U.S. Special Operations Forces have been increasingly aiming their night-time raids, which have been the primary cause of Afghan anger at the U.S. military presence, at civilian non-combatants in order to exploit their possible intelligence value, according to a new study published by the Open Society Foundation and The Liaison Office.
The study provides new evidence of the degree to which the criteria used for targeting of individuals in night raids and for seizing them during raids have been loosened to include people who have not been identified as insurgents.
Based on interviews with current and former U.S. military officials with knowledge of the strategic thinking behind the raids, as well as Afghans who have been caught up in the raids, the authors of the study write that large numbers of civilians are being detained for brief periods of time merely to find out what they know about local insurgents – a practice the authors suggest may violate the Geneva Conventions on warfare.
A military officer who had approved night raids told one of the authors that targeting individuals believed to know one of the insurgents is a key factor in planning the raids. "If you can't get the guy you want," said the officer, "you get the guy who knows him."
Even when people who are known to be civilians have not been targeted in a given raid, they have been detained when found on the compound of the target, on the ground that a person's involvement in the insurgency "is not always clear until questioned", according to military officer who has been involved in operational questions surrounding the raids interviewed for the report.
Raids prompted by the desire for intelligence can result in the deaths of civilians. The Afghan Analysts Network, a group of independent researchers based in Kabul, investigated a series of night raids in Nangarhar province in October-November 2010, and found that the raids were all targeting people who had met with a local religious cleric who was believed to be the Taliban shadow province governor.
Two civilians were killed in those raids when family members came to the defence of their relatives.
The report notes that many Afghans interviewed said night-time operations had targeted a number of compounds simultaneously, in some cases covering entire villages.
In a village in Qui Tapa district of Konduz province, SOF units, accompanied by Afghan army troops, conducted a raid that detained 80 to 100 people, according to the report. The interviewees said a masked informant pointed out those people to be taken a U.S. base to be interrogated.
The idea of using military operations to round up civilians to exploit their presumed knowledge of the insurgency has a long history in the U.S.-NATO war in Afghanistan.
The Pentagon official in charge of detainee affairs until the end of 2005 told IPS that concerns about "over-broad detention" in Afghanistan - meaning the practice of sweeping up large numbers of civilians - were countered by pressures for "more aggressive detention operations".
As then head of NATO intelligence in Afghanistan, Canadian Brig. Gen. Jim Ferron, explained in a newspaper interview in May 2007, "The detainees are detained for a reason. They have information we need."
It is not clear that civilians actually provide important intelligence on insurgents, however. The civilian victims of night raids are family and friends of Taliban fighters and commanders, who have now incentive to provide information that would make it easier for SOF units to track them down.
But another factor inclines the Special Operations Forces commanders in Afghanistan to focus more on people for whom the evidence of involvement in the insurgency is weak or nonexistent, according to the new report. After taking heavy losses, in 2010, Taliban commanders at district level and above are increasingly residing in Pakistan rather than in towns in Afghanistan where they can be more easily targeted.
Without those targets on their lists, SOF units in Afghanistan may have had to choose between going after more civilians or reducing the number of operations. And the growth in the number of operations and the statistics on alleged insurgents killed or captured are a key measure of the relevance of SOF units.
An average of 19 raids per night were conducted during the period from December 2010 through February 2011, according to data published by Reuters last February. But a senior U.S. military adviser interviewed for the report in April 2011 said that as many as 40 raids were taking place in a single night.
A military officer involved in the night raids told an author of the study that there were no longer enough mid- to high-level commanders still active in Pakistan to justify the present high rate of raids, and many raids were now likely to be targeting people who are known not to be insurgents but who might know something about specific insurgents.
Other officers interviewed for the report denied that contention, however, claiming there were still plenty of commanders left to target.
The report suggests that it is dangerous to detain family members in particular in order to exploit their knowledge of relatives in the insurgency, because it further inflames an already angry population across the country.
"If that is the criteria, they might as well arrest all southerners," said one Afghan journalist living in Kandahar. "The person who is an active Taliban is either my uncle, cousin (or) nephew…"
Based on interviews with residents in villages where raids have taken place in the past several months, the report concludes that communities "see raids as deliberately targeting and harassing civilians, in order to discourage communities from providing food and shelter to insurgents, or to pressure them to supply intelligence on the insurgency."
Most of those civilians targeted or swept up in night raids are released within a few days, according to the report. That assessment is consistent with the revelation, reported by IPS in September 2010, that roughly 90 percent of the individuals who were said by ISAF in August 2010 to have been "captured insurgents" were in fact released either within two weeks of initial detention or within a few months after being sent to Parwan detention facility.
The authors of the report conclude that deliberately targeting and rounding up civilians who are not suspected of being insurgents merely to exploit possible intelligence value "may constitute an arbitrary deprivation of liberty" and thus "inhumane treatment" in violation of Article 3 of the Geneva Conventions.
The report suggests there is "anecdotal" evidence that the targeting for the raids has become more accurate.
But that anecdotal evidence appears to be contradicted by other anecdotal evidence that the targeting has become more indiscriminate in deliberately targeting civilians.