Weighing the Strategic Impact of Killing Civilians in Counter-Militancy
Andrew Kenealy
In order to avoid putting ever more boots on the ground, the United States has increasingly turned to using airstrikes and drone strikes to damage a wide range of militant groups. Yet whenever President Obama contemplates dropping a bomb from the air, he risks killing civilians, which presents an unavoidable tradeoff. The benefits of damaging enemies are relatively simple to measure. More U.S. strikes on a group means it will have fewer weapons, less money, and fewer fighters. But in addition to significant and enormously important moral problems – not to mention larger strategic issues like potentially undermining international law -- how do we measure the on-the-ground costs of taking civilian lives? The answer lies in an assessment of the context of each mission.
Whether a militant group seeks to establish its own government or strike a subway system in the United States, the thoughts and opinions of the local population matter to its success. Beyond directly taking up arms and joining a terrorist or insurgent group, supportive non-combatants can help raise funds and resources, provide cover, or stay quiet when questioned by government officials. Those facts create a genuine strategic interest in preventing locals from joining, or even supporting, a militant group that the United States is seeking to eliminate. Accidentally killing civilians, by alienating locals and fueling extremism, can make the initial use of force counterproductive.
Initially, though, the only means through which killing non-combatants can exert a strategic cost on the ground is by altering the thoughts and feelings of community members. Accordingly, the first step in assessing the strategic weight of killing non-combatants is to determine how it will actually alter local public opinion.
The thoughts, emotions, and political stances that form a local public opinion are nuanced and complex, not black and white. While killing civilians may make local non-combatants hate the United States, we should avoid the false dichotomy that leads us to expect that U.S.-inflicted collateral damage will cause citizens to feel positively towards a militant group. As innovative polling in the Federally Administered Tribal Areas of Pakistan (where most drone strikes have occurred) revealed, civilians overwhelmingly disapprove of U.S. strikes on the Pakistani Taliban. But they also overwhelmingly disapprove of the Pakistani Taliban itself.
Strikes that kill a family member or friend are horrific, but they will not necessarily push a civilian towards a militant group that he or she would otherwise hate. Conversely, strikes that kill civilians in communities more initially sympathetic towards a militant group should be much more likely to increase the group’s support base. Understanding what the local public thinks about a militant group before a U.S. bomb falls, therefore, is necessary to predict how they may think about it after.
The second step in weighing the strategic impact of collateral damage is to determine what upset people will actually do. Although the underlying reasons for why people support terrorist groups or insurgencies are complex and the subject of intense scholarly debate, it is still possible to make some intuitive probabilistic conclusions on when and where the effects of increasing the number of civilians sympathetic to a militant group will be more or less severe.
For instance, it seems clear that radicalized civilians seeking revenge on a government practicing indiscriminate violence would be more likely to join a militant group that takes efforts to attract revenge-seekers than one that does not. Therefore, groups such as Al Qaeda in the Arabian Peninsula (AQAP) that recruit based onshared grievance and ideology should experience greater success in recruiting angry civilians than groups like Boko Haram, who have turned to recruiting heavily through forced conscription. Consequently, killing non-combatants in areas where AQAP operates may ‘create’ many more militants than would be created by killing non-combatants in areas where Boko Haram operates.
Similarly, civilians angered by collateral damage may refrain from sharing intelligence they have about militants with the United States, but this should negatively affect different missions to different degrees. Because large counter-insurgency campaigns place troops on the ground that can protect the local populationfrom militants (generally considered a precondition for civilians to cooperate with the government), COIN missions can collect a great deal of human intelligence. Killing innocents and alienating civilians who could share crucial information on the militants may significantly reduce otherwise promising human intelligence collection prospects, and in turn, harm the mission overall.
But many places where the United States carries out targeted strikes have no significant government presence, U.S. or otherwise. In these areas the local population likely has no readily accessible means of providing intelligence to the United States, and no credible guarantee of safety if they do. To compensate for these human intelligence collection constraints, U.S. intelligence is likely derived primarily from drone surveillance or signals intelligence, all of which can and will be collected regardless of civilian emotions. In these counter-militancy missions, killing civilians may not affect U.S. human intelligence collection prospects because prospects were very poor or nonexistent to start.
Reducing the public’s support of a militant group and collecting intelligence on the militants are only two parts of U.S. counter-militancy strategy whose success depends to some degree on the thoughts and feelings of the local population. And, as demonstrated, circumstances unique to each mission set this degree of influence. (Does the local population hate the militant group? Does the militant group recruit on ideological grounds? Can the United States access human intelligence to begin with?) Understanding these circumstances provides insight into how likely civilians’ anger is to influence each piece of U.S. counter-militancy strategy. This, in turn, provides a measure of how much killing civilians can negatively influence the overall success of counter-militancy missions.
To be clear, measures should always be taken to avoid noncombatant deaths. Moral considerations should weigh heavily. As we saw in Vietnam, images of killing civilians abroad can turn the American public against a mission. But because targets are often embedded within the local population and 100% certainty does not exist in the intelligence upon which these strikes are based (consider the U.S. bombing of the Kunduz hospital), approving a strike poses an inexorable risk of killing civilians.
Assigning a precise cost for each civilian killed in counter-militancy is beyond the scope of this analysis. However, as this piece suggests, the strategic impact of collateral damage depends in large part on the context of the specific counter-militancy mission, and it is possible to distinguish where this cost is relatively higher or lower. Determining when and where the strategic costs of causing collateral damage are acceptably low is an imperfect and blood-soaked calculus. But to neglect to do so is to forgo strategy.