Crooked Timber

Some of us had the death certificate inconsistency (or “inconsistency”) in good faith. It’s the one obvious point to make if you sit down and try to imagine reasons why this study might be wrong. (I then tried to figure out how the inconsistency could arise and fraud seemed to be the only answer.)

But anyway, it’s just another armchair objection.
People who don’t believe this study should be demanding that the government support an independent investigation by human rights groups and statisticians to finally settle the matter. Oddly enough, I haven’t seen any of the rightwing objectors urging this, but I’ve no doubt this is an oversight on their part that will soon be corrected.

Daniel, my impression is that the UN goes to the other hospitals and asks them for their death certificates. Do you have information to the contrary, or is this “Theoretically, this is how the Lancet study could be right?”

statistics agencies are often bottom of the queue after essential infrastructure, law and order and electricity)

Which makes it especially insane that aid agencies and donor countries all try to flood into post-conflict countries with grand plans for reconstruction, all based on no bleedin’ data! Is there any other field where so much is spent based on so little information? (obviously there are areas where much is spent despite information that the spender has chosen to ignore). But it never sounds good for someone to pipe up at the donor conference and say “how about a few quid for the central statistics agency?”

Even the numbers in Baghdad, which should be reliable since the death certificates are right there! in the building where they were issued! show the Lancet grossly overcounting deaths.

“Jane Galt” seems to be trying to set some kind of propaganda record for a single comments thread. Here’s the reply by the Lancet authors (from this paper: http://web.mit.edu/CIS/pdf/Human_Cost_of_War.pdf) to the various attempts to obfuscate their findings using the hopelessly incomplete numbers from the Iraqi health ministry:

The Ministry of Health in Iraq has published some numbers from time to time, but these are generally considered to be unreliable. The registration of deaths in Iraq has been an organized process for many years. Death certificates have traditionally been obtained for the deaths of all adults and older children. Death certificates are required for insurance claims, compensation, payment of benefits, and for burial. Cemeteries do not take bodies for burial without certificates. If deaths occurred outside of hospital, the bodies would be transported to the general hospital for the certificate to be issued. If there were doubts about the cause of death, a post-mortem examination would be carried out before issuing a certificate. Copies of the death certificates would go to the national offices managing vital registration. This process has continued through the current conflict, with death certificates being required for burial, and with information from certificates being duly recorded. However, the tabulation of data from registration of deaths in Iraq has suffered from the chaos of the current conflict. Beyond this, there is also a suspicion that records of death, particularly related to violent deaths, is being manipulated and only partially being released for various political reasons. Even with the death certificate system, only about one-third of deaths were captured by the government’s surveillance system in the years before the current war, according to informed sources in Iraq. At a death rate of 5/1,000/year, in a population of 24 million, the government should have reported 120,000 deaths annually. In 2002, the government documented less than 40,000 from all sources. The ministry’s numbers are not likely to be more complete or accurate today.”

Note that Juan Cole apparently doesn’t agree with you.

“Where are the bodies?”

This is a good question but according to Juan Cole, there’s a good answer: there are two major reasons that most bodies never pass through a morgue: (i): Muslim tradition demands that bodies are buried the day of the death in a simple wooden box, so there’s natural resistance and, (ii) while there are incentives for taking bodies to the morque to counteract this, they’re overwhelmed by the fact that the government is so infiltrated with death-dealing militias at this point that nobody wants to put their name on any sort of government form if they can help it.

http://www.juancole.com/2006/10/655000-dead-in-iraq-since-bush.html

Juan says there are no bodies because people are afraid to take bodies to the authorities.
Daniel says, Juan is wrong-there are bodies-and local government knows all about it-but the federal government doesn’t.

“When someone dies, you get a death certificate from the hospital, morgue or coroner, in your hand. This bit of the death infrastructure is still working in Iraq. Then the person who issued the death certificate is meant to send a copy to the central government records office where they collate them, tabulate them and collect the overall mortality statistics. This bit of the death infrastructure is not still working in Iraq.”

How do you know all this?

Sk

my impression is that the UN goes to the other hospitals and asks them for their death certificates

…

the numbers in Baghdad, which should be reliable since the death certificates are right there! in the building where they were issued!

…

So wouldn’t it be better technique to collect the data from the hospitals directly?

I think that people may be misunderstanding the issues here. Iraq is in a state of complete anarchy. It is not a place where you can just rock up to the hospitals, smile at the receptionist and pick up a neat manila folder full of the week’s death certificates. Even if anyone had the resources and armed guard necessary to carry out this little fantasy, they would use them for something else.

I’ll just continue to make be an asshole on this thread.

I understand that many here make their livings in some area of the the social science biz or the data biz, and this is how you like to talk about things. In the same way, I imagine that classical musical world knows what’s going on with the Baghdad symphony today, and the zoo people have the word on the Baghdad zoo. Great. So if statistics is your thing, this has been another chance to show your stuff.

Nonetheless, are the numbers decisive in any way? We seem to be arguing about a factor of four, maybe six (except for Soru, who believes that the Iraquis are actially better off now.) Granted that statistics, symphonies, and zoos are always important to people in concerned with them, is the difference between 100,000 and 600,000 at all important from our point of view ? Is there some kind of thershold between the two numbers?

Sure, 600,000 is a Quantitative Fact, but does having that fact tell us anything we couldn’t already know without the Fact?

Everyone paying moderately close attention has a lot of scattered, mostly qualitative information informing them that a LOT of people are getting killed in Iraq. There’s other information telling us that many normal functions (e.g. electricity) have been distrupted for years. We know since Rice’s visit that planes can’t routinely land safely in the airport. Reporters keep coming back and telling us that it’s much worse than we think. And so on.

None of the plausible interpretations or suggested revisions of the study really change the qualitative reality of what’s happening in Iraq.

Sebastian- A sample of hospitals is not useful in determining death rate that can be scaled up to the total population – even assuming that they have good records and that the officials in charge would let you see them – because you have no idea how many people use a given hospital. So you would have to collect the data from every single hospital in the country. With households, you can determine the average number of deaths per household, and the average number of people per household, and scale up to the size of the population. Read the study, especially the section on study design – it’s not that hard to understand.
http://web.mit.edu/cis/pdf/Human_Cost_of_War.pdf

Ok, this is where I have a problem. It is too chaotic for people to go to hospitals (which have a discrete small number) but they can do a household survey? That is the kind of thing that makes me skeptical.

See, we’ve fallen into the trap. Because Iraq is in chaos, we have no way of knowing that things are very bad there. No data, just anecdotal information. Your guess is as good as mine, really.

I notice that all the comments here about Part 2 of the post. This may be because Part 1 is unclear. Because of Davies’s normally crystal-clear writing style, I spent about five minutes wondering what my problem was that I had no idea what Part 1 was talking about. I was trying to figure out why Iraq being a young country had anything to do with its crude death rate being low.

Eventually I realized that “Iraq is a young country” should be read as “Iraq has a young population” (implying that its people, because young, are less likely to die of natural causes in a given year), rather than “Iraq was founded relatively recently” (implying that it doesn’t have a good records-keeping infrastructure or something like that).

“That is the kind of thing that makes me skeptical.”

I don’t understand. You don’t have any reason to believe that this study is invalid, you just think that they should have done a different study. A perfectly good response to a scientific article is: here’s a result, let’s see if we can confirm the results with a different study. But you seem to be saying, There could be more than one way to study this, therefore this way must be invalid and probably fraudulent.

No, I’m saying that a study which implies a civilian death rate higher than Germany during WWII deserves a skeptical eye since we aren’t, you know, firebombing cities.

But there wasn’t a civil war going on in Germany during WWII. Unless you count “Ubermenschen vs. Untermenschen”.

Guardian piece was great, by the way.

One thing I have been wondering about in trying to picture how the survey was conducted, is the number of households surveyed per day.

1,849 houselholds were surveyed over a period from May 20 to July 10. That works out at around 36 households per day. If we allow the researchers two weekend days off per week, it is 50 households per day. They were two teams, so each team would visit 25 household in a day.

If a working day is 8 hours, that’s around 3 households an hour, or 20 minutes per household.

Hm. I wonder, what’s the norm in this sort of work?

Let’s just perhaps take another example where there is a vast discrepancy between passive reporting systems and active surveillance: estimating the size of the AIDS epidemic. Case report systems, even with the strictest compulsory notification in place, under-estimate HIV cases by factors from 2 to 10 or even 20. The situation is even worse with reports of cause of death. That means that those who are serious about making estimates of the size of the population affected do careful sample or sentinel surveillance and come up with estimates which tend to be far larger than those captured under the passive surveillance systems, but over time have been found to be ‘more or less’ correct – the more and the less coming from the series of assumptions necessary in extrapolating the sample to the whole population.

Of course, AIDS case and AIDS death reporting is affected by stigma which is taken to cause much of the under-reporting, although a lot is also due to the inherent weaknesses of passive data collection systems. But it does not seem implausible to me that the lack of government in Iraq is at least as powerful a disrupter of completeness in reporting as is stigma in AIDS, ergo, I find nothing unremarkable in a big gap between statistically valid sample-based surveillance, and reported mortality numbers.

Emerson:

I don’t know why you feel it is unimportant whether the body count is 200,000 or 600,000. After all that’s 400,000 people we are talking about.

Besides, an actual number surely tells us something about the intensity of the Iraqi chaos. The comparison with Darfur has been made. You could equally compare with the Congo (the numbers of which I don’t have to hand).

Is the Iraqi death rate higher than that of Germany’s during WWII? Or is it only a quarter that?

I think that’s valuable information, worthy of a lot of painstaking nitpicking.

“Deaths need to be certified quickly, so that the body can be dealt with, but to tabulate and collate them at the same speed would require a ludicrously disproportionate amount of resources.”

Couldn’t you just make it a Sarbanes-Oxley requirement? :)

Jesus F Christ, I’m not talking about killing 400,000 people. I’m saying that for most of our purposes, “far, far too many” is as exact a number as we need, and that the study’s lower bound is “far too many”, and haf that or a quarter that would still be “far too many”.

If anyone here is involved in Iraqi demographics, Iraqi funeral planning, etc., my words do not apply to you.

And that’s in the UK, which is not a war zone.

And you say the flypaper strategy hasn’t worked

brucer:

While we’re at it, can we also all stop referring to the 600K violent deaths or 650K excess deaths in the Lancet as “civilian deaths?”

Yes. This seems to be Sebastian’s misunderstanding, above:

I’m saying that a study which implies a civilian death rate higher than Germany during WWII deserves a skeptical eye since we aren’t, you know, firebombing cities.

Sebastian takes the Lancet estimated death rate for everyone in Iraq, which is maybe 2.5%. Then he compares it to the civilian death rate for Germans during WW II, which was maybe 2.6%. The correct comparison, of course, is to the total German death rate, military and civilian (and death camp), which was 10.8%. (Germany numbers taken from Wikipedia.)

“No, I’m saying that a study which implies a civilian death rate higher than Germany during WWII deserves a skeptical eye since we aren’t, you know, firebombing cities.”

Posted by Sebastian Holsclaw ·

Sebastian, can’t you come up with some fresh talking points? That’s far from original.

The hospital/morgue count is an undercount: can we concede that much?

Undoubtedly. The fact remains, Hopkins finds that 80% of deaths are certified. I can accept that some of these certifications never see the light of day. But Hopkins also finds that between October 2004 – July 2006, 543,000 excess deaths have occurred (655,000-112,000 for the Mar 2003 – Sep 2004 period). This works out at a daily excess death rate of 822.

So 822 Iraqis are dying each day, of which approximately 658 are certified (80%). Yet as of 3 days ago, people were talking in terms of maybe 250 deaths per day. Meaning hospitals and morgues have been certifying 2-and-a-half times as many deaths as the least conservative estimates for the last 22 months.

It doesn’t mean Hopkins is baloney, but any objective observer should be forgiven for being sceptical.

Not that I’m objective…

barry,

What next?

Brownie, the real issue is the complete breakdown of civil society in Iraq. There are death squads in hospitals. Suicide bombs at funerals. Kidnapping is endemic and murder of kidnap victims is common. There is a shooting war, and with 3000 US dead (and an unusually low kill:wound ratio) there is almost certainly a high insurgent-to-coalition death rate.

Add all of these things up, and the possibility of a large casualty undercount by the usual methods is entirely possible. My main reservation is related to the clumpy nature of violent deaths and its correlation with neighborhood and ethnicity, which could skew statistical sampling. This can be difficult to quantify properly.

Jesus, I was just over at Galt’s. It was a Rorschach test of conservatarian posturing. “To me that looks like…. leftists epidemioligist bias!”

Gee you guys really fell off the carrot truck. OK, loved one dies (maybe by electric drill to the brain, lead poisoning, or one of the other popular methods). You don’t dare take him to the hospital or morgue because then they will know you are one of those and put you into the drill press. You know, or more likely your second cousins wife’s brother knows someone who works in the hospital and has access to a a stack of death certificates, complete with seal.

You need a death certificate. You give your second cousin 200 euros. He gives his wife’s brother 150 E. The brother gives 100 to his friend and back comes the death certificate. Everyone is happy (ok satisfied.)

Over on Jane Galt’s blog, one of her commenters suggests the peer reviewers should be shot.

Well, yeehaw! That’s how you settle a disagreement that good ole’ libertarian way. Intellectual debate is for girly men.

The 2004 report was hardly “debunked” by the UN report, which measured a different statistic. To the extent the two studies can be compared on an apples-to-apples basis, the UN report confirms the Lancet study. See here.

I’m still waiting to hear a defense of the study’s conclusion that there was little, if any, increase in the non-violent death rate. This is very hard to believe, given the well-documented public health disaster in Iraq (crumbling hospitals, failing sanitation, etc.).

While I think the qualitative conclusion of this study is right (hundreds of thousands of excess deaths), I don’t think it’s a particularly good study, and I think y’all ought to be careful about exagerating its virtues.

The concern troll has spoken.

If the goal of the study had been to estimate the percentage of deaths in Iraq for which death certificates had been issued, their data would yield a point estimate of 91%. Applying that to the actual number of death certificates issued yields a mortality estimate in the neighborhood of 50,000 or 60,000, which matches other studies (e.g. IBC) reasonably well.

The researchers, on the other hand, observed the sample mean of 91% and concluded that the population percent is highly likely to be less than 10%.

The defense of this conclusion given above is that a very large proportion of the issued death certificates are not being recorded. But this is hypothesis, not evidence. It is also a relatively easy hypothesis to test. The Johns Hopkins team could have taken the proffered death certificate data, compared them with official records and tallied the percent that are missing. They didn’t take that step, so they provide no evidence to either support or reject the under-recording hypothesis.

Until someone takes that step we have a study whose data appear to support two conflicting conclusions. One of those conclusions comes fairly close to other estimates (albeit derived in a very similar manner) and the other differs by something like an order of magnitude.

It is neither unreasonable to conclude that the Lancet case has thus not been conclusively made, nor to conclude that leaving this significant contradiction unexamined is a serious shortcoming of the study.

It really doesn’t matter whether it’s dangerous to go to the hospital, etc. The point is that the vast majority of families actually surveyed in this study were able to produce physical death certificates. They must have come from somewhere.

You can certainly theorize that there are a lot of fake death certificates or something – and then test that theory – but they apparently do exist, in large numbers, absent the possibility of actual fraud in the data collection process. You have to deal with this fact.

Daniel,

I don’t know how many interviews with Horton you’ve heard in the last few days, but he does a very good job of presenting these figures not just as the best estimate we have, but as likely to enjoy a high degree of accuracy. You know these are not the same thing.

Further, it’s not as if Horton restricts himself to presenting the figures, rather he is simply bursting with advice about foreign policy. Everyone is entitled to an opinion, for sure, and everyone is entitled to draw conclusions about the opinions of others and the appropriateness of those opinions with regard to time and place.

Whatever the degree to which you share Horton’s view of the situation in Iraq, do you not accept that less of the mud being thrown at the Hopkins study would stick were it not for Horton’s clearly and frequently expressed political views? Indeed, do you not accept that fewer people would be so quick to dismiss the study altogether?

Horton has, in my view, unnecessarily politicized the study. Which is not to say that it wouldn’t have been politcized without the additional commentary, just that his non-academic contributions have undoubtedly complicated matters.

Brownie, you’re by your own admission not objective, and you’re not a statistician, Now by your own criteria that means we’re allowed to sling mud at you. So will you just shut up and let us sling mud at you?

Eli Rabatt: Gee you guys really fell off the carrot truck. OK, loved one dies (maybe by electric drill to the brain, lead poisoning, or one of the other popular methods). You don’t dare take him to the hospital or morgue because then they will know you are one of those and put you into the drill press. You know, or more likely your second cousins wife’s brother knows someone who works in the hospital and has access to a a stack of death certificates, complete with seal.

You need a death certificate. You give your second cousin 200 euros. He gives his wife’s brother 150 E. The brother gives 100 to his friend and back comes the death certificate. Everyone is happy (ok satisfied.)

Out of interest, why you need a death certificate? (I think this was asked earlier in the thread.) And 200 Euros is a fair whack to pay for one of them, so is there some incentive for having one? (Presumably, compensation or insurance related.)

Matthew,

Thanks, my search didn’t throw that up for some reason. I hope you don’t think I’m being obtuse if I say that “a death certifcate was present” is still open to some interpreation i.e. was it “present” in that the families in question said they had one, or was it physically presented? My bet is the latter but definitive clarification would be good.

So will you just shut up and let us sling mud at you?

bi, the day you catch me trying to unpick the Hopkins methodology, you’d be entitled to throw as much mud my way as you can muster.

That comment looked better in the preview. I swear.

Lophakin: yep, you need the death certificate to get the body buried and to claim on life insurance, get bank accounts closed, establish probate, all the grisly business. (I suspect that overworked Iraqi undertakers might be stretching a point these days, but life assurers, never).

Brownie:

Whatever the degree to which you share Horton’s view of the situation in Iraq, do you not accept that less of the mud being thrown at the Hopkins study would stick were it not for Horton’s clearly and frequently expressed political views? Indeed, do you not accept that fewer people would be so quick to dismiss the study altogether?

Frankly no. I remember when people were rubbishing IBC as “absurdly high”. There is no way to portray the deaths of half a million people in a politically neutral way. I also note that the death counts in Darfur have not exactly been presented in a politically neutral way and nobody plays this kind of game with them.

At present, I think that the only critique which can’t be immediately dismissed as hackery is the one which Johan on the Harry’s place comments made; that the numbers-on-a-map method of geographical selection is new and hasn’t been field tested to see how it compares to “GPS grid” (used in the 2004 survey) and “pen-toss” (used in a lot of Roberts’ previous work). I don’t believe that there actually is a reason to believe that numbers-on-map would introduce a big upward bias, but it’s a sensible point. I think that the attempt to smear the Iraqi doctors as Sunnis and Ba’athists who obviously made the numbers up is just disgusting.

Brownie: You don’t get it, do you? I am already entitled to sling mud at you. If Hopkins is fair game for mud-slinging just because Horton supports his study and adds some “political baggage” to it, then since you’re not neutral either (you said, “Not that I’m objective…”), that means you are also fair game for mud-slinging.

So, shut up and get lost. And get me someone who’s actually neutral and objective to criticize Lancet’s study.

Well we have had literally hundreds of posts here on the Lancet study in the last few days. And yet, when all is said and done, the arguments against the study all boil down to three main points.

1: The argument from personal incredulity, much mocked (and deservedly so) by Richard Dawkins in the Blind Watchmaker. In this case it should be put ‘I, personally, having never been to Iraq, speaking no Arabic, not being a soldier, and never having been to a war zone (let alone a civil war), having no understanding of or training in statistics….I find these figures in credible! I don’t think I need to provide any more evidence than that.’

2: Dark hints about political bias, which mysteriously failed to bother them when similar studies in the Sudan and the Congo were published. (Have they checked the political affiliations of the authors of studies of excess mortality in the Congo and the Sudan? Do they care? Would they consider estimates of (e.g.) Sudanese fatalities to be grossly compromised if the authors of the relevant reports went on record as stating their opposition to murder and the actions of the Sudanese government?).

3: Arguments that it ‘doesn’t really matter’ because one opinion poll had a badly worded question that seemed to imply that a small majority of Iraqis thought it was ‘worth it’.

And that’s it. Note that most commentators don’t even bother trying to hide their political bias: the argument invariably goes ‘I found this paper emotionally unacceptable and so I went around trying to find evidence as to why it couldn’t be true’, not ‘I am a statistician and I happened to be looking through this article when…..’ etc.

Brownie s3z,

You do realise you sound like a three-year-old, don’t you?

Well, to put it briefly, I believe in adjusting my style to suit the audience. I’m sure you don’t have a problem with that.

Brendan: And for some reason, political bias on their Own Side™ is totally OK, while political bias on the Other Side™ is absolutely unacceptable.

I think that the attempt to smear the Iraqi doctors as Sunnis and Ba’athists who obviously made the numbers up is just disgusting.

I’m seeing rather a low degree of correlation betwen that and what I actually said. Care to reread it?

“I do happen to think that the fact successive polls show majority Iraqi support for the war is relevant and I do genuinely wonder at how widely known this is.”

I have pointed out that the question that was asked was grotesquely biased. However, even in the example you quoted, the question was not in any sense ‘do you support the war?’ It was: ‘do you think it was a good thing to get rid of Saddam Hussein?’ (that’s not a precise wording, but as I’ve argued, that was the jist of it). The key point is as I’ve always argued, to view the ‘war’ (actually an invasion) and the subsequent occupation as two discrete entities which have nothing to do with each other is a fantasy: it’s like stating that the Russian liberation of Czechoslovakia and Hungary had nothing to do with the fact that these countries were then assimilated into the Russian Empire.

In other words, it is invalid to infer ‘support for the war’ from the questions that were asked. Given your somewhat wide use of the word ‘error’ I think that counts as an error.

In any case, this is all totally irrelevant to the Lancet study. Even if 100% of Iraqis thought the invasion and occupation of their country was simply peachy, it would have nothing to do with whether nearly 700,000 of them are now dead who would otherwise be alive. Which makes it even more mysterious as to why you keep on raising this point in a comment thread that is, let’s not forget, about the Lancet study.

Incidentally, I’d like to apologise: apparently there aren’t 3 reasons as to why the Lancet studies (plural) couldn’t possibly be true: there are actually nine.

soru: And what, pray, is this “information directly available from this thread” you refer to?

(Oh, and don’t give me that “if you’re open-minded enough you’ll find it yourself because I’m open-minded but I’m still not going to bother finding it” line.)

I’m not sure why people give much credibility to the count of death certificates. The figures come from the Ministry of Health, which is controlled by Moktada Sadr, and is said to run death squads in the hospitals and morgues. In the past, Ministry officials have pretty much acknowledged that the figures are manipulated.
[both links require a subscription to the NY Times]

soru,

Good point about the survey apparently undercounting violent deaths in pre-invasion Kurdistan. There’s an obvious explanation for this: the interviewers spoke Arabic, not Kurdish.

Btw, I’m with Brendan et al in that not single substantial criticism appears to have be made against the study other than “F**k me that can’t be right.” Of course, with the lack of connections to international terrorism and no WMD, the humanitarian motive/justification is the only one left – and if these numbers are even vaguely correct than it’s a clean sweep for the entire strategy being clearly demonstrable as a complete and total disaster.

Clearly, the Lancet study presents a grotesque underestimate.

Who really thinks that interviewers are going to the most dangerous places?

The methodology itself is systematically biased to miss out lives lost when an entire family or village is wiped out and by not including the deaths of the widely reported huge numbers of foreign combatants.

Didn’t we go though all of this last time over the Fallujah ommissions where they deliberately excluded the worst figures to keep the numbers down?

Obviously a major science journal will round things down in the interests of not rocking the boat. They are no doubt responsible for the insensitive insistence that people produce death certificates for interviewers when clearly they should be allowed to grieve in peace.

Ragout, do you have a cite for the interviewers speaking Arabic and hte interviewees not? The more we know the more we can understand the survey.

David thanks for taking a break from sorting out these folks:
http://neweconomist.blogs.com/new_economist/2006/09/reproduction.html
It’s a shame that Daniel is letting his work get in the way of answering important questions.

Soru, how was the sampling done in your survey? Did they ask many women? People in Tikrit? Turkmen? I’m sure they must be extrapolating from a much, much larger sample.

Daniel: At present, I think that the only critique which can’t be immediately dismissed as hackery is the one which Johan on the Harry’s place comments made;

With respect, I don’t agree with that. For instance, there’s the quote to which Mettaculture drew attention in a comment at HP: ‘Coalition forces have been reported as targeting all men of military age.’ (along with references to a couple of newspaper articles.) This seems to me to be an inflammatory comment and out of place in what should be a scholarly study. Then there’s the misrepresentation of US Department of Defense statistics to which I drew attention in a comment at HP. They weren’t using DoD figures, merely relying on them as a backup, but still, they should get these things right. And then there’s the matter of the pre-war death rates.

Brendan: [As I pointed out in the other post, this is like asking Hungarians or Czechs in 1949, “Thinking about the hardships you have suffered since the liberation of your country by the Russians, do you personally think that forcing the Germans out of your country was worth it or not?” Of course the asnwer would be ‘yes it was worth it’ and of course this would spectacularly miss the point].

I must say, I fail to see why that would be an unreasonable question to ask of Hungarians or Czechs. In the other thread, you suggest posing that question in the 1950s or 60s, which might make it less sensible. FWIW, probably not all Czechs or Hungarians would have answered yes in 1949; and the overall response would be worth taking into account. Forgive me if I’ve missed an argument on this matter somewhere along the way, on some other thread.

Jack, the Lancet article says “All were medical doctors with previous survey and community medicine experience and were fluent in English and Arabic.” If any spoke Kurdish, the article surely would have listed it as a qualification.

It’s well-known that most Kurds in northern Iraq speak Kurdish as a first language.

Obviously a major science journal will round things down in the interests of not rocking the boat. They are no doubt responsible for the insensitive insistence that people produce death certificates for interviewers when clearly they should be allowed to grieve in peace.

Mmm … didn’t want to rock the boat about MMR either, did they?

What are you trying to gain? What is the magic number, lower than 600,000, that would make you happy, and make the Iraq war a good war? Why are you intervening here in such a kneejerk way?

You guys are like whipped boxers dead on their feet, waiting for the knockout blow, still punching feebly just because they have to, because they’re boxers. If you win this argument you win nothing. Why bother?

Our intervention has been a disaster. The statistical information here is more or less consistent with the qualitative reports we get—Iraq is descending into a goon squad civil war, and no one is safe except Americans in the Green Zone. The airport isn’t even safe. American leaders can’t announce their visits. There are large areas where no American can ever go.

The minimum estimate is 50,000, and it’s known to be an undercount. Does 50,000 seem like a small, OK nuumber to you?

[I realize that some of you are stats buffs and stats experts who have no opinion about the Iraq War. You guys carry on! We need freaks like you without political opinions if we are to get the truth!]

Obviously the people who did this study are illegal combatants. They released key information at a time calculated to interfere with a key part of the War on Terror—the Congressional Election. Only if this election is won can the terrorists be defeated.

Has anyone else noticed that every member of this “team” of so-called “statisticians” from the alleged John Hopkins university is American? I hardly think I need to point out what this might suggest about their unconscious biases…

According to calculations available on Tim Lambert’s blog, adding in the Fallujah data, while it increases the median prediction, also increases the variance sufficiently to make the 95% CI span zero.

Says Soru, yet again. It still isn’t true, though.

http://anonymous.coward.free.fr/misc/roberts-iraq-bootstrap.png

What are you trying to gain? What is the magic number, lower than 600,000, that would make you happy, and make the Iraq war a good war?

A number below zero would presumably make the question unambiguous, rather than debatable as at present. That is not remotely unlikely, given the smallest mistake in the survey technique.

Brendan:
This is really rather different from Brownie’s interpretation that a majority of Iraqi’s ‘supported the war’..

Is it a fair description of your position that Iraqis, collectively and according to the survey, can be assumed:

1. to support a war that deposed Saddam and resulted in a successful and complete withdrawl within the next two years.

2. to oppose a war that would result in the continued occupation of their country in 2014 (by analogy with 1956).

I’m guessing you think it is uncontested and not worthy of debate that ‘the war’ when used by Brownie refers to the latter.

[The question] was: ‘do you think it was a good thing to get rid of Saddam Hussein?’ (that’s not a precise wording, but as I’ve argued, that was the jist of it).

As Soru has said, that’s what is known in polite circles as a falsehood.

Two polls were cited, and the questions were in each:

“Thinking about any hardships you might have suffered since the US- Britain invasion, do you personally think that ousting Saddam Hussein was worth it or not?”

“As you know the United States, Britain and some allies removed the government of Saddam Hussein in Iraq in 2003. Do you think that this was the right decision or a mistake?”

When you are asked “was it worth it?” about anything, you are explicitly being asked to balance a benefit obtained against cost endured. You need to lift your patronization of Iraqis to new heights if you are seriously arguing that when answering such questions, Iraqis think only about Saddam and not the action taken to remove him, and all that this has meant.

When discussions reach a point where your interlocutor has nowhere to go but to deny the meaning of words written in black and white, it really is time to call it a day, absent even the semi-literate warblings of other commentators who insist on arguing against points I’ve never made and attacking positions I do not hold.

Having played host to numerous visits from the CT authors, I expected rather better from their commentators.

Time to stop racking up the comments over here and go back over there.

It was a blast. Kind of.

Brownie: You are completely wrong. “Was it worth it” is a completely accurate abbreviation of those questions. Go back to wherever you came from, you dishonest hack.

Brownie, though you’re right that saying “a death certifcate was present” is “still open to some interpreation”, that it follows “interviewers were to ask for a copy of the death certificate” I think the conclusion is that they were shown it.

You add (I think not in response to this), “the day you catch me trying to unpick the Hopkins methodology, you’d be entitled to throw as much mud my way as you can muster”.

You were critical of the original study methodology, and the statistical techniques used. Have you changed your mind on the original, or is difference in the methodology between the studies that makes the difference?

Well, well, if it isn’t David “I believe it is a statement of fact” Kane, open-minded torture apologist!

How did it feel like to undergo repeated waterboarding?

The furor about the numbers is successfully fogging the implications of the numbers. To me, one of the implications is how surprisingly little we know about the Iraq war – because it is reported (in English) almost exclusively with the American p.o.v. in mind. Nir Rosen’s book collecting his reportage of the war, In the Belly of the Green Bird, is one of the few war books I’ve read that actually mostly ignores the Americans and concentrates on the multiple factions that have been struggling in Iraq from the beginning. And it definitely presents a different war, one that even a reader of Informed comment and the daily news stories in the NYT and Washington Post does not see. At all.

Another implication is, of course, that the argument that Americans have a “moral responsibility” to the Iraqis and so must remain there is an even more brainless use of an empty phrase than we thought. Americans had a moral responsibility not to attack Iraq unless they were threatened, which they ignored – but the pretense that, after that moral lapse, now the moral Americans have a moral interest to mount an occupation is truly hilarious—rather like saying the Nazis had a moral responsibility to put France together after 1941. Putting that Welcome Wagon mask on a pillaging expedition has had the results we have been seeing for years. Moral responsibility has mainly consisted in Americans making completely cretinous Bungalow Bill decisions about Iraqi factions that they obviously have no business making, while profiting enormously from shoddy work or no work of reconstruction, which they are apparently incapable of doing, filling the country with private military forces and American soldier’s unconstrained by Iraqi law, and creating a massive jail system into which people regularly disappear for no reason. I suppose at least the latter shows that the Americans are following the you break it, you fix it rule – they broke Saddam Hussein’s system of torture prisons, and so they’ve been doing their best to restore them. And the end result has been an enormous death toll of the ‘natives’ and the continuing and incredible imperialist arrogance of the ‘occupiers’, who at the moment are engaged, incredibly enough, in trying to break up the country, fighting for the Badr brigades against Sadr’s militia, while Sciri rams through a federalism law in the joke parliament that is going to create a taliban like enclave of Islamicist rule in South Iraq.

Enough harm has been done to the Iraqis, I would think, to satisfy the American thirst to combine the pleasure of lynching with the pleasure of think that God blesses America. I certainly hope these Lancet figures count if the Dems win and they do investigate the conduct of Bush’s dirty war.

A number [of extra deaths] below zero would presumably make the question unambiguous, rather than debatable as at present. That is not remotely unlikely, given the smallest mistake in the survey technique.

As I have already explained to you, this might be true of the abstract statistical manipulation of the data, but the statistical interpretation of Iraq death tolls is only a supplement to a lot of other knowledge we already have.

But at the same time, notice how your fellow lefties are suddenly overwhelmingly convinced that the number has to be 600,000

My fellow lefties mostly responding to the inevitable barrage of criticisms by statistical illiterates pulling stuff out of their butts. I personally think that the study was a good-faith effort, but that in the present political climate kneejerk criticisms were inevitable. The ignorant, ad hoc nature of many of these criticisms has been effectively demonstrated. That’s a good thing, though to me it’s a bad thing that we’ve allowed trolls to set the agenda. The topic remains the war itself and the destruction it has wrought, not the differences between avrious quantitative estimates.

Brownie: you missed the point, again. Besides the fact that your quoted paragraph is not in my opinion an unreasonable description—- it wasn’t really meant to reflect only you and was badly worded. I probably should have left it out, but would you then have addressed any of the substance of the post? I somehow doubt it, but surprise me!

If you have a report from a reasonably reliable source (say, a well respected medical journal) that gives a result you don’t like, there is only one intellectually honest response to it. You simply cannot begin by trying to find flaws and expect to be taken seriously. That appoach may make you feel comfortable (being able to trumpet `ah, see, it’s broken’ at the first percieved flaw without any danger of learning something) but it is deeply flawed. The only honest approach is to first make sure you understand it (not somebody elses characterization of it) and then ask the question `can this be correct’? There is of course an opposite problem, that of grasping at numbers that support some position you hold and broading applying them without understanding either. I haven’t seen much of that going on in this current case.

Arguments from incredulity really don’t add anything of value to a discussion of this. You personally don’t give evidence (that I’ve seen) of understanding the study, other studies, or the methodologies involved. As far as I can see, you are merely (and ineptly) trying to poke holes in something you don’t like. This sort of nonesense is a waste of time to engage with, so I won’t any further.