Uncomfortable statistics but sound method

Most anyone who will read this already has heard of the recent study on Iraqi casualties. However, I originally started writing this thing for those who were out of touch with the news. Further, there is an interesting postscript to the study.

 

A group of researchers from Johns Hopkins Bloomberg School of Public Health completed their study of Iraqi casualties and released their findings on October 10 (published in Britain’s medical journal, The Lancet). Specifically, they studied the number of civilian deaths since the invasion began. Their conclusion: 601,027 died from violence (the majority from gunshot wounds; 31% ascribed to the coalition), while an additional 53,000 or so people died “by the effects of the war” due to “worsening of health status and access to health care.” To help put that in perspective: 15,000 violent deaths a month, 2.5% of the entire population of the country.

 

In order to compile these figures, they used a technique called clustering. (Thanks, Economist.) This involved the survey of 1,849 Iraqi families in 47 neighborhoods across 18 regions of Iraq. They sampled based on population size, not level of violence in each area. They asked the families about deaths and causes of such. This was not a phone interview: they visited each house, and asked if the families had certification of the deaths they reported. 92% of all the reported deaths were accompanied by certification. They then compared these figures with the mortality rate pre-invasion. By extrapolating from these figures, they concluded that over 600,000 died who would not have otherwise since the invasion began.

 

Mr Bush, General Casey, Iraqi officials—basically everyone instantly rejected the study. To quote Mr Bush, “the methodology was pretty well discredited… A lot of innocent people have lost their life—600,000, or whatever they guessed at, is just—it’s not credible.”

 

Discredited, guessing, not credible. Interesting. According to CNN, “Professionals familiar with such research told CNN that the survey’s methodology is sound” (http://www.cnn.com/2006/WORLD/meast/10/11/iraq.deaths/index.html). NPR interviewed a medical reporter and the head of the study, Gilbert Burnham. The former claimed that their methods were a “standard epidemiological tool” for natural disasters, war, etc. The head of the study, when asked about Mr Bush’s response, said that theirs were “Very standard off the shelf methods… the methods that the US funds all over the world and is responsible for much of the very excellent health data of countries of the world that the US has supported in its collection.” When pressed for examples of other studies that used the same method, Mr Burnham responded that the most well known are in the Congo (one of the Iraqi researchers was party to this study) and Darfur (including some Hopkins former students and the same methodology), the latter which is “widely quoted by US government and others in designing and encouraging a response to the serious situation in Darfur.”

 

In short, either the US needs to immediately distance itself from these other studies and whatever results came from such, or they need to give this study the same attention as the others. Either way, the government looks pretty stupid paying tax dollars to people for research which they maintain is “not credible.”

 

-W.