A new study published on bmj.com maintains that most estimates of global war deaths have been significantly underestimated – perhaps by as much as three times. Researchers have also been unable to find evidence to support claims that the number of war deaths has recently declined.
Estimates of violent war deaths are often gathered using survey-based techniques that are highly criticized for their possible biases and inaccuracies. Recently in Iraq, analysts used household deaths to estimate war mortality. However, there was some skepticism regarding the statistical validity of the results and there were fears that people were over-reporting deaths for political reasons. A second way to estimate deaths – one also used in Iraq – is from passive data collection from eyewitnesses and media reports. Of course, major biases affect these data, especially since eyewitnesses are unlikely to go to the most dangerous areas where the highest number of deaths occurs.
Ziad Obermeyer and colleagues from the Institute for Health Metrics and Evaluation in Seattle set out to overcome biases by creating a new way to estimate violent war deaths. Their method relies on data on siblings of respondents in large household surveys that are conducted during peacetime. The idea is that more accurate data on war deaths can be procured by comparing passive eyewitness and media accounts of violent war deaths in 13 countries over 50 years with data from the UN’s World Health Surveys collected in peacetime after conflicts. The researchers argue that this new technique not only reduces double counting and exaggeration of deaths by using sibling histories but also bypasses the measurement difficulties inherent to active combat.
In the 13 countries studied, the researchers estimate 5.4 million war-related deaths between 1955 and 2002, ranging from 7000 in the Democratic Republic of Congo to 3.8 million in Vietnam. Compared to previous reports, these estimates are about three times higher, on average. Specifically, previous estimates of violent deaths due to war between 1985 and 1994 were about 137,000 people per year, whereas the new estimates are about 378,000 people per year.
According to the authors, it is important to note that these new data contradict the common notion that violent war deaths have been declining since the middle of the 20th century. They also fail to support the argument that technological and strategic innovations have resulted in a minimization of civilian deaths in recent wars. The authors conclude by calling for a reevaluation of existing claims.
An accompanying editorial, written by Professor Richard Garfield (Columbia University), suggests that since these estimates still only focus on violent deaths, we are still underestimating the number of deaths due to conflict. He writes: “In the poorest countries, where most conflicts now occur, a rise in deaths from infectious diseases often dwarf the number of violent deaths during a conflict.”
“The importance of war as a public health problem and a social problem makes this imperative,” concludes Garfield.
Fifty years of violent war deaths from Vietnam to Bosnia: analysis of data from the world health survey programme
Ziad Obermeyer, Christopher J L Murray, Emmanuela Gakidou
BMJ (2008).
doi:10.1136/bmj.a137
Written by: Peter M Crosta