Lecture ~ Counting the Dead in Iraq



It's no wonder there was an outcry when Gilbert Burnham's group released its report on mortality in Iraq. The numbers of civilian deaths so overwhelmed body counts calculated by other groups that many were stunned or disbelieving, and Burnham earned the enmity of some U.S. and Iraqi government officials.

Burnham's public health team looked at pre- and post-invasion deaths. The 2004 study showed that the mortality rate among Iraqis before the invasion was 2%, and after, 7.9%. The 2006 survey, which polled more households and covered greater territory, was more devastating: In the three years since the invasion, crude mortality rose to 13.2 per 1000 people per year. The leading cause was gunshot wounds and deaths from car bombs. The majority of victims of violence were men, 15-45 years old, and children also died in great numbers. By the end of the analysis period, crude mortality rates approached 17 deaths per 1000 per year.

The most disturbing statistic is the report's estimate that there have been 654,000 excess deaths since the invasion of March 2003 -- 600,000 from violent causes. Critics, who are legion, Burnham acknowledges, point fingers at his study's methodology, accusing his group of inaccurate and inadequate record-keeping, or skewing the numbers for political purposes.

Burnham notes that getting actual body counts in Iraq is literally impossible, since there is no working system for keeping accurate track of the dead in hospitals and mortuaries, and "numbers are highly susceptible to manipulation." The backbone of public health studies are surveys, in which geographic clusters are chosen, households counted and individuals interviewed. As the number of clusters increase, "precision improves and confidence intervals narrow." This enables measurements "accurate and precise enough to make the right decisions even though we will never have absolute, true numbers to two or three decimal points."

At great personal peril, Burnham's on-ground Iraqi surveyors went house to house in neighborhoods all over Iraq, asking for death certificates. The author of the report "hid out at a basement of a hotel, and finally got out on forged U.N. documents." The 2004 survey reached 7868 people, and the 2006 contacted 12,800 individuals. The sample size was large enough to support the team's grisly conclusions. Civilians are doing badly in this war, dying in far greater numbers than combatants. Burnham's hope is to use such data "to protect people wrapped up in conflict," since this "won't be the only one in the 21st century."