October 12, 2006
From the desk of Jane Galt:It's possible that the Lancet study is right. Hell, it's possible that hte upper bound is right, and 4% of the population of Iraq has been killed in the past 3 years. I just don't think it's likely.
Any single study can be wrong. Especially a survey where, as the authors testify, the survey takers are afraid of getting shot. Obviously, the fact that they are afraid of getting shot testifies that all is not well in Iraq. But when you have a survey like this, it makes sense to check it by comparing it to other, relatively well known figures. How does the alleged mortality check against other civil wars? Against civilian deaths in other wars? Against military deaths in other wars?
Another way to check is against the passive statistics, like death certificates. Daniel Davies says these are wrong, because hospitals don't send them in to the central government. Perhaps they don't. But we can also check the figures for Baghdad, which has, everyone agrees, about a quarter of the population of Iraq, and also the worst violence. It is also where the central government is, and my understanding is that the morgue, at the medico-legal institute, is the one that issues and collects the death certificates. Undoubtedly some deaths in Baghdad are missed, but those statistics should be (for Iraq) pretty accurate. They are unlikely to be off by a factor of four or more.
In August, according to UNAMI, there were 73 violent deaths per day in Baghdad. In July, there were 93. Extrapolating this evenly across the remaining 75% of the population would give us approximately 125,000 deaths per year in the entire country, which is still below the lower bound of the Lancet study. But we also know that the other regions of Iraq aren't equally violent. The 10% of the population that is Kurdish lives in a basically stable area. So it's really 65% of the country, or another 16.5 million people that would have a similar violence rate. That gives me roughly 110,000 total. However, we also know that the country is, on average, safer than Baghdad, though some places, like Fallujah may be more dangerous. So that number needs to be revised downwards substantially. How substantially? No idea. Finally, we also know that the violence against civilians has been getting worse; even if the collection of death certificates is shoddy, the direction of the trend is clear. So earlier years would have fewer than 116,000 deaths even if the rest of the country were just as violent (on average) as Baghdad. Call it an upper bound of 300,000, being very charitable. That's a high number. But that's the top of the range, not the most likely number. What is the most likely number? No idea, and I'm not going to join the Lancet doctors in giving a false impression of precision. But it's less than a third of the Lancet study's upper bound.
On the other hand, to be fair, the death certificates undoubtedly miss a lot of people. After all, even in the US, we miss a few people--a fair number of missing persons probably fell into the river, got eaten by a bear or killed by a serial killer, or whatever. So we have to adjust that number upwards again by some unknown quantity. Anyone got any idea how easy it is to hide a body in Baghdad?
The upshot is that I have no idea whether the Lancet study is right or wrong. And neither does anyone else, including the authors. I do know that it implies an improbably large multiple over other known statistics, and that its error band is bigger than the lower bound, which always makes me feel kinda uncomfortable.
That doesn't mean Iraq was right, or wrong. All it means is that we're better off relying on qualitative data, not quantitative data, because the quantitative data is highly questionable. The emotional reaction of conservatives saying "That can't be right" is silly--and so is the sudden emotional attachement of liberals to the figure of 600K.
Posted by Jane Galt at October 12, 2006 4:07 PM | TrackBack | Technorati inbound linksWhy would anybody pay attention to anything Daniel Davies has to say? He is notoriously incapable of rational argument and personally offensive. And since he is fine with dishing out ad hominem, he should be equally fine to be on the receiving end.
Jack, you should pay attention to Daniel Davies because what he's saying is more important than his bad manners. The U.S. invasion has made life in Iraq worse (a lot worse), not better. That Davies might be a wanker doesn't change that.
All you are saying is that you find studies like this useless while the experts that work in the field don't. It is, of course, possible all the experts are wrong and the data they produce is rubbish.
The U.S. invasion has made life in Iraq worse (a lot worse), not better.
Let's assume, for the sake of argument, that that is true. It does not follow that every study which shows live has gotten worse in Iraq is correct.
Iraqis may well be worse off. The death rate has certainly increased from the immediate prewar period. But the methodology of the Lancet study is still *very* highly suspect. If you wanted to design a study for the specific purpose of making sure that deaths were overreported, it would look a lot like this.
All you are saying is that you find studies like this useless while the experts that work in the field don't.
The experts that work in this field have all gotten much lower death counts.
The U.S. invasion has made life in Iraq worse (a lot worse), not better.
This is not at all clear.
I'll advance a principle: When things are bad and not expected to improve, people will try to leave if they can. This, I think, is a sensible principle; as evidence I can cite various empirical data such as the Berlin Wall, Vietnamese boat people, and individuals floating from Cuba on innertubes.
Given this principle, one would expect that the net migration to Iraq would be strongly negative; that is, there would be many refugees and emigrants, and few immigrants. But the net migration to Iraq postwar is (estimated to be) -- zero.
The Iraqis aren't leaving.
Perhaps the poor, ignorant fools don't know just how bad things are. After all, they only live there, and probably don't read Crooked Timber or other leftist blogs on a regular basis.
Finally, I don't know which experts are working in which fields to generate these numbers, but I do know that nobody's working in the field in which 500,000 bodies are buried. That's something you'd notice, if only by the smell. Half a million bodies aren't easily hidden. Where are they?
(Yes, I see the possibility of a clever joke juxtaposing missing bodies and missing WMD. I leave it to the reader as an exercise).
But the net migration to Iraq postwar is (estimated to be) -- zero.
Reference ????
Reuters has a report that is only 12-15 old here where the Minister of Immigration and Displacement claims:
300,000 have migrated within Iraq "to escape violence", and
890,000 have moved to Jordan, Iran and Syria
Beg pardon? a 12-15 year old report on immigration has just *what* point here? Wasn't that pretty much between Gulf Wars I and II? So they were fleeing Saddam? Are Iraqis *still* fleeing Saddam?
The report concerns people fleeing since since Saddam's fall. It isn't 12-15 years old.
Jane says: "we're better off relying on qualitative data, not quantitative data, because the quantitative data is highly questionable."
I don't get this at all. It matters how many people are dying. If 200,000 people died, that is about twice as bad as if only 100,000 died (ceterus paribus). Such numbers are hard to estimate, but I see no reason why survey sampling shouldn't be part of the attempt. I'm not confident in evaluating how well this particular study was done without knowing a lot more about Iraq than I do.
And how do you know that the rest of Iraq is less violent than Baghdad? I'm not sure anybody really knows what's going on out there in some parts of Iraq.
And how do you know that the rest of Iraq is less violent than Baghdad? I'm not sure anybody really knows what's going on out there in some parts of Iraq.
Two ways come readily to mind:
1. All of the bombs and bullets that the media finds to bleed-leade are predominantly in the Sunni triangle, including Baghdad. Anyone can confirm this by simply taking note of where today's "IRAQ IS FAILING ARRRGHH!" stories are reporting from.
2. People HAVE been elsewhere in Iraq, and found conditions to be roughly as was described. Michael Totten, to name one. Of course, since stories from these regions generally do not bleed-leade, you frequently have to look outside of mainstream news outlets to find them...but being ignorant of these is not a good basis for claiming that the whole country is probably violent.
Any single study can be wrong
this is not a single study; it is a follow up of the 2004 study.
Undoubtedly some deaths in Baghdad are missed, but those statistics should be (for Iraq) pretty accurate. They are unlikely to be off by a factor of four or more
You have absolutely no basis on which to make this assumption.
The report I cited from Reuters is 12-15 hours old. Typo; sorry.
I haven't seen anyone discussing the number of wounded. If the wounded/dead ratio for these excess deaths is anything like previous wars (1.7 for WW1 rising to 6 for coalition forces in Iraq) then there should be a very noticable chunk taken out of the work force.
Including the dead, this would range from 6.75% at WW1 rates to 17.5% for OIF rates. In urban areas the proportions will be even higher.
Again, this doesn't pass the smell test. If a fifth of the population is either dead or confined to their beds it shouldn't take a survey to notice.
D9 - the situation with wounded is not comparable to the examples you mention. For one thing, a major reason for the 6:1 ratio Coalition troops have is better body armour - not something many Iraqi civilians will have. Second, a lot of these deaths are not taking place in a combat situation eg if a group of unarmed men are kidnapped at gun point, then taken away and shot, the death rate's going to be 100%.
I agree with Galt's conclusion that we don't know.
A fairly low daily rate of anything can build up impressive totals if carried on long enough.
Compare Iraq now to the Spanish Civil War: same populations, same length of civil war, same disputes (sectarian, rich v. poor), same intervention by foreign armies.
The historians accept a figure of 1 million dead in Spain. That's a suspiciously round figure, but it's what we've got.
Gen. Casey's figure of not much more than 50K is even more absurd than the most partisan critique of Lancet's figure: There have been around 15K deaths in uniform (a number not in dispute), and to get past 50K you have to assume only 2 civilian deaths per uniformed death.
"In August, according to UNAMI, there were 73 violent deaths per day in Baghdad. In July, there were 93."
No, these numbers are for deaths reported by hospitals and the Medico-Legal Institute. They do not encompass all deaths by any stretch of the imagination.
Iraqis may well be worse off. The death rate has certainly increased from the immediate prewar period. But the methodology of the Lancet study is still *very* highly suspect. If you wanted to design a study for the specific purpose of making sure that deaths were overreported, it would look a lot like this.
What are you talking about? Why does this study overreport? If anything the interviewers probably avoided the more dangerous areas, thus one could argue that it underreports. What reason do you have to believe this?
The experts that work in this field have all gotten much lower death counts.
To repeat. What are you talking about? I know of no other experts in the field doing any death counts. The media is not an expert at counting deaths. I would label government agencies, in general, as experts, but I think we can all admit that Iraq hardly has many fully or even minimally functioning bureaucracies that are reliable. The U.S. has refused to count civilian deaths. Are there any reports available based on sound evidence? I know of no others.
Bottom line is there is no good argument beyond fraud for this study being wrong. All Galt does is posit her own flawed theory as to why her numbers are superior to the Lancet study. She bases her analysis on a single number which, as Njorl points out, is undoubtedly missing a large number of deaths.
I would be curious how one obtains a death certificate in Iraq. Perhaps I missed the explanation in the actual study.
Jane,
The UNAMI figures for deaths in Baghdad are incomplete and not accurate.
First they are contradictory. For July 2006 the figures show 2800 killed, but only 1800 death certificates issued. How did they estimate the number killed then? There is no description of their methodology or reliability. So how can you use them as a test.
Second even in the peaceful period prior to the invasion the number of death certificates issued by the Iraqi ministry of health was far less than any plausible death rate, probably by a factor of 3.
So when you ask: "Undoubtedly some deaths in Baghdad are missed, but those statistics should be (for Iraq) pretty accurate. They are unlikely to be off by a factor of four or more."
The answer could very well be yes; especially because of the war and because the hospitals and Health Ministry are controlled by the Saadrists.
This from the companion piece to the report:
"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."
You are using the official Government figures as a benchmark to question the Hopkin's estimates. But you have no idea how accurate your benchmark really is. The Hopkins study on the other hand has explained their methodology, data sources and collection, and the resulting estimate of error.
It would be funny if the argument weren't about dead people. But as it stands, it is an argument about just how bad the US has made Iraq. Anyone who wishes to quibble about that fact would likely find a ready trade with an Iraqi, sight unseen, for lifestyle and possessions. As it stands, maybe I don't get out enough, but I don't fear drillbit-bearing nuts, nor feel the need to barricade my neighborhood at night. Your milage may vary.
Comments are Closed.