Psyche, Science, and Society

More evidence that Iraqi mortality figures are bogus

October 20th, 2006

According to this report, Le Monde is reporting that the Iraqi government has ordered the Ministry of Health to not release further figures:

Iraq ‘hiding true casualty figures’

THE Iraqi Government has told medical authorities not to reveal to the UN the true extent of civilian casualties in the country’s conflict, French newspaper Le Monde said today.

The daily quoted a telegram sent by the head of the UN mission in Iraq, Ashraf Qazi, to headquarters in New York, in which he said: “This development risks damaging the capacity of the UN’s Assistance Mission to report the number of civilians killed or injured.”
Since July 2005 the UN has used data provided by Baghdad’s Forensic Institute and the Iraqi health ministry to form an estimate. The estimate “was certainly imperfect but an indicator nonetheless of the growing number of civilian victims”, the telegram said.

The latest report said that 3590 civilians died a violent death in July and 3009 in August, figures which it said were “unprecedented”.

But the telegram quoted by Le Monde said that on September 21, one day after publication of the report, Iraqi Prime Minister Nuri al-Maliki wrote to the health ministry with instructions not to disclose more figures.

In an investigation published in British medical weekly The Lancet earlier this month, US and Iraq specialists estimated that more than 600,000 civilians died a violent death between March 2003 and July 2006.

Of course, this wouldn’t automatically invalidate earlier figures. But reports have frequently referredd to political manipulation of these figures, as in this May 25, 2006 article by the Christian Science Monitor’s dan Murphy:

What until late 2005 used to be a routine reporting visit to the morgue - the best way to figure out the level of violent death in Baghdad - has now become a minefield of frustration and implied threat.

After a person picks up a permission letter from the bureaucrat at the ministry public affairs office, he presents it to the morgue director’s secretary. But now, she immediately directs the person to go to the compound’s security office. “The director is in, but you have to talk to Major Kassem first.” Why? “You just have to.”

Going back outside, Major Kassem is tracked down inside a small air-conditioned trailer. He has a broad smile, and no uniform or badge.
Asked what his role is he explains: “I work for the Interior Ministry, but I’ve been assigned here to help coordinate the FPS.”

He then explains that there’s no need to speak to ministry officials, that he will provide all available information. He gives the monthly murder totals from the beginning of the year, though UN officials and Sunni Arab politicians say the ministry has taken to under-reporting the numbers, under pressure from Shiite militias.”

I, further, see no reason to believe that, under these conditions, the correct nation-wide data is even being carefully collected and sent into Baghdad. As one who has worked extensively with these government datasets in this country [we call them "administrative data sets"], I know how inaccurate and incomplete they can be, even under the best circumstances. Collecting and transmitting data is usually the last priority in human service settings. In times of stress, such as civil war, it is virtually unimaginable to believe that these data systems would work. Thus, I don’t find a large discrepancy between the Iraq Body Count numbers and other mortality figures to be at all surprising.

This applies to all the data systems in Iraq. Without extensive evidence that they are in fact working decently, I see no reason to believe that they are. Thus, discrepancies between survey-based mortality figures and official morgue and hospital figures are not only not surprising, but to be expected. This is a large part of why I think the IBC critique of the Lancet study is largely bunk. Its arguments rely to a gerat degree on the ptoper functioning of the countries information infrastructure. When electricity is down to under 3 hours a day in Baghdad, information systems are certainly not a priority in the best of times. These are near the worst of times.

If IBC has good, credible evidence that these information systems are functioning, they should present it. Otherwise, most of thie arguments can be reasonably interpreted as evidence that these systems are not, in fact functioning.

Of course, the weakness of IBC’s critique does not mean that the Lancet mortality study is necessarily correct. I’m still waiting for the main street bias issue to be sorted out. that seems to be to be a potentially greater threat to the validity. [Please note, I said "potentially greater threat to the validity." I am not endorsing this critique, only saying it should be sorted out.]

Entry Filed under: Iraq, Middle East, Mortality, Public Health, Research Methods, Science, Social Issues, War and Peace