Friday, October 13, 2006

600,000 dead in Iraq is a reasonable estimate

The Lancet study (PDF) that estimated deaths in Iraq at 600,000 is coming under considerable critique. However, all these critiques assume a priori that the number is unreasonable as an apriori. But if the number is actually quite reasonable, then the numerical methodologic critiques are irrelevant.

And in fact the estimate of 600,000 dead is actually quite reasonable. Juan Cole has a very lengthy post that lays it out on the line, including more detail about the methodology than most of the critics provide. He also notes that muslim burial practices sugggest that the estimates are, if anything, conservative. A lengthy excerpt:

I follow the violence in Iraq carefully and daily, and I find the results plausible.

First of all, Iraqi Muslims don't believe in embalming or open casket funerals days later. They believe that the body should be buried by sunset the day of death, in a plain wooden box. So there is no reason to expect them to take the body to the morgue. Although there are benefits to registering with the government for a death certificate, there are also disadvantages. Many families who have had someone killed believe that the government or the Americans were involved, and will have wanted to avoid drawing further attention to themselves by filling out state forms and giving their address.
Not to mention that for substantial periods of time since 2003 it has been dangerous in about half the country just to move around, much less to move around with dead bodies.

There is heavy fighting almost every day at Ramadi in al-Anbar province, among guerrillas, townspeople, tribes, Marines and Iraqi police and army. We almost never get a report of these skirmishes and we almost never are told about Iraqi casualties in Ramadi. Does 1 person a day die there of political violence? Is it more like 4? 10? What about Samarra? Tikrit? No one is saying. Since they aren't, on what basis do we say that the Lancet study is impossible?

There are about 90 major towns and cities in Iraq. If we subtract Baghdad, where about 100 a day die, that still leaves 89. If an average of 4 or so are killed in each of those 89, then the study's results are correct. Of course, 4 is an average. Cities in areas dominated by the guerrilla movement will have more than 4 killed daily, sleepy Kurdish towns will have no one killed.

If 470 were dying every day, what would that look like?

West Baghdad is roughly 10% of the Iraqi population. It is certainly generating 47 dead a day. Same for Sadr City, same proportions. So to argue against the study you have to assume that Baquba, Hilla, Kirkuk, Kut, Amara, Samarra, etc., are not producing deaths at the same rate as the two halves of Baghad. But it is perfectly plausible that rough places like Kut and Amara, with their displaced Marsh Arab populations, are keeping up their end. Four dead a day in Kut or Amara at the hands of militiamen or politicized tribesmen? Is that really hard to believe? Have you been reading this column the last three years?

Or let's take the city of Basra, which is also roughly 10% of the Iraqi population. Proportionally speaking, you'd expect on the order of 40 persons to be dying of political violence there every day. We don't see 40 persons from Basra reported dead in the wire services on a daily basis.

But last May, the government authorities in Basra came out and admitted that security had collapsed in the city and that for the previous month, one person had been assassinated every hour. Now, that is 24 dead a day, just from political assassination. Apparently these persons were being killed in faction fighting among Shiite militias and Marsh Arab tribes. We never saw any of those 24 deaths a day reported in the Western press. And we never see any deaths from Basra reported in the wire services on a daily basis even now. Has security improved since May? No one seems even to be reporting on it, yes or no.

So if 24 Iraqis can be shot down every day in Basra for a month (or for many months?) and no one notices, the Lancet results are perfectly plausible.

Also, to look at the issue another way, consider that almost 3,000 US sodiers have been killed in the war thus far, and that 14,000 Iraqi security forces have also been killed (the latter a conservative estimate based on published news reports only, source). Is it really beyond the realm of possibility that for every US soldier verified killed by the DoD or Iraqi security person reported killed by the media, that an additional 34 civilians have also been executed by militias or been killed by foreign jihadis? Because 35x17k= 600k right there.

The question is why the rush to deny the number. Does the argument for being in Iraq change if its 600,000 dead? If so, where is the threshold? Keep in mind that just a few months ago the same people were outraged at the suggestion that there might be 100,000 dead. Is that number still beyond the realm of possibility, too, in their eyes? And if not - if 100,000 is actually reasonable now whereas 600,000 is not, then when does 600,000 become reasonable?

I think that its telling of insecurity by the pro-war right that they have seized with such ferocity on the report, because they are essentially arguing a detail rather than a strategy. John (not Juan) Cole explores this further.


  1. Zeyad is actually a former dentist student, not a doctor, and the study asked for and received death certificates in 92% of cases.

  2. On the other hand, Zeyad of "Healing Iraq" thinks the numbers a bit high, but in the right ballpark.

    His own guess is about half the study's, and he suggested (as have others) that the study didn't take into account the fact that violence doesn't scale linearly with population density - and oversampled in high population areas.

  3. Guys, you expect a Islam Supremacist like Aziz to be reasonable and logical? His God forbids reason!

  4. It's funny how Iraq Body Count has become so important for some people...

    Crooked Timber have the Lancet report pretty much covered.