Documenting Iraqi Casualties

A new RAND report, An Argument for Documenting Casualties: Violence Against Iraqi Civilians 2006 by Katharine Hall and Dale Stahl is likely to reopen the debate on the quantification of civilian deaths in Iraq:

Protecting the civilian population is one of the central tenets of U.S. counterinsurgency doctrine. Until very recently, however, the U.S. military has not had a formal system for documenting the level of violence directed against Iraqi civilians. Therefore, other groups (such as nongovernmental organizations, the United Nations, and Iraqi ministries) have filled the vacuum in reporting, relying on media accounts, surveys, death certificates, and other open-source information to generate datasets of varying transparency and quality. The resulting statistics have generated widespread debate over sources, methods, and political biases. This study examines available open-source data on Iraqi civilian fatalities and assesses problems associated with previous collection and analysis efforts. The authors present a more robust RAND Corporation Iraqi civilian violence dataset from which they derive new observations about trends in targeting and weapons in 2006. RAND’s dataset reveals that the majority of attacks in the year 2006 against civilians were directed against individuals without any identifiable affiliation, and that most attacks were carried out using firearms (rather than via improvised explosive devices or suicide attacks). These findings lead to a proposed framework for future civilian fatality data-collection efforts in Iraq and beyond.

The other groups to which they refer of course include the infamous Lancet reports (latest version) and the Iraq Body Count website, both of which generated considerable controversy when published. If we recall, The Lancet put the approximate total number of Iraqi civilian casualties between March 2003 and June 2006 at 655,000 based on a hotly disputed cluster sample survey. The IBC total ranges from 75,000 to 82,000 over the period February 2003-October 2007, using data derived principally from media reports. Suffice to say, the last year in the RAND study, 2006, was a bloody one indeed, corroborated by United Nations and Iraqi government figures. And the RAND total? Well, you’ll just have to download the report to find out …