Iraq death rate up since invasion

Rev · Post by **Rev** » Fri Oct 29, 2004 4:36 pm

The BBC says the Lancet (a medical journal) says 100k extra deaths in Iraq vs before the war.

Most other estimates are down in the tens of thousands, but they are often restricted to violent death, and some of them actually count only specifically reported violent deaths (thus being intentionally conservative to some degree). The US military does not even release an estimate of Iraqi civillians killed. Either they keep it secret for propaganda purposes or, more shockingly, they aren't even making one.

Wars normally kill a lot more people through more indirect disruptions. Food shortages, poor medical care, sanitation, etc wiping out the young, old, and previously sick.

3278 · Post by **3278** » Fri Oct 29, 2004 7:30 pm

3278 · Post by **3278** » Fri Oct 29, 2004 7:30 pm

BBC wrote:A study published by the Lancet says the risk of death by violence for civilians in Iraq is now 58 times higher than before the US-led invasion.

Unofficial estimates of civilian deaths had varied from 10,000 to over 37,000.

The Lancet admits the research is based on a small sample - under 1,000 homes - but says the findings are "convincing".

Responding to the Lancet article, a Pentagon spokesman defended coalition action in Iraq.

"This conflict has been prosecuted in the most precise fashion of any conflict in the history of modern warfare", he said.

And he's right. Moreover, the study is clearly biased and poorly executed. Plus, the estimate that the risk of death by violence is 58 times higher flies in the face of the facts: the most reputable - but still anti-war - estimates place the civilian deaths at under 15,000 in the last year and a half, one tenth of the number Hussein ordered killed in a single day when he got angry at the Kurds.

I really like the notion that it's the military's responsibility to count the number of civilian deaths. Has that /ever/ been done during a war?

Luckily, deaths by "other" have gone down over 300 percent. Think the US will get credit for that?

Rev · Post by **Rev** » Fri Oct 29, 2004 8:57 pm

It is quite possible that the military has not counted civillian deaths in Iraq. However it is absolutely certain that they have in the past done so. The military has entire books (by generals like Curtis LeMay) about how many people will be killed by bombing raids of various types compiled during and after WW2 because killing civillians was a parameter they were trying to maximize. Here they are supposed professionals in fighting, killing people, and nowdays supposedly not killing the wrong people. How can they possibly accomplish thier mission of reducing civillian casualties if they do not even attempt to estimate the number of civilian casualties they do cause. Unfortunately I also find it entirely possible that they are not collecting these statistics in Iraq either because it has not occured to them, or because the number is politically embarassing. Perhaps they are collecting it, and just not releasing it.

Some of those 10-30k numbers are talking only about people blown up, shot, burned, etc. I am not sure if all of them are, but I know at least two, including that big website one are. Yet if we bomb a sewage treatment plant and the incidence of water born disease doubles did the bomb not kill whoever dies from that as well? That is why it is important to study overall numbers in addition to immediate combat fatalities. The overall degredation in living standard associated with war is usually more deadly than the actual fast moving bits of metal. LeMay would have included those people when he was trying to maximize civillian deaths.

How is the study "clearly biased and poorly executed"?

How can it fly in the face of the facts when the whole point is that we don't know what the facts are? Are you saying that anti-war estimates of civillian deaths are indisputable facts, and that they perport to be complete estimates?

I pointed out that it conflicts with other reports, and speculated on a reason why, you just claim this number is wrong with no reasoning.

Rev · Post by **Rev** » Fri Oct 29, 2004 10:04 pm

Another story with responce from the british government and one of the "counts of individual deaths reported" groups. Sounds like he thinks it is probably too high. Note of this 100k 50% would be the sort of deaths reported by the other groups eg direct combat, so thier estimate is around 50k of those. Other estimates range from 10-27(or maybe it was 37?) thousand.

3278 · Post by **3278** » Sat Oct 30, 2004 1:32 am

Let me come back later with a less knee-jerk response. I do think the "study" is intentionally misleading and terribly biased, but not for the [pretty illogical] reasons I've stated so far.

Serious Paul · Post by **Serious Paul** » Sat Oct 30, 2004 4:35 am

From CNN:

While the major causes of death before the invasion were heart attack, stroke, and chronic illness, the risk of dying from violence after the invasion was 58 times higher than in the period before the war.

WTF? So the risk of being gassed or shot in the Hussein regime was less than the chance of your ticker giving out? Am I the only person who finds this hard to believe?

I mean, okay in the war you're likely to die-it's a war. But prior to the US invasion Iraq was not candy land.

The researchers surveyed nearly 1000 Iraqi households in September, asking how many people lived in the home and how many births and deaths there had been since January 2002.

So they asked, and accepted the answers given?

How scientific. Say Paul, how did you get those warts? Oh yeah frog got me man. It's not VD.

Also there's how many people in Iraq? Oh right:

While the researchers said the sampling was small, in an editorial alongside the survey, Lancet editor Richard Horton said interviewing more households would have improved the precision of the report, "but at an enormous and unacceptable risk to the team of interviewers."

Right. Well then this should be labeled GUESS.

While I certainly think it is fair to say people have died, and lots of them, I am not sure this is the best way to say it.

And not to cast too much doubt on Mr. Roberts, but this article plays nicely into a course of study that was introduced a year ago, or so. I am sure it doesn't help his position as Director of Health Policy at the International Rescue Committee, an NGO based in New York that provides relief to victims of war.

Crazy Elf · Post by **Crazy Elf** » Sat Oct 30, 2004 6:41 am

Paulie wrote:WTF? So the risk of being gassed or shot in the Hussein regime was less than the chance of your ticker giving out? Am I the only person who finds this hard to believe?

I mean, okay in the war you're likely to die-it's a war. But prior to the US invasion Iraq was not candy land.

Actually, Iraq wasn't that bad at all. It was pretty liberal for a Muslem country. Sure, it was a military leadership where the people really didn't have that much say in the election, but same goes for America with Bush running it.

"Those that we love and those that we hate are reflections of ourselves."

Serious Paul · Post by **Serious Paul** » Sat Oct 30, 2004 6:45 am

However, as much as I dislike Mr. Bush on occassions he did not, unlike Mr. Hussein, gas large segments of his population. There's no Republican campaign of genocide.

Anguirel · Post by **Anguirel** » Sat Oct 30, 2004 12:07 pm

Serious Paul wrote:However, as much as I dislike Mr. Bush on occassions he did not, unlike Mr. Hussein, gas large segments of his population. There's no Republican campaign of genocide.

That's just what the man wants you to believe!{/tin-foil-hat}

Crazy Elf · Post by **Crazy Elf** » Sat Oct 30, 2004 7:11 pm

No, instead he bombed large sections of other people's populations. I don't see how that's much better.

Serious Paul · Post by **Serious Paul** » Sun Oct 31, 2004 1:47 am

I do. And so, I would think, does the American people. You may not, and that's fine. I's funny and ironic considering who claim to be, but it's cool Hombre.

Crazy Elf · Post by **Crazy Elf** » Sun Oct 31, 2004 3:56 am

He's not blowing up goths. I don't see how there's any irony in me thinking he's a fuck.

3278 · Post by **3278** » Mon Nov 01, 2004 4:40 pm

Like Paul, I'm extraordinarily dubious of the figures claimed by this study, as well as the circumstances surrounding its release. If you read the full text of the study, several suspicious elements stand out.

Firstly, this item was fast-tracked for publication, and given only limited peer review. The Lancet claims this is because of the dangerous conditions in Iraq, but release only a few days before the US presidential election - when nothing more or less dangerous is occurring in Iraq - is dubious, to say the least.

Second, this is a cluster sample survey, meaning that 33 clusters of 30 households were sampled: instead of giving an answer about what is occuring in Iraq during the last 17 months, this tells us what's been happening only in those 33 areas. Furthermore, as Paul said, the questions are simply answered by the respondants, and not verified in any way whatsoever, lending further bias to the report.

Third, even when answers are clearly suspicious - such as the claim that "most" civilian deaths were women and children - no attempt is made to verify or explain those answers. Clearly, those committing this study had no interest in objective fact-finding, but only in reporting whatever would make the US and the US military look as bad as possible.

Even the most critical assessments of civilian deaths do not reach half of this "study." That alone makes me suspicious. That almost no information about the methodology of the study was released makes me additionally dubious. Every time I read further, more and more unscientific elements become clear.

Don't take my word for it: <a href="http://image.thelancet.com/extras/04art ... .pdf">read the study</a>. Pay special attention to the methodology given, and look out for such words as, "assumed," "believed," and "essentially." Read their own comments about the limitations of their verification process.

Look, I don't doubt civilian death is up during this period, compared to other periods. However, it is similarly clear that it is down compared to many of the periods of similar duration under Hussein. Moreover, the number of civilian deaths during wartime should hardly be compared to civilian deaths /not/ during wartime. What will more accurately portray the success or failure of this invasion is a comparison of civilian life and death in the ten years after the invasion to the lives and deaths in the ten years before. This study makes no attempt to make such a comparison, but only does whatever is necessary to assess the highest number of civilian casualties, despite groups like Human Rights Watch publicly claiming such numbers are impossible.

Does anyone believe these numbers are correct, or is it similarly clear to everyone else that this study, fast-tracked through peer review, pushed out to the international press days before the election, is an attempt to cast the current administration in a bad light immediately prior to the US elections?

Ultimately, if the numbers aren't believeable, and the intent is unscientific, we should ignore this study and move on to more likely claims. There are millions of reasons to suspect the invasion of Iraq is a spectacularly ill-founded affair: we don't need to invent new ones because 100,000 is a bigger number than 15,000.

Rev · Post by **Rev** » Mon Nov 01, 2004 6:00 pm

Awesome, 3278, you found the whole article! I was looking for that friday but could not find it. Will be reading it later.

They did make an attempt to go beyond just taking people at thier word, that is the part where they asked for death certificates to verify some of the deaths.

The study that I think should be done is a calibration study for the reported deaths type studies that are more common. What one does is given a fairly easy to collect peice of data that one knows is showing only a portion of the picture one does an intensive and therefor expensive study of a subset of the sources for that data to figure out how much the easier data set misses. Then one has a calibration to modify the easier to collect data. Thus one might say that for every reported death in bagdad there are approximately x additional unreported deaths, and y additional indirect deaths.

3278 · Post by **3278** » Mon Nov 01, 2004 6:39 pm

Rev wrote:They did make an attempt to go beyond just taking people at thier word, that is the part where they asked for death certificates to verify some of the deaths.

Specifically:

The Lancet wrote:Several limitations exist with this study. Most importantly, the quality of data about births, deaths, and household composition is dependent on the accuracy of the interviews. We attempted to confirm two non-infant deaths per cluster, but in four of the 33 clusters no noninfant deaths were reported, and in some clusters
interviewers confirmed deaths in more than two households. In 63 of 78 (81%) households where confirmations were attempted, respondents were able to produce the death certificate for the decedent. When households could not produce the death certificate, interviewers felt in all cases that the explanation offered was reasonable—eg, the death had been very recent, the certificate was locked away and only the husband who was not home had the key. We think it is unlikely that deaths were falsely recorded. Interviewers also believed that in the Iraqi culture it was unlikely for respondents to fabricate deaths.

Obviously, not much of a verification system, as they themselves admit. Most deaths are left unverified, and even those for which verification was attempted and failed were written off as "reasonable" or defended by an assumption that Iraqis are unlikely to lie about deaths. Very unscientific.

Which is unsurprising! The intent of the study and its release are, I think, surpassingly clear particularly in conjunction with the <a href="http://image.thelancet.com/extras/04cmt ... >companion comments</a>, which, amongst more defense of bad polling, attacks "democratic imperialism," and "this political and military failure." Again, I think the intent of the "study" is very obvious, and, insofar as its desire is to prevent further death and injury, suited to a medical journal. What is inappropriate, then, is not the inclusion of temporal, military issues in a medical journal, but the unscientific means by which they have attempted to make their point.

Rev · Post by **Rev** » Wed Nov 03, 2004 7:11 pm

I think your perception of what is "scientific" is pretty far off.

The essence of science is the collection and analysis of data. That is what the authors attempted to do. They collected data and analyzed it, pointing out that the data was damned hard to collect and thus imperfect. They attempted to validate thier data by checking some of it against death certificates. They discuss at length possible problems with the data. This is also what I am doing: collecting and analyzing data. Baldly rejecting the only study of its kind is what is unscientific. You don't have to be convinced, but it clearly is useful data.

It is not unusual at all for scientific data which is felt to be "important" in some time sensitive way to be rushed through the normally leisurely review process, most scientific data just isn't time sensitive in this way.

Interestingly you pick out only the "anti-american" conclusions for criticism. They also found that individual american ground soldiers were not killing very many people and that they were probably reasonably confused, or killing combatants, and that they apolagized.

They also make the point I was making in my first post, "It seems difficult to understand how a military force could monitor the extent to which civilians are protected against violence without systematically doing body counts or at least looking at the kinds of casualties they induce. This survey shows that with modest funds, 4 weeks, and seven Iraqi team members willing to risk thier lives, a useful measure of civilian deaths coudl be obtained. There seems to be little excuse for occupying forces to not be able to provide more precice tallies."

They also say that calibration of those other casualty counts is necessary because they are known to be conservative. Read the Iraqbodycount.net website to see that they agree they are underestimating even the direct combat deaths (let alone the indirect deaths which they do not attempt to estimate), they just spend almost all thier effort fending off right wingers who claim that they are vastly overestimating.

The refusal to collect and analyse data is truely unscientific, and that is (according to tommy franks and others) our military and governmental practice.

Luckily there are some scientific people out there so we will see more studies over time to give this one better context. Iraqbodycount page on other studies

Marius · Post by **Marius** » Wed Nov 03, 2004 8:28 pm

It is not unusual at all for scientific data which is felt to be "important" in some time sensitive way to be rushed through the normally leisurely review process, most scientific data just isn't time sensitive in this way.

Neither was this.

WillyGilligan · Post by **WillyGilligan** » Wed Nov 03, 2004 8:30 pm

Well, from the "life and death" angle, it could be seen as time-sensitive, but I'm leaning towards 32 on this one. It's more time-senistive from the "elections are around the corner" angle.

Marius · Post by **Marius** » Wed Nov 03, 2004 8:41 pm

Oh, well that's okay then, isn't it? It's totally ethical to rush bad methods and data past peer review if you need to get it done in time to hurt a politician you don't like.

WillyGilligan · Post by **WillyGilligan** » Wed Nov 03, 2004 9:30 pm

Sorry, was I too vague?

I think it's unethical to do it for political reasons (which this sounds like). I put the first reason (lives are being lost, maybe this data will help, rah rah rah) as a possible reason for rushing this that doesn't have sinister motives. Because, you know, I like to give me fellow humans at least some doubt that they might not realize when they're being evil. It's how I get along so well with people.

Marius · Post by **Marius** » Wed Nov 03, 2004 9:38 pm

Sorry, was I too vague?

No, I left off the sarcasm tags.

3278 · Post by **3278** » Thu Nov 04, 2004 3:11 am

Rev wrote:The essence of science is the collection and analysis of data. That is what the authors attempted to do.

No. That's in fact my problem: I believe it is quite clear that's /not/ what they were trying to do. I believe the essence of statistical analysis is the /objective/ collection of data, not the collection of data in specific ways which will prove a point you wish to make.

Rev wrote:They collected data and analyzed it, pointing out that the data was damned hard to collect and thus imperfect.

They collected data imperfectly, and analyzed it in such a way as to match their intent.

Rev wrote:They attempted to validate thier data by checking some of it against death certificates.

Two death certificates per cluster. Out of 33 clusters. Confirmation was attempted at only 78 of 988 households; only 81 percent of those were confirmed. In total, 63 death certificates were collected to confirm 100,000 deaths.

Rev wrote:Baldly rejecting the only study of its kind is what is unscientific. You don't have to be convinced, but it clearly is useful data.

I don't feel it's particularly useful data at all, which is why I reject it; my rejection isn't "bald:" it's informed by the horrendous misuse of scientific research to further a political agendy.

Serious Paul · Post by **Serious Paul** » Wed Nov 24, 2004 2:22 pm

So we pretty much agree this study is crap?

lorg · Post by **lorg** » Wed Nov 24, 2004 2:26 pm

It has been 20+ days since the last post. I heardly even remembered the thread ...

The study might have been somewhat lacking in certain areas. But the Lancet still remains a scientific journal in good standing.

Serious Paul · Post by **Serious Paul** » Wed Nov 24, 2004 2:31 pm

Yeah I'm not questioning the magazine, just the study.

3278 · Post by **3278** » Wed Nov 24, 2004 2:34 pm

lorg wrote:The study might have been somewhat lacking in certain areas.

I don't think "somewhat lacking" really covers it. That's like saying the evidence for creationism is "somewhat lacking," or the evidence for WMDs in Iraq was "somewhat lacking." The study was misleading and improperly conducted, and its results are thus scientifically worthless.

lorg wrote:But the Lancet still remains a scientific journal in good standing.

Perhaps in your mind. I would strongly question the ethics of any publication which fast-tracks questionable - and, in many ways, outright deceitful - research. One bad study doesn't ruin the reputation or standing of an otherwise ethical journal, but the timing and tactics of this study's release do call into question their methods and intent. Furthermore, the Lancet's standing - and I'm not certain if this was your implication or not - has no bearing on the worth of the study; "good standing" or otherwise, the study is, indeed, crap.

Salvation122 · Post by **Salvation122** » Sat Dec 04, 2004 2:21 am

Someone check my math.

Of the seven percent of deaths they attempted to confirm, eighty-one percent did so. That makes a little over 5% total. Five percent of the 142 deaths the survey describes means that they based a finding of an additional hundred-thousand dead Iraqis off of eight confirmed deaths.

So: 7 * .81 = 5.67; .0567 * 142 = 8.
That can't possibly be right, but I'd appreciate it if someone would tell me where I went wrong.

MooCow · Post by **MooCow** » Sat Dec 04, 2004 2:29 am

The BBC says the Lancet (a medical journal) says 100k extra deaths in Iraq vs before the war.

Yeah, I heard about this on Michael Medved's show. He's a conservative pundit, so of course he'll disagree and everything he says must be taken witha grain of salt. However, according to his expert, the study is hopelessly flawed. Primary reason: They claimed the death rate prior to the invasion was lower then the US death rate.

Now call me crazy, but I find that notion to be positively ridiculous. Iraq couldn't possibly have a lower accidental death rate lower then the US. Just not possible.

Secondary reason: Osama Bin Laden doesn't even claim the US is responsible for 100k deaths. The number he's been using is somewhere around 15k.

mrmooky · Post by **mrmooky** » Sat Dec 04, 2004 11:50 am

I'm not going to weigh in much on this one, because important as the issue is, it's fucking impossible to get relaible stats of death rates either under a murderous dictator *or* during a war, and so even the best estimate is unlikely to be close to the real figure. But I will say this:

MooCow wrote:Iraq couldn't possibly have a lower accidental death rate lower then the US. Just not possible.

It's possible, if very unlikely. It all depends on how many Americans kill themselves and make it look like an accident.

3278 · Post by **3278** » Sun Dec 05, 2004 5:00 pm

MooCow wrote:However, according to his expert, the study is hopelessly flawed. Primary reason: They claimed the death rate prior to the invasion was lower then the US death rate.

Death rate, or accidental death rate? Below, you're talking about one, and above, you're non-specific. Either way, the dubious nature of a thing is no reason to disbelieve it; it is, however, a fine reason to doubt it and verify it.

MooCow wrote:Now call me crazy, but I find that notion to be positively ridiculous. Iraq couldn't possibly have a lower accidental death rate lower then the US. Just not possible.

I don't find it impossible at all. When discussing accidental - non-violent - deaths, I don't find any reason to doubt it. There are many more available ways to die accidentally in the US than in Iraq. Again, though, speculation doesn't aid in comprehension.

MooCow wrote:Secondary reason: Osama Bin Laden doesn't even claim the US is responsible for 100k deaths. The number he's been using is somewhere around 15k.

As has been stated several times in this thread, and in the original study, the 15k number is for directly invasion-related deaths; the 100k figure is for indirect deaths, caused by food shortages, invasion-related illness, and so on.

Look, the study is crap, but there are good and supported reasons to believe it's crap. There's no need to break out poor or unsupported reasons for its utter crappiness.

MooCow · Post by **MooCow** » Sun Dec 05, 2004 5:34 pm

There's no need to break out poor or unsupported reasons for its utter crappiness.

Of course there is.