J. Bohannon/Science

Uncounted dead? The toll of war victims, like these remains in Bosnia, may be grossly underestimated according to a controversial study.

War Deaths Grossly Underreported--Study

John is a Science contributing correspondent.

The 20th century was a bloody one by any account. But it may have been even worse. According to a study of war-related deaths reported by siblings in 13 countries, the number of fatalities since 1955 is three times higher than most experts think.

Getting an accurate figure on war deaths has never been easy. But over the past half-century, researchers have developed a tool kit for estimating the toll, including household surveys, tallies of media reports of violence, and comparisons of census data before and after conflicts. By pooling such data, scientists conclude that the number of violent deaths due to armed conflicts has dropped in the 50 years since the Korean war.

A team led by Ziad Obermeyer, a public health researcher at the Institute for Health Metrics and Evaluation in Seattle, Washington, used an alternative source to estimate war deaths: published surveys that asked people about their siblings. The rationale is that the most accurate source of data about whether someone died in a war is their brothers and sisters who survived. (They did not account for the chance that all siblings in a family are killed.)

The team focused on World Health Organization data from the 13 countries in which people reported the highest number of sibling deaths due to war. These include some of the deadliest conflicts in the latter half of the 20th century, such as those in Vietnam, Zimbabwe, and Bosnia. Of the 43,874 deaths noted by siblings in the surveys, 917 were reported to be the result of war injuries. Using census data for each country, the team extrapolated a figure of 5.4 million war-related deaths in these 13 countries between 1955 and 2002.

That figure is three times higher than estimates compiled a few years ago for global war-related deaths over the same period, Obermeyer's team reports today in the British Medical Journal. In Bangladesh, for example, the researchers estimate that 269,000 people died violently in the 1971 war for independence, compared to previous estimates of 58,000. And comparing their estimate with previous estimates decade by decade, the researchers conclude that "there is no evidence" for a decreasing trend in war-related deaths since the 1950s.

Many experts are skeptical. Nils Petter Gleditsch, the editor of the Journal of Peace Research in Oslo, Norway, a co-author on an earlier analysis, says the team's analysis of Bosnian fatalities "is most concerning" because that conflict has been so thoroughly studied. The discrepancy with Bosnian studies "undermines the credibility of their entire study," he says.

Posted in Biology