Wednesday, October 11, 2006

Iraqi civilian deaths: 655,000?

Oh, you know this is going to be controversial...

The John Hopkins Bloomberg School of Public Health did a study of the mortality rate among Iraqi civilians since 2003, comparing those figures to the mortality rate before the war. The idea was to determine how many people had died as a direct and indirect result of the war.

Up until now, civilian casualties have been measured by body count. The official number by the Bush administration is around 30,000 deaths. Other independant surveys put it at 44–49,000. The John Hopkins study puts the number of deaths since the war began at 655,000, or about 500 per day.

I found the news item here:

The study surveyed 1,850 families from 47 areas around the country, comprising some 12,800 people. They asked them about members of their families who had died. 629 family members had died since 2002. Of those, 87% had died since the war began. When you divide the 540-odd deaths by the 12,800 people over the time period investigated, you get a mortality rate or 13.3 people per 1,000. The mortality rate before the war was 5.5 per 1,000. Multiply the rate across the entire population of 25 million and you get 655,000.

Note: I take exception to the first line of the BBC article, which suggests that 655,000 people died as a direct result of the war. The mortality rate before the war would have resulted in a number of deaths in the 270,000 range for the same time period, so the comment that, "An estimated 655,000 Iraqis have died since 2003 who might still be alive but for the US-led invasion" is just flat wrong. The right number would be 385,000.

The reason for checking mortality rates is to uncover deaths due to things like disease, accidents, and preventable medical conditions that occurred due to a break down in the Iraqi infrastructure. However, according to the survey 601,000 of the estimated deaths — again, if the numbers bear out across the rest of the country — are due to violence.

When I see numbers like this, I immediately wonder about the methodology used. The researchers have suggested that they may be under counting deaths, because entire families could have been wiped out with no one to speak for them, and infant deaths could be under reported.

That having been said, how many of the deaths listed were for family members who were party of the insurgency, and thus not counted by the Pentagon among "civilian" deaths. The team said 80% of the family members had death certificates. Could any have been faked? Then there's the clusters they used for the survey: where were they, how were they derived, and how do they map to the Iraqi population as a whole? How do they map to the level of violence in an area? If there were 655,000 deaths, has anyone found physical evidence by researching grave sites? (This may be difficult, given the level of violence inflicted on Western civilian researchers over there.) The number of deaths directly attributed to violence seems awfully high (that's not to say it is in error, just that it sends up a flag).

The White House have come out against the study, saying the methodology has been "discredited". A similar study back in 2004 was said to be "discredited" by the White House when it stated 100,000 civilians had died since the war at that point. I'm not exactly sure how they can say this current study is discredited when the study isn't published until tomorrow, in the British medical journal Lancet. I would have hoped they'd at least wait until they saw the information before they said it was "discredited". And I'd like to see their reasoning why it is discredited, instead of just a flippant remark. I hope their is some coverage of this online, because I'd really like to see how it clears the peer review process.

This will be hotly debated. When the WMD dust had settled, the administration fell back on three key reasons for fighting in Iraq: Iraq had connections to Al Quaida, the U.S. is safer from terrorists due to the war in Iraq, and the people of Iraq are better off with Saddam Hussein deposed. The first statement was shot down when the 9/11 Commission's report was released, and the Commission's conclusion was reiterated last month. The second point was also shot down last month. If these numbers hold, then it could be argued that the people of Iraq are demonstrably worse off due to the invasion, killing the final justification for going to war.

The president's response was interesting. When asked about the report, he said this:
No, I don’t consider it a credible report, neither does General Casey and neither do Iraqi officials. I do know that a lot of innocent people have died and it troubles me and grieves me. And I applaud the Iraqis for their courage in the face of violence. I am, you know, amazed that this is a society which so wants to be free that they’re willing to — you know, that there’s a level of violence that they tolerate.

Missing from the President's remarks are the fact that — according to the associated press — some 890,000 Iraqis have moved to Jordan, Iran, and Syria due to the violence, and another 300,000 have fled to other parts of Iraq. According to a poll result obtained by the Washington Post, 71% of Iraqis want the U.S. out of Iraq. That doesn't sound like they are tolerating the violence to me.

Lost in this story will be any real analysis of the numbers. The administration and pro-conservative groups will stick with the "official" number of 30,000. Anti-administration and pro-liberal groups will stick with the distorted message that "655,000 Iraqis have died due to the war". I hope to find see an analysis at some point, but I'm not holding my breath.


JAM said...

Wow. You're posting them faster than I can read them.

Yeah, I smelled something fishy when the Prez came out discrediting the report so fast. Before the report is published. But I guess it's possible they had an advance copy.

I would guess the true number is way north of the official count and somewhere south of the 650K or so of this new report.

Is Lancet something a person can buy? I attended both ULM (formerly NLU) and La Tech, although both experiences were a while back. I wrote papers that used both JAMA and Lancet at both schools, so I know they usually get and archive them. It would be interesting to get and read the study, I don't know if mere mortals can get these magazines anywhere in public otherwise. I've never seen it at our public library down here, but it very well may be. I'll have to look.

Michael said...

The survey methodology has a huge margin of error. But even the lowest possible estimate provided by this study shows over 300,000 excess civilian deaths since the war. That's considerably higher than the previous high number, and obviously much higher than that presented by BushCo.

The CBC has an interesting analysis of the study. Obviously something like this is going to raise all sorts of methodological questions. But it's worth noting that the study was peer-reviewed before being published, which means that the methodology itself is sound. Where I'd have questions is in the actual data-gathering. Just how did the survey teams get access to a "random" sampling of Iraqi civilians, given the chaotic, violent nature of the country?