Ascertaining an accurate count of the number of civilians to die in the Iraq war — a bloodbath that they neither invited nor deserved — has been impossible. Some deaths are never reported, some are suppressed and the people who keep track of the carnage often have an ax to grind, which results in a predisposition to under or over report.
If there is anything approaching a consensus view, and I use that term advisedly, it is that somewhere between 80,000 and 87,000 civilians have died since the U.S.-led invasion in March 2001. That is the range arrived at by the folks at Iraq Body County based on data cross-checked from media reports and hospitals, morgues, non-government organizations and official figures.
That number stands in stark contrast to the claim that 601,027 civilians died in a Johns Hopkins University cluster-sample survey published three weeks before the mid-term elections in The Lancet. By contrast, Iraq Body Count listed 47,702 deaths during the same period.
The survey provoked a firestorm. Anti-war advocates seized upon it to advance their view that the civilian toll was far worse than was being acknowledged, while pro-war advocates argued that it was deeply flawed.
Drawing on my own experience in survey taking and analysis and despite a healthy dose of skepticism, I myself came down somewhat on the side of the survey. But just as there still are people looking for proof of those elusive Iraqi WMD, there have been people hard at work trying to debunk the survey.
“Data Bomb,” the most thorough and persuasive effort to date, has just been published in The National Journal.
Authors Neil Munro and Carl M. Cannon write that the survey authors:
* Were ideologically predisposed to conclude that there was a much higher body count and the timing of the survey’s release just before the election was no accident.
* Followed a model that ensured that even minor components of the data, when extrapolated over the whole population, would yield substantially higher numbers.
* Have made it difficult to resolve apparent inconsistencies in their methodology and analysis by not making available to other researchers the surveyors’ original field reports and response forms and not just collated survey results.
* May have engaged in fraud.
The survey was led Gilbert Burnham, a Johns Hopkins University professor, with assistance from Riyadh Lafta, and Les Roberts.
Burnham defends the survey, although not all that persuasively, in an interview with Pajamas Media editor Richard Miniter.
Please click here to read more at Kiko’s House.