Since the drive through Kansas is long and mostly eventless—save for a truly great record store in Lawrence—I managed to listen to a bunch of back-episodes of This American Life. Good stuff. Particularly good was a segment on Iraqi casualties, which featured a lengthy interview with Marc Garlasco, an analyst for Human Rights Watch. Garlasco used to do targeting for the Pentagon, and now he analyzes the fallout from those bombing runs, which makes him an interesting dude in his own right, but almost as an aside he brought up something that should really be a major story.

Basically, when Pentagon planners select targets to bomb, they figure out the value of the target and weigh it against the number of civilian casualties that might result. If, say, they want to take out a house where some suspected insurgent leader is sleeping, but there's a hospital next door, they might select a smaller bomb, alter the angle of attack, and basically fiddle with various factors in order to reduce the number of estimated casualties below a certain arbitrary number—I think thirty civilian deaths is the cutoff point for the occupation of Iraq. If the number is above the cutoff, then the Secretary of Defense and president need to sign off on the attack.

On some level, the military really does try to minimize casualties in these sorts of operations, but what's especially stunning is that the Pentagon never examines the aftermath of their attacks to see if their casualty projections actually panned out. So they have no way whatsoever of knowing whether their algorithms for estimating civilian deaths are accurate in the first place. They basically just guess. When the reporter asked military people why they didn't track civilian casualties, many of them seemed to shrug and say it was just their business to win wars. Maybe so, but then it's also quite clear that the United States isn't doing everything possible to avoid civilian casualties. Not even close. Recklessness would be a better word.

P.S. Garlasco, as it happens, was also the Human Rights Watch analyst who was widely cited because he supposedly dismissed the original Lancet study on Iraqi casualties. But according to TAL, what actually happened was this: Garlasco got a call from a Washington Post reporter while riding a train. He protested that he hadn't seen the study and didn't actually know much about the statistical methods involved. But the reporter was on deadline and really, really needed a quote, so Garlasco basically said he guessed that the casualties figures seemed kind of high. That quote ended up in the paper, under a paragraph noting that "experts immediately challenged the new estimate." Journalism, needless to say, is truly awesome.