100,000 Dead In Iraq

The startling news that some 100,000 people, nearly all of them Iraqi civilians, have been killed in Iraq since the U.S. invasion in March 2003 should give Americans pause when considering the costs of the war. Until now, the official line from Washington has been that "we don't count" Iraqis killed, only American soldiers killed. That total, more than 1,000, is bad enough. With this new scientific estimate, the consequences of this misbegotten war should be front-and-center in the coming days.

What is likely to cause considerable rancor in this new estimate is that most of the deaths reported were from U.S. bombing — nearly 80 percent — and not as a result of the insurgency. This scale of destruction is on par with the war in Vietnam.

The estimate, to be published next week in The Lancet, a leading British medical journal, comes from a distinguished group of social and medical scientists at Johns Hopkins University, headed by Dr. Les Roberts of the Bloomberg School of Public Health. The team also included researchers at Columbia University and the College of Medicine at Al-Mustansiriya University in Baghdad. They went house-to-house in 33 neighborhoods that reflect Iraq society as a whole and interviewed residents about deaths in their households since the U.S. invaded. The death rate, they found, averaged about 300 percent higher than normal, attributable to the war's violence.

"The risk of death was estimated to be 2-5 fold higher after the invasion when compared with the pre-invasion period," notes the story in The Lancet. Extrapolating from this data to the nation as a whole, that translates into a minimum of 100,000 war dead among civilians, the researchers said.

As reported in Friday’s newspapers, the estimate is being treated with considerable skepticism. The motives of the editors of The Lancet were questioned in The New York Times story (tucked away in a single column on page A8), because the study was released in a special Web version before The Lancet’s usual publication date. It is as if the Times were implying that such major news should have been held until after the election. Over at the Washington Post, a researcher for Human Rights Watch criticized the study's method by alleging that the sample was too small. HRW and some other groups have tried to count casualties by using documents, such as press reports. Those estimates have been around 17,000 Iraqi deaths.

But the estimate by the Hopkins team is sound in terms of how the data has been gathered and what it says about the casualty rate. A sample of 988 households with more than 7,800 people, in a country of 25 million, is a sizable sample. By comparison, pollsters in this country, using similar techniques of sampling (so the people interviewed in aggregate represent the demographics of the country as a whole), consider a sample of 1,500 people in a country of 280 million to be adequate for extrapolation and reliable results. Wherever possible, too, the researchers verified claims of fatalities with documents. A larger sample would be worthwhile, and as in any important empirical research, it would be useful to repeat the data collection to compare results. But the method is sound.

As a result, the estimates are likely to be quite a bit more accurate than the clumsy attempt to count through press reports, which is partial and not a representative sample. Indeed, the study directors believe the estimate of 100,000 deaths may be conservative. (Fallujah, for example, was not counted due to the extreme level of violence in that city.)

"There is a real necessity for accurate monitoring of civilian deaths during combat situations," study co-author Gilbert Burnham, MD, said in a statement. "Otherwise it is impossible to know the extent of the problems civilians may be facing or how to protect them."