May 16, 2005

Lancet Study considerations

Silly hypothetical question: Is a person responsible for the death of someone he starves to death?

I have a few more things to say about the Lancet Study in light of Joe Malchow's offhand dismissal of it. Malchow seems to have concluded that the study is somehow worthless because, "While the Lancet Study does take the difference in pre- and post-war deaths, it is still wrong in terms of magnitude. It still counts deaths not related to war violence."

This is a) a misleading characterization of the study, and b) making what is in principle a completely absurd, inhumane argument about how a nation that beings a war and invades a country should be held responsible for the civilian deaths it causes.

(a) What the Lancet Study demonstrates through its well-established scientific methodology is an estimate of the deaths due to the US invasion of Iraq, whatever form those deaths took. That means deaths related to war violence but not necessairly directly related. Just thinking through the study's methodology, explained in an earlier post, for a second will make this clear. It does not just measure the number of Iraqi civilians who die from being shot, bombed, or otherwise destroyed immediately by US forces. It measures all kinds of deaths from disease, malnutrition, etc., that are the inevitable side-effects of war. It's a logistical reality that many cons like Rumsfeld would not even care to include under "collateral damage," since those who die under collateral damage are those who die only directly from being accidentally bombed or shot, presumably.

Just image if any region in the US were devastated as parts of Iraq have been. As is, without such a devastation, there are large numbers of people who are in need of healthcare, food, shelter, etc, and most of whom manage to scrape by thanks to many services that are provided to such people. Now take the logistical damage that would happen: cut off some water, power, overcrowd the hospitals and soup kitchens and shelters, and think about what would happen. Now imagine this scenario in a nation like Iraq, with a fraction of the resources the US has.

The Lancet Study has a few problems with it, mainly problems of scale or sample size. It happened to include Fallujah in its sample, where the damage wreaked by a prolonged and heavy assault probably skewed the results slightly. But this study still deserves to be taken seriously, and provides a good estimate of the real magnitude of civilian death in Iraq. Having looked critically at the major studies of the death toll, I believe a good estimate of the total deaths caused by the invasion is in the range of 50,000 to 100,000, with 75,000 as the most likely, and, I think, conservative number.

(b) I just don't get it: people seem to be criticizing the Lancet Study for its very fundamental, and fundamentally humane, purpose: to figure out how many people the war has actually killed. Do these critics believe that if one person starves another, no one is responsible because no shots dealt the deathblow? Would they also argue that the embargo on Iraq was not resonsible for any deaths? (Try on the order of hundreds of thousands after looking at some evidence here and here.)

1 comment:

  1. Joe Malchow's priorities:

    --- You wrote:
    Date: 16 May 2005 13:58:57 EDT
    From: Christopher J. Bateman
    Subject: Re: another post
    To: Joseph I. Malchow

    Chris...

    The reason I am being so passive about the Lancet study is that it is old news. It was released before the election in a [poor] attempt to influence it, and the battle over its statistical methods and conclusions has been fought many times over. It is old news. I would be inclined to react in the same way if I had made mention of Dan Rather's having used fake documents on a 60 Minutes report, and a critic in turn had posted: "How dare he make such an accusation! Let's debate this!"

    Regards,
    -joe
    --- end of quote ---

    Great. A possible 25,000 to 75,000 unacknowledged deaths: never worth an update.

    ReplyDelete