Coalition Casualties in Afghanistan

Far be it for me to claim there is no problem with Afghan soldiers attacking coalition troops in Afghanistan. Obviously the problem is real—and so is the fallout. Witness French President Nicolas Sarkozy’s threat to pull French forces out of the country after four French soldiers were killed by an apparent Afghan soldier, or at least a gunman in an Afghan uniform. But the New York Times still appears to be overhyping the threat with its lead article this morning which proclaims:

American and other coalition forces here are being killed in increasing numbers by the very Afghan soldiers they fight alongside and train, in attacks motivated by deep-seated animosity between the supposedly allied forces, according to American and Afghan officers and a classified coalition report obtained by the New York Times.

One has to read literally to the end of the article to find the details which make these claims less alarming than they appear at first blush. In the first place, that “classified coalition report” was not, apparently, a document issued by any U.S. or NATO headquarters. It was, according to the article, “prepared for a subordinate American command in eastern Afghanistan” (presumably Regional Command-East, but why not use the official title?) “by a behavioral scientist who surveyed 613 Afghan soldiers and police officers, 215 American soldiers and 30 Afghan interpreters who worked for the Americans.” In other words, this was one of countless documents prepared for various commands in Afghanistan by contractors and employees; it is not, as far as I can tell, an official statement of policy.

Which is not to deny that the problems outlined in the report, as summarized in the Times, are real—especially the lack of trust that too often characterizes relationships between U.S. and Afghan units. But such issues are hardly new; they were also prevalent in Iraq, and in both countries they can be addressed by better training (on both sides) and closer mentoring. Paradoxically, coalition soldiers are likely to be safer the closer they get to their Afghan counterparts—this builds bonds of trust that are the best protection anyone can have in this tribal society. By trying to hold Afghan soldiers at arm’s length—for example by walling off their compounds from U.S. bases or preventing them from carrying weapons on U.S. bases—coalition troops are actually increasing the risk to themselves because they send a clear message to their Afghan partners that they are not trusted.

Episodes of Afghan soldiers attacking their coalition partners certainly do much to fray trust, but it may be doubted this is quite as much of an epidemic as the Times claims. Deep in the article we read as follows:

The classified report found that between May 2007 and May 2011, when it was completed, at least 58 Western service members were killed in 26 separate attacks by Afghan soldiers and the police nationwide. Most of those attacks have occurred since October 2009. This toll represented 6 percent of all hostile coalition deaths during that period, the report said.

Twenty-six attacks in four years works out to fewer than seven attacks a year. And the casualty total from these attacks has been inflated by one particularly grisly and unexpected outrage which occurred in April 2011 when an Afghan Air Force colonel killed eight U.S. officers and contractors in Kabul.

The larger picture, which goes unmentioned entirely in the article, is that coalition casualties are not only considerably lower in aggregate than in Iraq but are also declining. According to icasualties.org, 566 coalition troops were killed in Afghanistan last year, down considerably from the 711 killed in 2010—an achievement especially impressive because in the meantime the total number of coalition forces in Afghanistan had surged and they had moved into some of the Taliban’s toughest strongholds.

Again, I don’t mean to minimize the problem with “blue on blue” attacks in Afghanistan, but alarmist reports such as this one can give a skewed presentation of the on-the-ground reality.