TPMM on Petraeus's death count methodology.
Exclusive: Petraeus' Sectarian Death Count Methodology
By Spencer Ackerman - September 21, 2007, 4:44PM
In the debate over the surge, there have been a number of questions raised within the government about an important metric for understanding whether the U.S. military's strategy is succeeding -- how Multinational Force-Iraq calculates sectarian violence.
Earlier this month, David Walker of the Government Accountability Office testified that he could not "get comfortable" with General David Petraeus' methodology for determining sectarianism, considering it too inferential to be reliable. His report, echoing objections from senior intelligence officials, instead tabulated the pace of attacks on civilians and found the surge didn't appear to have a significant effect on civilian-targeted violence. However, relying on data interpreted through the MNF-I methodology, Petraeus testified that sectarian violence had fallen in Iraq to mid-2006 levels.
The actual methodology MNF-I employs has remained unknown. Until now.
In response to a Freedom of Information Act request I filed two weeks ago, MNF-I has provided TPMmuckraker with its criteria for identifying ethnic and sectarian violence. We've added the methodology to our Document Collection, and you can read it here.
[...]
An important thing to note, in addition to the effect the season has on violence in Iraq, is that unless they're using the same accounting now that they did in the pre-surge past they're comparing apples and oranges and saying there are fewer oranges now...
Cheers,
Scott.