Caught in the crossfire

On September 2, 2010 an airstrike conducted by Joint Special Operations Command in Afghanistan’s Takhar province killed a man named Zabit Amanullah and nine of his companions. NATO forces in Afghanistan believe the raid killed a Taliban deputy governor called Mohammed Amin, but there is ample evidence that all those killed were in fact civilians who were caught in the crossfire of a military intelligence case of mistaken identity.

I began investigating the Takhar air strike as soon as it happened because I knew Zabit Amanullah, who had previously worked with me as a human rights researcher. With the help of another Afghan friend who had acted as Amanullah’s security focal point, by the following day, I had discovered the identities of those civilians killed in the attack. It took me six months to find the real Mohammad Amin and work out the relationship between him and Zabit Amanullah. Special Forces helpfully supplied the Afghanistan Analysts Network, which recently released an authoritative investigation into the Takhar airstrike, with the sketchy biographical details they had on Amin. I sought the help of contacts within the Taliban in northern Afghanistan to find the real man who matched their profile. 

The Amin I met in March of 2011 clearly was the man Special Forces had been hunting. The clinching details were that he had served as Taliban deputy governor for Takhar, elaborated on family relationships that the Special Forces had notes on, and even carried an identity card. He also shared enough information on his family background and career to locate him in the northern Afghanistan sociopolitical landscape like a marker on the terrain in Google Earth. From our conversation it was clear that he was still active in the insurgency, although he was predictably reticent about his current role, as he was fully aware that the September 2 strike had been intended for him.

The relationship between the military’s intelligence and the kill order was critical in the September 2 disaster. Once Mohammad Amin’s name was on the military’s target list, the military was essentially on the lookout for any opportunity to kill him. If Special Forces had found him herding a flock of sheep, they could still have been authorized to kill him.

The September 2 tragedy could have been avoided if the dossier on Amin’s insurgency career had been used to put him on a watch list rather than a kill list. A watch list would trigger follow up surveillance to spot actual insurgent activity. Accurate surveillance of the election convoy prior to September 2 would have revealed a pattern of public meetings and visits to government buildings that is inconsistent with insurgent activity. The men would not have been spotted laying mines or loading lots of weapons into their vehicles and so the kill order would never have been given. Eventually, NATO’s intelligence might even have been reexamined and Special Forces might even have found the real Mohammad Amin.

The obvious alternative to kill targeting is capture. The fundamental point in favor of capture is that mistakes can be undone. However, an argument against capture is that Afghanistan’s detention system is highly flawed. Afghan authorities find it difficult to keep insurgents in jail and there is a high rate of recidivism. But detention is preferable to killing the innocent. Those on the watch list — against whom there is strong intelligence that they may be involved in some kind of insurgent activities — would be far better captured and interrogated than killed. Shah Jehan, the police chief in Takhar, made this point regarding Zabit Amanullah: the men knew each other and with one phone call, Shah Jehan could have summoned him for investigation.

Another issue the Takhar case raises is the unseen third alternative to kill/capture: persuade. It is deeply troubling that NATO is investing so much of its resources in trying to kill men of the caliber of the Mohammad Amin I met. He could be considered a "grey area insurgent" — someone who was indeed fighting against the government and NATO, but who was a rational actor with a plausible list of grievances who could potentially be reconciled to the Afghan government. Currently, the reintegration program is meant to provide the mechanism for fighters to exit the conflict with Afghan government security guarantees and some prospect of economic assistance. The program is bogged down, unable to get funds to where they are needed, lacking after care for those who have reintegrated, and subject to questions over what percentage of the men reported reintegrated were actually insurgent fighters before they joined. Critically, the program lacks a way for commanders to pledge to cease operations when it is unsafe for them to associate with government overtly. The lack of an effective reintegration program means that people like Mohammad Amin are being targeted unnecessarily.

For a cross-check on how incidents like this play out in other insecure areas of Afghanistan, I asked a veteran Afghan researcher for an assessment of the impact of airstrikes in Paktia, a province which has seen an alarming rise in the incidence of insurgent attacks. He immediately listed militant commanders, operating from or supported from Waziristan, in Pakistan’s tribal areas, who were killed in precision airstrikes and ambushes on the battlefield, killed while laying mines or travelling in pick-ups loaded with weapons. My colleague was confident that the killing of these men hurt the insurgency. But he was far less confident in the security benefits of the night raids, another component of the kill/capture campaign, in which residential compounds are surrounded and searched. In some cases, there were no insurgents. In other cases, the actual insurgents managed to evade capture in the confusion of the raid. In both cases, the families of those targeted in the raids tend to protest, and their complaints are echoed by insurgent propagandists and the Karzai government, which has repeatedly called on the U.S. to halt the night raids.

The most striking, and troubling, aspect of the whole Takhar case has been the way NATO and JSOC have continually asserted that they got it right. Quite apart from the human loss and the damage to NATO’s credibility among Afghans, this insistence on sticking to an implausible case hinders military learning. In a challenging environment like Afghanistan, the military needs to learn and adapt. The reluctance to learn exhibited in this case suggests that intelligence failures like that which confused the identity of Zabit Amanullah and Mohammad Amin may not be so rare. Despite the military’s confidence that technological advance has radically transformed its ability to conduct these surgical strikes, the system is fallible. There should be more emphasis on targeting those actively engaged in hostile acts rather than those who are merely watch listed. Both detention and reintegration must become preferred outcomes. But as the U.S. increasingly relies upon the high tech combination of signals intelligence and precision airstrikes for counterterrorism, policymakers and military officials deciding when to deploy these potentially deadly tools need to know how reliable they are. If the reluctance to address social knowledge and acknowledge errors becomes part of the military’s culture surrounding these new weapons systems, policymakers will simply be in the dark about the consequences of this push-button warfare.

Michael Semple is a fellow at the Carr Center for Human Rights Policy at the Harvard Kennedy School.