HOW MANY ARE WE KILLING? There's a lot of discussion out there about civilian casualty counts, as the argument that casualties in Iraq have dropped would appear to be one of the central measurable claims that the White House will make in defense of the surge. As Ilan Goldberg and others have noted, the statistics are very twitchy; they don't count intra-Shia and intra-Sunni violence, and they don't count car bombings. The latter are excluded because such violence isn't strictly "sectarian", while the latter ... I'm not sure, but I'd guess it has something to do with the exclusion of statistically exceptional events, which makes about as much sense as excluding three run home runs from the final score of a baseball game.

I may have missed it, but it also doesn't look as if casualties caused by U.S. action are included. This is problematic, given that the surge has involved an increase in the number of troops in Iraq and a significant increase in the tempo of offensive operations. This has included airstrikes, despite the fact that FM3-24 (the counter-insurgency manual) specifically warns against the use of air power in counter-insurgency warfare. These operations must have killed a lot of people, both insurgents and bystanders; at least regarding the latter, I wouldn't be at all surprised if those numbers exceeded the lives saved by the "drop" in sectarian violence. Indeed, even the former (insurgent deaths) is meaningful for this calculation, since many extra-judicial killings involved members of competing militias or insurgent groups.

--Robert Farley