Over the weekend, I highlighted the UN report which found that the number of civilian casualties in the war in Afghanistan rose for the fifth year in a row, at 3,000 in 2011. This puts the number of civilians killed in Afghanistan since 2007 at over 12,000. The UN report claimed that insurgents were responsible for 2,332 civilian deaths in 2011, while the U.S.-led forces were responsible for 410 civilian deaths. The most deadly NATO tactic was airstrikes, which killed 187 civilians.
But another 63 people were killed in U.S.-led night raids, the report said. A commenter rightly pointed out that previous UN and NATO estimates of civilians killed in night raids drastically underestimated the figures. Gareth Porter has investigated these and I cited them in a December piece on night raids. A July UN report which claimed that only 30 civilians died in raids during the first six months of 2011, Porter et al. found, included only a very small fraction of night raids in which civilians were killed, according to officials of the independent Afghan commission that co-produced the report. According to other data later released by the U.S. and NATO and scrutinized extensively by Porter here, well over 1,500 civilians were killed in night raids in less than 10 months in 2010 and early 2011.
So suffice to say that the estimates in the latest UN report are almost certainly too low. And it should also be reiterated that Obama has made night raids a central tactic in the Afghan war. “An estimated 12 to 20 night raids now occur per night,” a study in September found, “resulting in thousands of detentions per year, many of whom are non-combatants.” And many of the associated tactics, like “mass detention operations, holding entire villages for questioning on site for prolonged periods of time,” may violate international law. This fuels resentment and has exponential effects for death and bloodshed even if no one is killed in a specific raid.
In John Tirman’s new book, The Deaths of Others, night raids as a tactic in Iraq were briefly covered and this passage is particularly relevant here (from page 230):
Clearing out villages and towns involved house-to-house searches that frequently violated Iraqis’ social norms of privacy (especially of women and girls) and honor (especially of men). Many thousands of Iraqi men and boys were arrested in front of their families and sent to detention camps or prisons like Abu Ghraib without any evidence of wrongdoing. “Probably 99 percent of those people were guilty of absolutely nothing,” a two-star army general later told Congress in 2006, “but the way we treated them, the way we abused them, turned them against the effort in Iraq forever.” Some 120,000 Iraqis were detained. In the house-to-house searches, if Iraqis resisted in any way, violence could readily follow. The procedure itself was violent. “You grab the man of the house. You rip him out of bed in front of his wife. You put him up against the wall,” recalls an army scout posted in Baquba, the capital of Diyala province north of Baghdad. Army privates gather the other family members in one room. “Then you go into a room and you tear the room to shreds” looking for weapons. A sergeant adds that the house is turned upside down. “You’ve just humiliated this man in front of his entire family and you’ve destroyed his home. And then you go next door and do the same thing in a hundred homes. Now, next week ten roadside bombs go off…And nobody can understand why.”
I'm glad to see you question the UN estimates in Afghanistan, which have gotten a free pass in the news media. Like Iraq Body Count, these are "passive" in that they merely add up other counts. In this, one never knows what's missed, and they accept the military's definition of who is a civilian, what actually occurred during operations, and so on. In other words, the reporting mechanism is weak and intermittent if not haphazard. The household survey technique used in Iraq, which is far more reliable, has never to my knowledge been attempted in Afghanistan. Hence, we have grossly underestimated the actual toll, which I suspect is about 100,000 for all causes and all Afghans.