fallujah. The 100,000 total excess widely cited (with about 57,000 violent deaths, as the National Journal article mentioned) was widely cited and these more conservative numbers were widely dismissed, though now supported by the latest study.
Posted by: Donald Johnson | January 15, 2008 2:10 PM
Your Lancet 1 violent death estimate seems to exlude Falluja. Is that correct? What would the violent death estimate be if you included Falluja, as the L1 authors did in every discussion that they give in the paper of violent deaths?
David, i used this opportunity to again work through the Lancet 2004 paper with a "falluja" search and i would advice everyone to do the same.
http://web.mit.edu/humancostiraq/reports/lancet04.pdf
they give numbers for both cases (with and without Falluja) in most cases and argue SPECIFICALLY about the exclusion for the confidence interval. i still do not see the basis for your attack on the authors.
More thana third of reported post-attack deaths (n=53), and twothirds of violent deaths (n=52) happened in the Falluja cluster. This extreme statistical outlier has created a very broad confidence estimate around the mortality measure and is cause for concern about the precision of the overall finding. If the Falluja cluster is excluded, the post-attack mortality is 7·9 per 1000 people per year (95% CI 5·6-10·2; design effect=2·0). ... As mentioned above, the Falluja cluster is an obvious outlier and might not belong with the others. When included, we estimate that the rate of death increased 2·5-fold after the invasion (relative risk 2·5 [95% CI 1·6-4·2]) compared with before the war. When Falluja was excluded, we estimated the relative risk of death for the rest of the country was 1·5 (95% CI 1·1-2·3)
and i can not understand, how you manage to sum up this:
We suspect that a random sample of 33 Iraqi locations is likely to encounter one or a couple of particularly devastated areas.
with this:
There is no doubt that the L1 authors feel that Falluja is representative of the devestation in many parts of Iraq and that, were another survey done at the time, it would have included at least similarly horrific cluster.
the restriction "at the time" is an important one, btw. the concentrated destruction that happenend in Falluja has not been repeated afterwards. and a poll today would most likely not bring up a similar cluster in falluja, as the city has changed massively as population numbers sank and increased again.
Posted by: sod | January 15, 2008 5:53 PM
and so the misrepresentation continues... David, is there anything you have ever read that you haven't turned completely on its head?
Posted by: SG | January 15, 2008 7:09 PM
So if IFHS proves in the minds of David Kane et all that L1 and L2 were fraudulent it presumably also proves ILCS is fraudulent.
I look forward to David's posts on this.
Posted by: Ian Gould | January 15, 2008 8:52 PM
What is everyone's problem? I quoted the study. If I felt like it, I could provide numerous quotes of Roberts emphasizing over and over again how the 98,000 number was "conservative", how the true number was almost certainly much higher. It is clear from every public statement that I have seen from the Lancet authors that their honest belief, as they say in the paper, is that the Falluja cluster is not "atypical." It may be an "outlier" in the sense that it is a much bigger number. But is an accurate outlier. There were, in 2004, many equally decimated neighborhoods in Falluja and elsewhere which the Lancet authors could have sampled but didn't. That is why they are so comfortable, in the paper, saying things like "Most individuals reportedly killed by coalition forces were women and children," when such statements are only true if you include Falluja.
If you had asked the L1 authors, in 2004, what their estimate for violent deaths in Iraq was, they almost certainly would have included the data from Falluja, just as they included Falluja when making claims about all the women and children killed by US forces. And that is a perfectly reasonable and defensible opinion. And, since they do not cite an estimate for total violent deaths in the paper, Tim has no good reason to exclude that data. Certainly, an estimate that throws out all of Anbar is hardly comparable to an L2 or IFHS estimate that does not.
And, even better, if you include all the data from L1, the results for violent deaths for L1 and L2 match up almost perfectly! It would be nice if Tim gave us details on the calculation. But it looks like he takes a 57,000 violent death estimate from L1 which excludes Falluja and then scales up by 5.1. If you take the 200,000 estimate for violent deaths from L1 (which includes Falluja), then multiply by 5.1, you get 1,020,000. Presto! The L1 and L2 numbers in the table would be a perfect fit!
Posted by: David Kane | January 15, 2008 9:20 PM
David Kane claims:
Not so. From an interview with Richard Garfield:
Even Lancet denialist Neil Munro used the 57,600 figure in his NJ piece.
Posted by: Tim Lambert | January 15, 2008 9:30 PM
Why not at least assign blame for the deaths of women and children accurately: the cowardly insurgents who hide behind them. Would anyone here do the same thing should we be in a fight with a foreign force? I doubt it. I'm certain that we would send our women and children as far away from the fight as possible.
Posted by: ben | January 15, 2008 10:04 PM
I stand corrected! I did not recall that Garfield interview. But I have never seen Les Roberts or Gilbert Burnham say anything along the same lines. (Counter examples welcome.) So, please correct my claim to "If you had asked Les Roberts . . . "
Perhaps this was the first example of how/why Garfield had a falling out with the other authors, perhaps because he felt that their estimates were too high?
But describing Munro as a "denialist" seems a bit much. Would you also apply that terminology to the authors of IFHS? They claim that "Both sources indicate that the 2006 study by Burnham et al. considerably overestimated the number of violent deaths."
And, not to get off topic, but does anyone have thoughts on why Burnham and Roberts are confused about death certificates?
Posted by: David Kane | January 15, 2008 10:05 PM
To stay on the topic, David, how many L1 critics can you name who said that 57,000 deaths was a reasonable estimate for Sept. 2004? Why are you only interested in the possible mistakes of the Lancet authors?
And anyway, so what if Roberts did think one Fallujah-type outlier might be typical in a sample of 33 clusters? It was a perfectly valid thing to suspect, given the evidence. It turns out it probably wasn't typical and the more conservative analysis was (if the IFHS paper is right) closer to the truth.
Posted by: Donald Johnson | January 15, 2008 10:31 PM
Uh, none. Why would they? Who would care about a number that does not appear in the paper? Why not ask me about L1 critics who thought that 40,000 or 80,000 or pick-your-favorite number was reasonable?
Because we are trying, I thought to have a good faith discussion of how to compare the IFHS number of 151,000 with other numbers. It seems fairly comparable (but not perfectly) with the 601,000 number for L2. (The comparison is not ideal because L2 includes several deaths from July 2006, while IFHS does not.)
But, if 151,000 for IFHS is fairly comparable with 601,000 for L2, what is the comparable number for both of them for L1. Since L1 does not answer that question directly, we need to answer it ourselves. Throwing out Falluja (and, therefore, all of Anbar) does not seem like a good idea to me. Although it does make L1 look more like IFHS, it also makes L1 look less like L2.
And again, my claim is that if you had asked Les Roberts for his best guess, based on the data that he had collected for L1, about violent deaths in Iraq, I have no doubt that he would have said something much closer to 200,000 than 50,000. Indeed, anyone who still thinks that L2 is closest to the truth would be even more likely to say that today.
Perhaps Tim could ask Les: What is the best estimate for the total excess violent deaths in Iraq through Sep 2004 using just the data that was available at that time from L1?
Do you really think that Roberts, who is always happy to use Falluja data when complaining of US bombing, would really say 57,000? I'll take the other side of that.
In terms of violent deaths, L1 and L2 are quite consistent.
That's why I think that Tim's table is misleading.
Posted by: David Kane | January 15, 2008 10:56 PM
"Uh, none. Why would they? Who would care about a number that does not appear in the paper? Why not ask me about L1 critics who thought that 40,000 or 80,000 or pick-your-favorite number was reasonable?"
Okay, I will. There's nothing sacred about that precise figure of 57, 600. But it's clear to anyone reading L1 (including non-experts like me) that if 100,000 excess deaths occurred then something like 40,000 or 60,000 or 80,000 were by violence. Something in that neighborhood is a reasonable estimate, given L1 data. Somehow I suspect you know that David, and you're just trying to dodge the question. So why didn't L1 critics say that the number of violent deaths implied in L1 is quite reasonable if you exclude Fallujah?
They didn't do this, of course, because even without Fallujah the death toll was much higher than IBC's and before L1 was published, even IBC was sometimes criticized by prowar types as giving numbers that were too high.
As for Tim's treatment of L1, he could break it into two pieces if he wants--one with Fallujah and one without. That'd be fine. Lancet1 did the analysis both ways. But somehow you think he should only do it with Fallujah, because it just sticks in your gut that any aspect of the Lancet papers might be supported by later evidence.
And yes, Les Roberts might pick the higher number. But David, outside of your own head the world doesn't revolve around your vendetta with your nemesis. Tim's site was the place to go for Lancet 1 debates and people on both sides of the argument generally used the more conservative Fallujah-less numbers.
Posted by: Donald Johnson | January 15, 2008 11:17 PM
Ah, but it's part of the strategy. If like David, you have nothing new to say, and if what you say is often wrong, and if people point this out to you, and show you evidence, what to do, what to do.
What David does, is endlessly recycle at length and claim that everything and everyone who does not agree with him is wrong, false, degenerate, loves Sadaam, etc.
Soon (OK some of us have more staying power than others) David hopes that his bad dream will go away as he blathers on.
The posts that really get me, are the ones where David discovers something new. And he shares it with us, at length as usual, and it usually is about something that anyone with half a brain who was paying attention thought about 20 years ago because he or she was paying attention while David was not paying attention. Probably because in his usual little davidlike monochromatic grinding way he was doing something else, at length, in detail, with twenty repetitions.
And oh yes, about 20 years ago anyone who was paying attention figured out that that bright new davidthought had a huge flaw. So we had a beer with a friend and smiled.
**20 years is roughly when the first Gulf war and the Afghan expulsion of the Soviets occurred, but obviously the current war and the Lancet studies are not that long ago, it's just that David makes time fly.
Posted by: Eli Rabett | January 15, 2008 11:39 PM
Well, as for me, it never really came up. Don't forget, I was quite comfortable with high numbers more than a year ago.
Those numbers (pre IFHS) are, obviously, perfectly consistent with 57,000 violent deaths through Sep 2004.
Again, I have no dog in this fight. If you want to compare IFHS/L2 with Anbar to L1 without Anbar, be my guest. I just don't think that the comparison is that useful and, in fact, causes the L1/L2 numbers to be much more different than they should be.
Perhaps we agree that two lines in the table (one for L1 with Falluja and one for L1 without) would be better than what Tim has now?
Posted by: David Kane | January 15, 2008 11:56 PM
How many of these deaths are combatants?
Posted by: wildlifer | January 16, 2008 12:33 AM
All,
You are struggling in vain. Much as I admire your tenacity in trying to corner Kane into making some kind of sense, you must never forget the wisdom of Upton Sinclair:
It is difficult to get a man to understand something when his salary depends upon his not understanding it.
In Kane's case, it is not his salary that is put in jeopardy by understanding, but his hopes for wealth and status through adoption and celebration by the right-wing establishment.
Posted by: Sortition | January 16, 2008 12:36 AM
Yes, the excess death rate in Iraq is terrible!
(PS, lets hope nobody catches on to the fact we oppose the surge tooth and nail, despite it reducing death rates dramatically. People might think we are huge hypocrites who don't really give a toss about Iraqi deaths after all, and are simply using this issue as propaganda!)
Posted by: Davis | January 16, 2008 1:15 AM
"Why not at least assign blame for the deaths of women and children accurately: the cowardly insurgents who hide behind them."
The question of how many people died is entirely separate from the question of who is responsible for those deaths.
Other than the IBC people, who have their own ax to grind, virtually everyone criticising the Lancet figures appear to be doing so because they support the decision to invade Iraq.
Political partisanship is not a good starting point for rational analysis.
Posted by: Ian Gould | January 16, 2008 1:38 AM
Davis (Jefferson I assume) writes: "we oppose the surge tooth and nail, despite it reducing death rates dramatically. People might think we are huge hypocrites who don't really give a toss about Iraqi deaths after all, and are simply using this issue as propaganda!"
Can you cite ANY examples of anyone on this site saying anything about the surge?
Posted by: Ian Gould | January 16, 2008 1:43 AM
"Would anyone here do the same thing should we be in a fight with a foreign force? I doubt it."
Yes because during World War II the British were widely condemned for housing women and children in their cities as human shields against German bombing.
Posted by: Ian Gould | January 16, 2008 1:47 AM
BTW Ben during World War II, the Americans actively encouraged Filipino children to take part in attacks on Japanese troops.
Posted by: Ian Gould | January 16, 2008 1:49 AM
Not to mention the Japanese using women and children as human shields to protect Hiroshima in 1945.
Posted by: dhogaza | January 16, 2008 1:55 AM
Ben, for several years I lived one block away from a very important Australian Naval base, which happens to be in the very middle of one of its most densely populated areas. Should that base have been bombed, and me accidentally killed in the resulting carnage, does that mean the aggressor force can blame my "cowardly government" which hid "behind me"?
When one is fighting a war to defend one's homeland from invasion, one tends to have to fight in one's homeland. Just because you don't think the Iraqis should defend their homeland doesn't make them any more or less cowardly for doing so in the cities and towns where they live.
Plus you might like to consider that the Geneva convention requires civilised nations to make a bit of an effort to avoid civilian targets. Or is that irrelevant to you?
Posted by: SG | January 16, 2008 2:26 AM
David, a "conservative" estimate of deaths excluding Fallujah maybe doesn't mean what you think it means. It means "biassed towards the null", i.e. not including data which you think are likely to bias your conclusions away from the null on account of their wierdness. As a critic of the lancet papers you should welcome their decision to adopt a conservative analysis plan.
My God, your ability to misrepresent commonly understood words and phrases is unparalleled!
Posted by: SG | January 16, 2008 2:29 AM
In keeping with my policy of avoiding colloquy with the trolls directly (and for all they puff him up, Kane is just that - a typical anti-scientific net troll), I want to point out that one of the problems here is a very common troll tactic. The Climate Audit people live and die by it.
Roughly, they don't, or pretend not to, get that if you only eliminate the error that doesn't favor you, you're not increasing precision, you're increasing distortion.
There are always elements of error and uncertainty. If there is no one with an axe to grind evaluating them, you nearly always find that overestimation and underestimation balance out if the science and statistical methods are fundamentally sound, especially over large samples or numerous different studies.
But what they claim is that if you can find any overestimation error at all (if they want an underestimate) or any underestimation error at all (if they want an overestimate) then that proves that you have to modify your conclusion in their direction. That's saying you have to base your conclusions partly on noise.
To denialist trolls, all reality is political and economic. There really, truly is no real world of laws and accurate data "out there", and everyone's a hustler and a demagogue. To this sociopathic worldview, the rules of science are just a con game, and the only way to determine reality is by a shouting match. They just assume the people shaving points in a "not us" direction will always balance out the people (them) shaving points in an 'us' direction, and a servicable truth will emerge from this adversarial process.
They routinely confuse this courtroom or forensic approach with the actual, also in a sense adversarial but also honest and cooperative process of peer review and filtering out of unsound science.
To give a concrete example, if your 1st grade class project wanted to know the average temperature at your elementary school, and you had testing that showed that in the field, children taking temperature were about as likely to go over as under, if you as a teacher who for some reason wants to brag about how hot the temperature is at your school remeasured only the lowest temperature spots, and only corrected the ones where the children had underestimated the temperature, and you ignored the others, both accurate and overestimated, you would end up with a hotter average temperature, but you would not be justified in saying the estimate should be moved in the direction of your revised estimate, nor would you be justified in saying you had improved precision, even though you are replacing erroneous measures with demonstrably better measures (for instance, you use a couple different thermometers and take a lot of measurements in the same spot).
Posted by: Marion Delgado | January 16, 2008 4:09 AM
Ben, taking the last refuge of a scoundrel, said, "Why not at least assign blame for the deaths of women and children accurately: the cowardly insurgents who hide behind them".
Ben, how much longer can you infuse such utter nonsense into these debates? First, according to international law, which the current DC regime ignores on a serial basis, the occupying force is responsible for providing security in the land they occupy. Thus, the US and UK are responsible, or should be, for every single death that has occurred there since they invaded and occupied the country. What you are saying is akin to saying that the deaths of French civilians in German-occupied France during WW II was due to the cowardly behavior of the resistance fighters who hid amongst the general population. One can only make such bland assertions as you do if one believes that the US owns and runs the world.
Second, you are intimating that the occupiers really care about the life of Iraqi civilians (much like they allegedly cared about south Vietnamese civilans during the Viet Nam war, Korean civilians during the Korean war, and, for that matter, the million or so dead Iraqis who died under the crippling sanctions regime between 1991 and 2003). The fact is, that while the occupiers do not intentionally set out to kill civilians (well, at least most of the time), probably mainly for propaganda reasons, the truth is that they don't really care much about them either. Not when there is a larger economic and political agernda to pursue. In my view, the occupiers view Iraqi civilians much in the same way we think of ants. As Noam Chomsky said last year, whenever he goes for a walk he probably kills many thousands of ants by inadvertently stepping on them. He didn't intend to kill them, but its an unfortunate by-product of his walking. This is a horrific perspective, but in all probability a true one.
Finally, you should read John Tirman's excellent post yesterday. The reason for the insurgency is that people tend not to like their countries to be occupied by an invading force, especially one which invaded on the basis of a brazen economic agenda. As I have said a million times, Iraq and the region surrounding it was described as the 'Greatest material prize in history' and a 'Source of stupendous strategic power' by the State Department more than 50 years ago. Nothing has changed. A recent poll suggests that 99% of Sunnis and 84% of Shia want the occupation to end. But now the Bush administration has chyanged the wording of 'permanent military bases' to an 'enduring relationship between Iraq and the United States', a sleight of hand that suggests that the invaders will never leave the country. But they never intended to: the invasion was all about controlling a region of immense economic importance to state and corporate planners, and stay they will, irrespective as to what the Iraqis want.
Posted by: Jeff Harvey | January 16, 2008 4:16 AM
I have no doubt that the death toll in Iraq is at least in the hundreds of thousands, and since the US started the war, we are indirectly responsible for most of those deaths. However, most of those deaths are NOT from US fire, deliberate or inadvertant. They are from car and suicide bombings and inter-tribal massacres and "ethnic cleansing." The situation is very like the one that prevailed in Yugoslavia after the death of Tito -- a strong dictator was holding the country together by sheer force, including secret police and labor camps/mass executions, and with the dictator removed, the various ethnic factions tried to kill each other. I place more blame on the Serbians, but that's beside the point.
We should have seen the insurgency coming. If we were going to conduct a successful occupation of Iraq (like the successful US occupations of Germany and Japan), with minimal casualties, then we should have gone in with at least three times the troop strength we actually had. We should never have fired the Iraqi army before disarming them. We should have controlled and suppressed looting from the very beginning. And if an insurgency broke out despite all that, we should have used an area-specific "oil spot" strategy like the British successfully used against the insurgency in Malaysia.
The war was not necessarily an immoral one in intent; removing a genocidal dictator and would-be conqueror is a good thing. But it was conducted in such an incompetent manner that hundreds of thousands of innocent people have been killed. The present administration deserves the blame for all of that.
Posted by: Barton Paul Levenson | January 16, 2008 8:14 AM
Posted by: Sortition "...you must never forget the wisdom of Upton Sinclair:
'It is difficult to get a man to understand something when his salary depends upon his not understanding it.'
In Kane's case, it is not his salary that is put in jeopardy by understanding, but his hopes for wealth and status through adoption and celebration by the right-wing establishment."
That brings up a question - who pays Kane's salary?
Posted by: Barry | January 16, 2008 9:10 AM
If you had asked the L1 authors, in 2004, what their estimate for violent deaths in Iraq was, they almost certainly would have included the data from Falluja, just as they included Falluja when making claims about all the women and children killed by US forces.
and
I stand corrected! I did not recall that Garfield interview. But I have never seen Les Roberts or Gilbert Burnham say anything along the same lines. (Counter examples welcome.) So, please correct my claim to "If you had asked Les Roberts . . . "
David Kane at his best. making a completely baseless claim. refuted immediately. instead of being all ashamed, rethinking his position, trying to be humble for a moment, he immediately brings up a minor adjustment to his nonsense.
In Kane's case, it is not his salary that is put in jeopardy by understanding, but his hopes for wealth and status through adoption and celebration by the right-wing establishment.
this is the truth.
Posted by: sod | January 16, 2008 9:23 AM
Tim could include an L1 entry with Fallujah if he wants and if one can decide on what number to assign, but what is more interesting is that the conservative analysis of L1 data, the analysis that the statisticians here supported, where one throws away the outlier that is so hard to interpret, ends up giving numbers extremely close to the IFHS. That's something that should be stressed to all the L1 critics, because they were so vehemently opposed to it and now embrace a report that supports one of the two analyses given in L1. Since you are so often cited as an authority by certain people online, maybe you should point this out to them.
Your claim that including Fallujah makes L1 more similar to L2 is sounds correct to my nonexpert mind, but it's not the spin you were putting on it last summer, when by your own version of statistical analysis, including Fallujah would have widened the CI so much the lower end would have included decreases in the death rate. That doesn't remotely resemble L2.
Barton--It's likely that most of the deaths in Iraq are caused by other Iraqis, but "most" could mean 70 percent or 90 percent or 95 percent. I think it's unlikely to be the case that the US kills several thousand insurgents per year (as shown by the statistics released by the government to USA Today some months back, where they gave a year-by-year breakdown of the number of insurgents they claim to have killed) and only kills several hundred per year (excluding the opening two months and the two assaults on Fallujah), which is the figure reported by Iraq Body Count. I'd bet that in the uncounted civilian deaths you'd find a higher fraction caused by the US than you do in the ones that reach the media.
Posted by: Donald Johnson | January 16, 2008 9:23 AM
First couple of paragraphs above were in response to David.
Posted by: Donald Johnson | January 16, 2008 9:25 AM
But, if 151,000 for IFHS is fairly comparable with 601,000 for L2, what is the comparable number for both of them for L1. Since L1 does not answer that question directly, we need to answer it ourselves. Throwing out Falluja (and, therefore, all of Anbar) does not seem like a good idea to me. Although it does make L1 look more like IFHS, it also makes L1 look less like L2.
David, it is all smoke screen with you. rather obviously, an overview looking at violent and total excess deaths, gives some variation but there is no major doubt about the effect of this war.
it is pretty absurd, that Kane, who thinks we can NOT refute the null hypothesis, is trying to argue which study L1 agrees with, if we make his weird adjustments.
again: the authors write A LOT about why they think Falluja is an outlier and should be handled as one. the war in Iraq changed massively after Falluja. the US army gave up the tactics of clearing towns by force. Violence increased MASSIVELY (this is not shown by the IFHS numbers. a major problem, that David avoids), but is more spread out now. it would be extremely unlikely for any later poll, to chose a "falluja L1" type of cluster. David is obviously doing nothing, but throwing sand in peoples eyes here....
Posted by: sod | January 16, 2008 9:51 AM
Surely this suggests that if the IHFS study had come across a single Fallujah-like cluster they would have got results similar to L2? (and given that they didn't survey Anbar or the dangerous bits of Baghdad, well, you can fill in the reasoning)
Posted by: dsquared | January 16, 2008 9:55 AM
The IFHS visited 1,086 clusters so one Fallujah-like cluster wouldn't make that big a difference (said he, not bothering his arse to do the arithmetic). But dsquared has a point nonetheless: the IFHS can reasonably be compared to Lancet 1 excluding Fallujah, since in different ways they both exclude the most violent areas.
Posted by: Kevin Donoghue | January 16, 2008 10:11 AM
Regarding salaries:
I think David runs a hedge fund. This leads me to believe that either he knows a lot more statistics than he presents himself as knowing, or my former default position, that the world of high finance is largely one of fluff and bluster, with a bit of luck thrown in.
Posted by: jodyaberdein | January 16, 2008 10:51 AM
Barton, I agree with some of the things you said. But the aim of the Iraq war was not to remove a genocidal dictator. If that was the case one must ask why the US has a habit of supporting genocidal dictators and even helping them into power. One can list Suharto, Montt, Marcos, The Shah, Pinochet, Habre, and even Saddam Hussein who were given huge financial and military support in full knowledge of their crimes. The reason, as Edward Herman has written, is not that the US government likes regimes that torture and murder their own citizens. It is just that these countries tend to be good places for businesses to invest, and for profits to be repatriated. It was recently shown that 20 of the top 25 regimes that receive the most US economic and miltary 'aid' have abhorrent humans rights records.
The aim for aggressing against Iraq was (1) regime change, and (2) to control a region of vital importance ot the US (and global) economy. The war party must have reckoned on the possible quagmire that would result, but they clearly balanced this against the benefits of controlling a region that planner George Kennan once said would give the United States 'Veto power over the global economy'.
As far as successful occupations go, Iraq is a proud nation with a long history and they know exactly what the motives of the US and UK planners were and are. They know full well what the invasion was all about, even if our corporate media in the west aims at 'dumbing us down' as to the real motives. People just do not like their countries to be occupied, especially by a foreign power that has contributred significantly to their misery in the past. They also do not like to be occupied when they know the real reason for the occupation has nothing to do with liberation and everything to do with plunder.
Bush and his regime weren't interested in allowing Iraqi economy to develop internally, as the war was aimed at outsourcing the profits to US contractors and corporations. All of the points you made were valid but must have been considered by Bremer and others in the Bush administration. But they were dismissed.
Posted by: Jeff Harvey | January 16, 2008 11:05 AM
Jeff Harvey writes:
[[The aim for aggressing against Iraq was (1) regime change]]
Wasn't that what I said?
[[All of the points you made were valid but must have been considered by Bremer and others in the Bush administration. But they were dismissed. ]]
Sorry, I don't buy into conspiracy theories, including ones where the administration knew ahead of time that the war would become a gigantic mess but deliberately did it badly for sinister reasons. I didn't vote for Mr. Bush, but he is not exactly a master of subtlety. If it looks like a giant fiasco and talks like a giant fiasco, it probably is a giant fiasco.
Posted by: Barton Paul Levenson | January 16, 2008 12:37 PM
Jeff writes:
Ben, how much longer can you infuse such utter nonsense into these debates? First, according to international law, which the current DC regime ignores on a serial basis, the occupying force is responsible for providing security in the land they occupy. Thus, the US and UK are responsible, or should be, for every single death that has occurred there since they invaded and occupied the country
That's utterly absurd logic. Yes the occupying force does have a responsibility to set up security forces, and they are doing that, but to blame them for every single act by Al Qaeda and anyone else who decides to blow up a funeral march is complete moral gibberish. You cannot believe that for a second.
Moreover, the very same people who say this are also those who oppose the surge which has reduced casualties. You people are extraordinary hypocrites.
Posted by: MikeT | January 16, 2008 4:58 PM
Imagine if we had handed the country over to Al Qaeda a few years ago as people like Jeff were begging us to do.
Posted by: MikeT | January 16, 2008 4:59 PM
Mike, so instead we handed it over to G. Bush and he did the job better than O. Laden.
Posted by: Eli Rabett | January 16, 2008 5:15 PM
Were the resistance fighters in France busy blowing up the general population as well?
Yes I am.
I disagree.
Oh brother.
Methinks he grossly overestimates the number of ants he kills on a walk.
It might be true for Noam, but I don't think it's true for the American Military. While the "insurgents" are busy bombing children, the Americans are busy trying to save them.
Posted by: ben | January 16, 2008 5:16 PM
HOW MANY OF THESE DEATHS ARE COMBATANTS???
Posted by: wildlifer | January 16, 2008 6:09 PM
46, MikeT writes:
"Yes the occupying force does have a responsibility to set up security forces, and they are doing that"
We're 4 years and somewhere between a half million and a million dead, and nearly 5 million refugees, into this war, and we 'are doing that.' That is as searing an indictment of the mass-murderous incompetence of this administration in the planning and execution of this utterly bungled war, as I've seen anywhere. Thanks, Mike.
"oppose the surge which has reduced casualties" I'd argue that the reduction of violence is at least a much a function of what US military has taken to calling the 'self-sorting' of the Iraqi population. Much of that violence was aimed at driving certain factions out of certain neighorhoods. It worked - the target populations are largely gone. The violence is dropping because its purpose has been accomplished. We failed to protect people - the violence is dropping because what we promised to protect people from, continued through the early surge, and now has largely happened already.
Meanwhile, the political objectives of the surge, the primary objective of the surge, has failed completely.
Pointing all this out does not make us hypocrites. It does show why we are outraged.
Posted by: Lee | January 16, 2008 6:13 PM
Lee, the most recent and credible study finds that the terrorists are responsible for killing 150,000 people, not half a million. I hate them even more than you do, but lets keep the record straight. We shouldn't use propaganda.
You can't on the one hand claim the political objective of the surge has failed, then pretend you are concerned about Iraq casusalties. The mainstream does understand the hypocrisy of this; that's why people rarely state these high figures as fact. You people cannot be trusted under any circumstances.
Posted by: MikeT | January 16, 2008 8:34 PM
MikeT, would you at least try to look at the top of the thread, before you make such absurd statements. the most recent of many studies finds some 150,000 VIOLENT deaths through mid 2006. Using their number, others have found that the total excess deaths derived form their data are in line with Lancet 1 and Lancet 2, even though the violent deaths are off by a factor of 4 from Lancet 2. Pretending that this means that there are only 150,000 deaths attributable to the decision to invade, is laughable.
"You can't on the one hand claim the political objective of the surge has failed, then pretend you are concerned about Iraq casusalties." Why in the hell not. Is it necessary, if one is concerned about Iraqi deaths, to stick one;s head in the sand and pretend that the political goals have been achieved?
The heinous number of excess deaths is clearly due at least in large part to the incompetence of this administration in planning and prosecuting this war - saying so does not mean in any way that I don't care about those deaths.
The political goals of the surge have failed completely - saying so does not mean that I'm not very, very much in favor of lower death rates. The MILITARY goal has largely failed - the protection of people where they live. Saying so does not mean I don't see the tragedy of the dead and the refugees that we failed to protect in their neighborhoods, or that I don't consider it a major improvement to see the falling death rates. Fewer people dying is good.
Ethnic purification of Iraqi neighborhoods is not good, and a failure. To point out that the reason fewer people are dying is a result of a failure to protect them from death and displacement over the past years, is NOT to favor additional death. Rather the opposite in fact, despite your despicable charge.
Posted by: Lee | January 16, 2008 10:23 PM
"most of those deaths are NOT from US fire, deliberate or inadvertent." We may find out that this is not the case.
There are new books about personal war experiences in Iraq (a publishing growth industry) reviewed here:
http://www.nybooks.com/articles/20906
which indicate a different story. We read that nothing much happened in the initial invasion for three days, until hell broke loose. The Marines took fire and lobbed about 2000 artillery shells (155 mm) over thirty-six hours into Nasiriyah, a city of about 400,000 people. Entering the city, the Marines found plenty of dead, including women and children.
This sort of thing was replayed and intensified as the forces got nearer to Baghdad. When the story is finally told, we may learn that (wild guess) 100,000 to 150,000 civilians were killed BEFORE Baghdad fell.
There is also the not-so-little matter of aerial bombardment, which has gone unreported in the U.S. news except for campaigns in 2005 (I seem to remember,) the beginning of 2007, and right now. This too leaves few respondents to interview.
Posted by: Lee A. Arnold | January 16, 2008 11:09 PM