Countering Terrorism with Targeted Killings

Congress should

• suspend or drastically reduce targeted killings, as executed by drone strikes and special operations forces, in Pakistan, Somalia, and Yemen;
• rely instead on more indirect methods of counterterrorism, such as military assistance for foreign forces combating terrorists, or nonmilitary methods, such as intelligence operations;
• require the military, Central Intelligence Agency, or a congressional panel to publish studies on the efficacy of lethal counterterrorism methods, especially targeted killings by drone strikes;
• hold oversight hearings on those studies; and
• vote on authorizing the use of force in nations where targeted killings might occur and include geographic, temporal, and possibly other limits in the authorizing language.

Although the phrase “war on terrorism” has fallen out of rhetorical favor in Washington in the 15 years since the 9/11 attacks, the war is still going strong. U.S. forces today target terrorists in seven foreign nations. Nearly 10,000 U.S. troops remain in Afghanistan, fighting Taliban militias as well as a splinter group claiming affiliation with the Islamic State of Iraq and Syria, or ISIS. U.S. airstrikes, aided by small numbers of U.S. ground troops, are attacking ISIS in Iraq and Syria. U.S. aircraft recently commenced bombing the ISIS-affiliated group in eastern Libya. And U.S. unmanned aerial vehicles (drones) target terrorists amid those wars and in Pakistan, Yemen, and Somalia, where they are aided by U.S. special operations forces who identify airstrike targets, support local forces, and occasionally conduct killing raids.

This chapter evaluates those U.S. efforts to combat terrorists outside larger wars, focusing on armed drones and the gray area between peace and traditional warfare that they inhabit. According to the Bureau of Investigative Journalism, the United States has conducted 424 drone strikes in Pakistan, 136–156 in Yemen, and 32–36 in Somalia between 2002 and October 1, 2015, collectively killing between 3,318 and 5,288 people, 10–20 percent of them civilians. Both the military, through Joint Special Operation Command, and the Central Intelligence Agency (CIA) conduct the strikes, primarily using Predator and Reaper drones armed with Hellfire missiles. A recent Obama administration decision is reportedly shifting the CIA out of that paramilitary role.

Keep in mind that drones are weapons serving a policy of targeted killing. That policy is what is evaluated here. The general utility of drones for the U.S. military — where they are primarily used as surveillance platforms — is not the subject. Nor does this chapter analyze the wisdom of the wars in Afghanistan, Iraq, and Syria, which other chapters in this volume cover.

Evaluating the efficacy of targeted killings to counter terrorism is difficult for several reasons. One is the murky nature of counterterrorism policies. Observing and measuring their effect requires some information about the operations and motivation of terrorist organizations, which is not something that terrorists eagerly share. The risks of targeted strikes, meanwhile, come in forms that resist measurement, such as militants’ motivation to attack the United States.

The other obstacles to evaluation result from institutional dysfunction in the U.S. government: excessive secrecy and anemic oversight. Some secrecy is necessary, of course, to protect intelligence sources and methods. But our government uses secrecy to prevent people from second-guessing its policies. The Obama administration refuses to divulge even its general strategic and legal reasoning concerning the targeted killing of terrorists.

That abuse of secrecy undermines Congress’s ability to provide oversight. Members of Congress cannot meaningfully oversee policy without being able to publicly critique it. And that would bother Congress were it inclined to scrutinize U.S. wars. But Congress allows the executive branch to launch strikes on terrorists as it sees fit, under the dubious legal authority of the 2001 Authorization for Use of Military Force. That law actually authorizes the use of force only against the perpetrators of the 9/11 attacks and those who aided or harbored them. A Congress more jealous of its war-making powers would hold hearings before authorizing strikes against a new group or nation. Those hearings could force the administration to explain its rationale for strikes.

Evaluating Success

These circumstances — the inherent difficulty of evaluating counterterrorism policy and the government’s refusal to help — make evaluation of counterterrorism programs rough and uncertain. Still, to assess the worth of counterterrorism programs, we must weigh their costs and benefits, however speculatively. What follows is an outline of such an evaluation, first considering possible benefits and then potential costs.

“Benefits” here means successful counterterrorism — terrorism that did not occur. That is different than attacking or even killing terrorists. Defenders of targeted killings, especially executive branch officials, tend to conflate strikes with counterterrorism, muddling means and ends, presumably because the former is easier to measure.

A program’s benefits should be measured by comparing its results to what would occur in its absence. That may seem obvious, but advocates of drone strikes often insist that the only alternative is something more costly, like sending in ground forces. Were they right, drone strikes would be a bargain. But doing something nonmilitary, or doing nothing at all, is always an option.

Terrorism’s limited threat to Americans bounds the benefits that targeted killings provide. The benefits often attributed to military counterterrorism programs — such as the absence of major attacks in the United States since 2001 — rely on high baseline expectations for what terrorism might accomplish. In fact, the terrorist threat to Americans is tiny compared to even moderate health risks and far smaller than political rhetoric generally suggests. Globally, terrorists killed 13 Americans in 2013 and 32 in 2014, according to the National Consortium for the Study of Terrorism and Responses to Terrorism, or “START,” database. Annual totals were not much higher in the late 1990s, casting doubt on the argument that the vast increase in U.S. counterterrorism efforts subsequent to the 9/11 attacks is what is keeping deaths low. More plausibly, credit should go mostly to relatively simple and relatively cheap measures, such as tracking terrorists through intelligence work, policing, and immigration controls at borders. As John Mueller writes in Chapter 66 on homeland security, most counterterrorism programs at best make improbable dangers slightly less probable.

This take on the danger posed by terrorists remains controversial, and the prior paragraph alone is unlikely to change many minds. Skeptical readers should consult the suggested readings below. Still, all readers should at least agree that, if this take on terrorist capability is correct, then counterterrorism policies with uncertain and distant costs are probably not worthwhile.

Even the limited potential benefits of targeted killings are difficult to achieve. That’s because killing terrorists is a tactic serving a counterterrorism strategy that is difficult to execute. That strategy, which I call disruption and decapitation, uses those two methods to suppress and ultimately destroy terrorist groups. “Disruption” undermines terrorists’ ability to function and launch attacks by keeping members on the run, deterring recruits, and demoralizing members. “Decapitation” means killing a group’s leaders to undermine its organization and morale. Strikes can aim at both goals. If the group is small enough, this strategy might morph into “elimination,” such that strikes aim to kill off an organization entirely.

There are indications that the strategy succeeds in suppressing terrorism. Intelligence gleaned from killed or captured terrorists, including Osama bin Laden, suggests that the drones do cause terrorists to devote great efforts to self-preservation. At least according to some U.S. officials, al Qaeda’s organization in Pakistan has been nearly eliminated. Off the record, the Obama administration gives that reason for slowing down drone strikes there. Only 3 have occurred in 2016 after a peak of 128 in 2010, by the Bureau of Investigative Journalism’s count.

Still, there is reason to suspect that this strategy is insufficient to defeat terrorist organizations. Disruption and decapitation grow out of strategic airpower theory, which holds that airstrikes aimed at the enemy’s industry and political leadership can achieve victory in a conventional war, independent of armies. That theory has a remarkably poor track record. Historical studies, like the Strategic Bombing Survey undertaken by the U.S. government after World War II, show that strategic airstrikes tend to do less harm than planned and to trigger people’s sense of national solidarity, heightening leaders’ support without much dampening their ability and desire to fight.

Studies of decapitation as a counterterrorism tool largely follow the same critique. With some important exceptions, the literature takes a skeptical view, contending that decapitation is effective only in narrow circumstances, such as when targeted groups are young, small, and unpopular. That description poorly fits current terrorist adversaries, with certain exceptions, such as the ISIS affiliate in Libya. Some studies suggest that successful decapitation may fragment terrorist organizations into pieces that are even more dangerous. A simpler reason to doubt that disruption and decapitation work is that U.S. drone strikes continue in each nation where they have occurred. The enemy may be disrupted, but it evidently persists.

Hidden Costs of Targeted Killings

One reason that policymakers are inclined to support targeted killings in spite of these difficulties is that the direct costs are low. Budgetary costs — including some portion of war and intelligence budgets, plus much of Special Operations Command and the cost of buying and operating armed drones — are probably less than $20 billion annually, which is just 3 percent of U.S. military spending. Moreover, the annual number of U.S. personnel killed in service of these missions is generally either a handful or none. These seemingly low costs encourage a “why not?” approach toward targeted killings.

The problem is that the strikes may have far higher costs than is initially evident. Because those dangers are nebulous and slow to arrive, policymakers focused on the short term will tend to overlook them. One such risk is blowback, which occurs when airstrikes or raids anger people in targeted nations in ways that cause them to retaliate. Individuals might become terrorists themselves or support the targeted organization, meaning that strikes generate ever more terrorists to hunt. Blowback might create support for anti-American insurgents or foreign leaders’ disinclination to cooperate with U.S. requests.

Proponents of drones argue that they create little blowback because their surveillance and precision targeting capability allows discriminate strikes. As the Stimson Center’s Task Force on Drone Policy recently noted, drones generally produce fewer civilian casualties than weapons systems that would be used in their stead. However, the scholar James Igoe Walsh points out that drone strikes are typically used in remote, dangerous regions where reliable targeting intelligence is tough to get and verify. Strikes relying on shaky intelligence inevitably strike the wrong target on occasion, as in 2015 when a CIA strike in Pakistan killed an Italian and an American hostage. So although drones likely produce less blowback than alternative military means that might be used in their stead, they may produce a considerable amount compared with not striking at all.

Another risk of targeted killings, especially drone strikes, is their potential to serve as a slippery slope to wider or continual war. Drone strikes make wars easier to start but not necessarily easier to win. Terrorism tends to occur amid insurgencies driven by political disagreements that drone strikes do not settle. Even strikes that succeed in suppressing terrorism may seem to fail as the conflict persists and the terrorist group continues to function. That perceived failure might also encourage a president, especially one facing domestic criticism for failing to eradicate a terrorist organization, to escalate the war.

Another slippery slope scenario arises if targeted killings create new enemies. By targeting Islamist insurgents — due to fear that they are either planning attacks on the United States or helping groups that are — strikes might turn insurgents focused on local enemies into terrorists working to kill Americans. Strikes, in other words, might produce what they are intended to combat. Because the terrorists that the United States targets in Pakistan, Yemen, and Somalia are enmeshed in broader Islamist insurgencies, it is likely that this scenario has already occurred to some extent.


These potential costs, along with the limited benefits of counterterrorism and the difficulties of executing a counterterrorism strategy of disruption and decapitation, justify skepticism about the value of targeted killings. The United States should vastly reduce drone strikes and special operations raids outside broader war zones and rely on either more indirect methods of counterterrorism (such as military assistance for local forces) or nonmilitary methods (such as intelligence cooperation). These alternative approaches are not silver bullets, but they carry far less risk of blowback and a slippery slope to broader war.

Congress should heighten oversight of targeted killings. That might require statutory reform to ensure that Joint Special Operation Command’s lethal actions outside traditional war zones are reported to the armed services committees. Congress should consider suspending targeted killing in Pakistan, Yemen, and Somalia, at least until the executive branch provides studies or public testimony explaining why the policy makes sense in each nation. At that point, Congress should debate the matter and vote on whether to authorize strikes. Authorization should be limited to a particular state or region and expire after a time period or under certain conditions.

If the next president or Congress will not stop, drastically reduce, or provisionally suspend drone strikes outside war zones, they should at least require the military, CIA, or a congressional panel to publish studies of the benefits and risks of current military counterterrorism efforts, especially drone strikes. Although any studies’ judgments would be uncertain, the exercise might at least produce debate conducive to insight.

In general, U.S. security policy should reflect greater wariness of methods said to achieve ends historically delivered by war while avoiding its traditional costs. Whether the agent is drones, special operations forces, cyberattacks, or bomber aircraft, policymakers should keep in mind the old saying that you can’t get something for nothing. The full cost might be hidden, with the bill coming. Or the good might do less than promised. If it costs so little, it probably isn’t worth too much.