Burnham et al (NEJM) estimated excess deaths from March 2003-June 2006.
The US military estimates of civilian deaths are from January 2006-December 2007.
Comparing Roberts et al and Burnham et al to the US military estimates is comparing apples to oranges prima faccia from dates (I don't know the exact methodology for the US military estimates, although they tracked closely to the Ministry of Health numbers from what I recall, at times being slightly greater in number). However, just like IBC, these numbers capture a slightly different statistic than Roberts et al and Burnham et al, which capture "excess" deaths due to not only violence, but also to different mortality rates from disease, birthing, etc. Thus, even if the dates were the same, you can't compare the two statistics without first stripping out non-violent deaths from the Roberts et al and Burnham et al studies.
In terms of my own confidence intervals, as Herodotus will attest to since we've been kicking around this subject for almost four years now between here but mostly at another board, the Lancet studies grossly overestimate deaths from OIF. Although I'm sure there are some false deaths captured in IBC due to insurgent propaganda, I'm pretty confident that even with all the scrubbing of news reports that many deaths go unreported, and so I consider the IBC to provide a floor in terms of the lower bound of the number. The Burnham study comports very well with the range of excess deaths I had felt was likely based on the floor provided by IBC and how IBC didn't capture deaths due to disruption of health services and sanitation services, etc. (notice, I'm not attaching casuality here, just stating that these services ebbed and flowed based on the security situation).
The Burnham study, while conducted past the peak violence months in Iraq, didn't capture the peak months of violence, and so I wouldn't be surprised if a study were conducted today that the total wouldn't have increased by 50-100K in the intervening time. However, we're at a point now where violent deaths are probably at or below pre-war levels and services are close as well.
Any investigation into civilian deaths is probably going to detail targeting and ordnance information as well as other operational details, most if not all of which should be classified. Any report takes on the classification level of the highest classification piece within it (at least within the US military), so if there's only one piece of information (say a single phrase within one sentence) in a 1000 page report and it's classified "super duper extra sensitive secret," then the entire report has to be classified "super duper extra sensitive secret."Originally Posted by mkenney
I don't know if the UK MoD has a different policy with regards to the classification of material, but going with the assumption that it's not different, then even if all the information regarding #s of non-combatant deaths/injuries isn't classified, if it's contained within a report that's classified, then it's a crime to pass on the report or any portion of the report unless you have the information screened by someone with the authority to declassify or authorize its release.
You're going to need to explain this one, because I'm reading this as baiting as an insinuation that NATO military elements deliberately target non-combatants and should try to figure out how to be more effective in this. There will be zero tolerance for future statements like this if my read is not mistaken.Originally Posted by mkenny