Coming Clean About Casualties

The U.S. military in Afghanistan needs to acknowledge what it doesn’t know

One of the challenges the U.S.-led coalition faces in the war in Afghanistan is controlling the narrative surrounding its actions. Often, the accounts given by officials differ so sharply from those of local eyewitnesses that the coalition’s portrayal of events seems disconnected from reality. The recent bombing controversy in western Afghanistan is only the newest case. By examining how various stories diverged over the days after the incident, a clear pattern emerges: the coalition has a problem with damage control.

The undisputed facts are fairly sparse: in the early pre-dawn hours of Friday, August 22, a small joint American-Afghan force came under attack in a village called Azizabad, in the Shindand District of southern Herat Province. While offering support to the embattled unit, an AC-130 gunship fired into a cluster of buildings. Beyond this, the facts become murky, even contradictory.

On Saturday, August 23, Voice of America reported that, while Afghan officials complained that the strike had killed “at least 70 people, including women and children,” a U.S. coalition spokesman claimed that a battlefield assessment indicated only thirty Taliban were killed, and that a follow-up investigation was under way.

On Sunday, August 24, The Washington Post published a story noting that a U.S. military spokesperson maintained that thirty Taliban insurgents died during the fight. Local officials, including a spokesman for the western regional command of the Afghan National Army, claimed they found the bodies of sixty children and nineteen women among the dead.

This was a potentially explosive issue—civilian deaths in Afghanistan have led to violent riots before, most infamously the “traffic riots” in Kabul in 2006—so local and international groups traveled to the area to investigate.

By Monday, August 25, they were counting the dead in Azizabad: a local human rights group claimed “at least 78 people were killed” during the clashes and the air strike, while the Ministry of Interior claimed seventy-six people died, including fifty children under the age of fifteen. The Ministry of Defense adhered to the new U.S./NATO line of twenty-five dead militants and five dead civilians (probably a clarification of the initial count of thirty).

Meanwhile, Afghan president Hamid Karzai increased the anger in his public statements about the incident, upping the body count to a reported eighty-nine. He also fired two high-ranking Afghan National Army (ANA) officers for their involvement in the firefight.

The situation took an altogether different turn by Wednesday, August 27, when the U.N. announced that its own investigation revealed “convincing evidence” that at least ninety civilians, sixty of them children, had died during the Azizabad incident. They also directed stern words at NATO and the U.S. over the safeguarding of civilian lives during combat operations, though a Pentagon spokesman maintained that the U.S. strike was “a legitimate one, a Taliban target.”

By Thursday, August 28—six days after the initial incident—the Pentagon leaked its own version of events to The New York Times, claiming that photographs of the scene after the firefight revealed little evidence for a higher civilian death toll, the U.N.’s investigation was cursory, and other reports relied too heavily on the testimony of local villagers rather than physical evidence such as freshly dug graves or injured people in local hospitals. One U.S. military investigation repeats the claim of twenty-five dead insurgents and five dead civilians, but still proposes additional “joint” investigations of the incident—leaving one to wonder what, exactly, they are still investigating.

This creates an unfortunate case of he-said/she-said for outside observers. While a BBC Persian crew traveled to the area and interviewed locals who corroborated a higher body count, English-language news services have not yet completed their own investigations of the incident. This leaves everyone outside of Shindand reliant on the statements of the U.N., coalition media relations, and local activist groups—none of which are neutral or disinterested actors (the U.N. typically uses local stringers for investigations and surveys).

There is, however, precedent to follow. On July 17, almost exactly one month before the current incident, NATO claimed to have killed two Taliban commanders in this same area, while local tribal elders claimed fifty civilians were also killed. For days, NATO officials denied civilian casualties, though they eventually admitted some deaths. In April 2007, too, upwards of fifty-seven civilians died in various coalition operations in Shindand District.

In other bombing incidents, such as the wedding parties attacked in Nangarhar province and in Nuristan Province (both, weirdly enough, on July 6), U.S. officials have also denied the presence of any civilians among the dead for many days. Regarding the Nuristan incident, one coalition media officer went so far as to say the coalition had “no reports of civilian casualties,” even while Al-Jazeera English was publishing the testimony of locals who were wounded or lost relatives in the attack.

When asked about the latest incident in Shindand, General James Conway, the commandant of the Marine Corps, said the Taliban intentionally surround themselves with civilians so as to increase the civilian death toll—part of their sophisticated propaganda campaign to trick the world into thinking the U.S. and NATO are reckless. Given the categorical denials of civilian death during the previous six weeks that have since been retracted, the coalition is running out of sympathetic ears when new civilians are reported killed.

The way the coalition has handled these incidents creates the impression that they are callous or even casual about dead civilians: repeatedly denying non-coalition body counts without evidence to back their claims, calling the dead “Taliban” when they are nothing of the sort, and disparaging human rights groups trying to confirm ground conditions. All of this serves to isolate the Pentagon from real social currents on the ground. Moreover, it sets up an expectation that, no matter what actually happened, the official response will be to deny until forced to admit—which, when its account differs so greatly from local accounts of these incidents, encourages the idea that the coalition is lying.

That is almost certainly not the case. It is difficult to imagine the U.S. military or NATO not caring about non-combatants dying by their hands. Yet without better reporting from the scenes of these alleged tragedies, it is difficult to move beyond hearsay accusations. There is, however, one simple solution: admit to having imperfect information. General Conway came close when he explained that troops aren’t “on the inside looking out” of compounds when they order air strikes. If the coalition, instead of rushing to declare itself blameless and diligent immediately after combat, admitted the possibility of innocent casualties, that would increase the likelihood for widespread confidence in the follow-up investigation.

This runs counter to the very real need to tightly restrict information. Yet in a war that relies as much on perception as it does success in battle, this tendency to hunker down and deny an unfortunate reality does far more harm than good. While the military waits for its follow-on investigation, Afghans are left wondering—rightly or wrongly—why the coalition seems not to care when they die.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Joshua Foust is a military consultant. He is a contributor to PBS Need to Know, a contributing editor at Current Intelligence, and blogs about Central Asia and the Caucasus at