Nova Scotia Scott

Based on media reports, I would have assumed that bombings killed more civilians in Iraq between 2003 and 2008 than any other cause. According to a study published earlier this week in the New England Journal of Medicine (NEJM), however, that assumption would be wrong.

Utilising the detailed database assembled by Iraq Body Count, a team of researchers found that, of non-combatant deaths from armed violence during the five-year period from March 2003 through March 2008, one-third were inflicted by execution, 20 percent by small-arms gunfire, 14 percent by suicide bombing, with smaller proportions for vehicle bombs, roadside bombs, mortar fire, and air attacks. Of those who were executed, almost three in ten showed signs indicating torture.

The greatest proportion of victims — 19,706 of 60,481, or 33% — were killed by execution after abduction or capture. Of the bodies of those who were executed, 5760, or 29%, showed marks of torture, such as bruises, drill holes, or burns. (A typical media report about this particularly appalling form of violent death reads: “The bullet-riddled bodies bore signs of torture and their hands were tied behind their backs.”) Iraqi civilians also suffered heavy tolls from small-arms gunfire in open shootings and firefights (20% of deaths), apart from executions involving gunfire, and from suicide bombs (14% of deaths).

In events with at least one Iraqi civilian victim, the methods that killed the most civilians per event were aerial bombings (17 per event), combined use of aerial and ground weapons (17 per event), and suicide bombers on foot (16 per event). Aerial bombs killed, on average, 9 more civilians per event than aerial missiles (17 vs. 8 per event). Indeed, if an aerial bomb killed civilians at all, it tended to kill many. It seems clear from these findings that to protect civilians from indiscriminate harm, as required by international humanitarian law (including the Geneva Conventions), military and civilian policies should prohibit aerial bombing in civilian areas unless it can be demonstrated — by monitoring of civilian casualties, for example — that civilians are being protected. [footnote omitted]

Women and children were disproportionately represented among victims of attacks using relatively indiscriminate weapons—air strikes and mortar fire. Men tended to be killed using more precise methods—gunfire and executions.

Assyrian International News Agency has an additional tidbit not found in the text of the NEJM article.

One author of the study thinks many of those involving torture “had to do with trying to get people to move out of their houses” because they were from rival Muslim sects.

The article lists six co-authors, three from Iraq Body Count; two from Royal Holloway College, University of London; and one from King’s College, London.

The analysis encompassed only victims of short-duration events (two days or less). Civilians killed in prolonged violence, i.e., the two sieges of Fallujah and some episodes during the initial stage of the invasion (20 March-30 April 2003), were excluded from consideration. This was necessary because deaths occurring in those situations could not be reliably linked with type of weapons used to kill. Thus, the overall number of civilian deaths—91,358—was reduced to 60,481 included in the analysis.

h/t: personal communication from Michael Spagat

Tags: Iraq, Methodology, Middle East