<<previous  |  index  |  next>>

V. Airstrikes: Policies for their Use, Problems in Implementation

Whether civilian casualties result from aerial bombing in Afghanistan seems to depend more than anything else on whether the airstrike was planned or was an unplanned strike in rapid response to an evolving military situation on the ground. 

When aerial bombing is planned, mostly against suspected Taliban targets, US and NATO forces in Afghanistan have had a very good record of minimizing harm to civilians. In 2008, no planned airstrikes appear to have resulted in civilian casualties.  In 2007, it appears that only one planned airstrike resulted in civilian casualties.⁵⁶ In 2006, at least one attack resulting in civilian deaths may have been a planned attack.⁵⁷

Planned attacks allow the US and NATO to use civilian risk mitigation procedures, including formal risk estimates to model and minimize civilian casualties. This includes a “pattern of life analysis,” which looks for civilians in the area for hours or days before an attack using “eyes on the target” ranging from ground observers to technical reconnaissance. According to NATO Judge Advocate General (JAG) staff, the US and NATO also require positive visual identification of the target during a planned strike, allowing the pilot to look for civilians and call off an attack based on those observations. Planned strikes also allow the US and NATO to develop a target over time, thereby using far more detailed intelligence to understand who is and is not in the target area.⁵⁸ 

US and NATO forces have been far more likely to cause civilian casualties in unplanned situations, normally when ground troops call in airstrikes as tactical support when under attack from insurgent forces, or to target insurgent forces on the move. The vast majority of known civilian deaths and injuries from airstrikes in Afghanistan come in these situations. Often these situations involve “Troops in Contact” (TIC). A TIC is an unplanned engagement occurring when US or NATO ground forces unexpectedly come into contact with insurgent forces. According to US and UK operational planners, in such situations, the US or NATO forces are usually instructed to withdraw from the area without any airstrikes being carried out due to concerns for civilian casualties; airstrikes are most often carried out when withdrawal is not an option, as when troops are ambushed and find their retreat blocked.⁵⁹ A number of the cases described here, however, suggest that airstrikes sometimes are used in other circumstances, as when forces are pursuing insurgents and are uncertain about the extent of civilian presence in the area. But even in TIC situations US and NATO forces sometimes have gone beyond close air support for troops in danger. A number of battles (such as Chora, described above) began as TICs but lasted for several hours or days, with airstrikes used to support small troop numbers on the ground resulting in civilian deaths. This suggests that the US is not taking all feasible precautions during prolonged battles, including using adequate forces to minimize civilian harm, employing low-collateral damage bombs, and positively identifying the locations of combatants and civilians. 

While US planned airstrikes require both positive identification of the target and a “pattern of life” assessment to determine if civilians are in the area, TICs use far fewer checks to determine if there is a civilian presence. The tactical collateral damage assessment performed by the Joint Terminal Attack Controller (JTAC), a service member qualified in directing airstrikes on the ground is one of the only checks done, and, of necessity, such assessments often are made under the stress of hostile fire. 

A senior US military lawyer told Human Rights Watch, “We will never engage [with airstrikes] when we know civilians are in a compound. Problems happen when we don’t know.”⁶⁰ According to Colonel David Diner, a US Judge Advocate General, US planes have dropped bombs when they did not know for certain who was in a compound. So long as a valid military target is identified, such attacks are not unlawful on their face, but they raise concerns about whether “all feasible precautions” have been taken to minimize civilian loss, as required by the laws of war.⁶¹

An important factor contributing to civilian casualties is the US dependence on Special Operations Forces, as highlighted by the Sangin airstrikes, discussed above, that resulted in at least 21 civilian deaths. Because such forces typically operate in small groups and are relatively lightly armed when conducting counter-insurgency missions, they often require rapid support in the form of airstrikes when confronted with superior numbers of insurgent fighters. Several NATO officials told Human Rights Watch of their grave concerns with the lack of coordination between NATO and US Special Operations Forces on counter-insurgency missions. NATO forces have complained that OEF operations in their regions are not communicated to them, but civilian casualties from airstrikes called in by OEF forces are left for them to address. 

Another problem may be the differing missions and Rules Of Engagement (ROE) between US and NATO forces. When on offensive operations, both NATO and US forces use the same ROE. However, when on the defensive they use different ROEs, the US rules allowing for use of force—including airstrikes—in a broader range of situations.⁶² Offensive operations are defined here as planned engagements while defensive operations range from TICs to responding to perceived threats.

NATO and the US both require “hostile intent” for aerial munitions to be employed to defend their forces. NATO defines “hostile intent” as “manifest and overwhelming force.” The US ROE defines hostile intent as “the threat of the imminent use of force,”⁶³ a much lower threshold than NATO for employing airstrikes, permitting anticipatory self-defense. According to a senior US general in Afghanistan: “For the United States, ‘hostile intent’ is the use of imminent force. One difference is the US says imminent does not have to mean instantaneous. US troops have a different standard [than NATO].”⁶⁴  

This subjective definition significantly lowers the bar for US forces to call in airstrikes (or other forms of support, such as ground-based ordinance), allowing US forces, even those under NATO control, to use airstrikes in circumstances in which their NATO allies cannot. In Afghanistan, a nation where the civilian population is often armed, and where insurgents do not wear any defining uniform or insignia that would differentiate them from civilians, the expansive ROE is likely to lead to mistaken attacks against civilians.

This ROE “conflict” creates an imbalance in response options that is of concern to some NATO officials, who believe the US relies too much on airstrikes. One ambassador interviewed in Kabul told HRW, “Some Afghans think the US is worse than the Russians [who occupied Afghanistan from 1979 until 1989]. The problem is in the TIC they call in air support in a hurry, and special forces go too far on the ground and call in airstrikes too often. There is a cultural problem with the US – they are cowboys.”⁶⁵ In addition to the operational confusion this may cause forces on the ground, it also creates the impression that US forces are less concerned about protecting civilians than their NATO counterparts. 

A top-ranking US commander, speaking to Human Rights Watch on condition of anonymity because he was not authorized to speak to the organization, defended the US use of the more elastic definition of hostile intent: “You want to give the commander on the ground flexibility within the laws of armed conflict.” He was not concerned with the separate US and NATO ROEs, saying, “The ISAF [NATO] definition has no teeth as all countries have their own standards.”⁶⁶ 

It is unsurprising that civilian casualties are higher in TIC situations. These ground actions between insurgent and US or NATO forces often only occur in the first place because of inadequate intelligence about the presence of insurgents. This may come about because of poor communication with local residents or because of attempts by the insurgents to draw coalition forces into an unanticipated ambush, since the element of surprise is always a prominent tactic of insurgent movements. These intelligence problems become exacerbated when airstrikes are called in to help extricate US or NATO forces or target insurgents on the run. 

NATO lawyers involved in airstrikes told Human Rights Watch that in some TIC situations in which airstrikes have been called in, US and NATO forces did not know who was in the area they were bombing. Civilian casualties increase when forces on the ground do not have a clear picture of the location and number of combatants and civilians in an area. Such gaps in knowledge, when combined with fear and the “fog of war” at times mean that forces resort to airstrikes when options less likely to cause civilian loss are available. In winning the tactical battle quickly on the ground with bombs, US and NATO forces risk losing the strategic battle for the support of the population, essential in counter-insurgency operations.

In July 2007, following public outrage and President Karzai’s criticisms, NATO announced that it would change its tactics to reduce civilian casualties. The changes included using smaller munitions in bombs, suspending attacks when civilians might be harmed, and letting the Afghan National Army conduct house-to-house searches. NATO’s Secretary General Jaap de Hoop Schaeffer said NATO “was working with weapons load on aircraft to reduce collateral damage.”⁶⁷