On March 5, the United States used unmanned drones and manned aircraft to drop bombs on a group of what it described as al-Shabab militants at a camp about 120 miles north of Mogadishu, Somalia, killing approximately 150 of them. The administration claimed that the militants presented an imminent threat to African Union troops in the region with whom US advisers have been working, although it produced no evidence to support the claim. The news that the United States had killed 150 unnamed individuals in a country halfway around the world with which it is not at war generated barely a ripple of attention, much less any protest, here at home. Remote killing outside of war zones, it seems, has become business as usual.
This is a remarkable development, all the more noteworthy in that it has emerged under Barack Obama, who came to office as an antiwar president, so much so that he may be the only person to win the Nobel Peace Prize based on wishful thinking. Our Peace Prize president has now been at war longer than any other American president, and has overseen the use of military force in seven countries—Afghanistan, Iraq, Syria, Pakistan, Libya, Yemen, and Somalia. In the latter four countries, virtually all the force has come in the form of unmanned drones executing suspected terrorists said to be linked to al-Qaeda or its “associated forces.”
That an antiwar president has found the drone so tempting ought to be a warning sign. As Hugh Gusterson writes in Drone: Remote Control Warfare:
If targeted killing outside the law has been so attractive to a president who was a constitutional law professor, who opposed the war in Iraq from the very beginning, who ended the Central Intelligence Agency’s torture program, and who announced his intention to close the Guantanamo Bay detention camp on assuming office, it is unlikely that any successor to his office will easily renounce the seductions of the drone.
And it is not only President Trump or Clinton we need to worry about. Other countries are unlikely to be reticent about resort to unmanned aerial warfare to “solve” problems beyond their borders. Already, Israel, the United Kingdom, Iran, Iraq, Nigeria, and Pakistan have joined the US in deploying armed drones. China is selling them at a list price of only $1 million. In short order, most of the developed world will have them. And when other nations look for precedents, Obama’s record will be Exhibit A.
So what is Obama’s record? If it is to be a guide for future conduct, it is important to understand precisely what he has and has not asserted and done. And as Obama looks to the end of his tenure, the critical question is, what can and should he do now to mitigate the risks that a world armed with drones will become a place in which lethal force is a first rather than a last resort?
Some critics equate Obama’s drone record with the war crimes of his predecessor, George W. Bush. Glenn Greenwald, for example, charges in his afterword to The Assassination Complex that Obama’s drone policy “embodies the worst of what made the Bush-Cheney ‘war on terror’ approach so destructive.” Fordham Law School professor Karen Greenberg maintains, in the collection Drones and the Future of Armed Conflict, that Obama’s drone war “is but a refocused and incrementally more legally rationalized version of standards and assumptions that have persisted since the beginning of the War on Terror.”1
Greenwald summarizes Obama’s approach to drones as follows:
The centerpiece of his drone assassination program is that he, and he alone, has the power to target people, including American citizens, anywhere they are found in the world and order them executed on his unilateral command, based on his determination that the person to be killed is a terrorist.
If this were indeed Obama’s policy, the charges leveled by Greenwald and Greenberg and echoed by many other critics would be justified. But it is not accurate. First, Obama has not claimed the power to kill “terrorists,” but only those fighting on the other side in an armed conflict authorized by Congress against al-Qaeda and organizations allied with it. The power to kill the enemy in an armed conflict is as old as war itself.
Second, Obama has not asserted the power to use lethal force “anywhere…in the world,” but only in war zones—where drones are just another weapon—and, outside war zones, only where an enemy fighter poses an imminent threat that cannot otherwise be addressed, usually because the host country is incapable of capturing the fighter. When the host country is capable of arrest and prosecution, according to the administration, killing is not an option. Thus, under Obama, hundreds of persons suspected of engaging in or supporting terrorism have been arrested—in the US, the UK, and many other nations—and brought to trial for their alleged crimes. Obama has never claimed the authority to kill individuals who are outside a war zone and subject to capture.
Third, multiple sources, including Greenwald’s own website The Intercept, have reported that Obama hardly chooses targets on his own, but has set up an elaborate process that involves the review and input of many high-level military and government officials before any targeted killing is approved. Obama has insisted on taking ultimate responsibility, as he should, but it is hardly “he, and he alone,” who makes the decision.
It is also important to note that Obama’s policy and practice of using drones have evolved significantly over the course of his presidency. His initial years in office were marked by an aggressive expansion of the drone program. In Pakistan, for example, according to the New America Foundation, President Bush oversaw forty-eight drone strikes, killing between 377 and 558 people, whereas President Obama has overseen 355 strikes, killing between 1,907 and 3,067 people. But the number of drone strikes in Pakistan under Obama peaked at 122 in 2010, and has dropped every year since then. There were only ten strikes in Pakistan in 2015, and thus far only three in 2016. The number of strikes has also dropped in Yemen, from a high of forty-seven in 2012 to twenty-four in 2015 and nine thus far this year. In short, President Obama has shown significantly less proclivity to rely on drones in his second term than in his first.
Critics charge that drones have killed large numbers of innocent civilians. Here, too, the record is complicated. Drones have often killed innocents, including one strike that hit a wedding party in Yemen in 2013, killing twelve; and another, under President Bush, that hit a religious school in Pakistan in 2006, killing more than sixty children. For years, the US refused to acknowledge any of its drone strikes, and therefore offered no accounting of who was killed, combatant or civilian. Several independent organizations have sought to fill the void, although obtaining accurate information is extraordinarily challenging because of the secrecy surrounding strikes, the obstacles to reaching many of the targeted regions, the difficulty of piecing together evidence after explosions rip bodies apart, the fear of local residents about speaking to outsiders, and the propaganda interests of terrorist groups in exaggerating the number of civilians killed.
Nonetheless, the London-based Bureau of Investigative Journalism estimates that as of May 24, 2016, between 493 and 1,168 civilians have been killed in US drone strikes in Pakistan, Yemen, and Somalia combined. The New America Foundation more conservatively estimates between 370 and 448 civilian deaths from US drones in the same three countries. On July 1, the Obama administration, after seven and a half years of silence on the subject, reported that it killed between 64 and 116 civilians in 473 “counterterrorism strikes outside areas of active hostilities” between January 2009 and December 2015. It specified that it deems Iraq, Syria, and Afghanistan areas of “active hostilities” and so its report did not include civilian casualties there, despite reason to believe they may have been considerable.
The administration attributed the discrepancies between its figures and those of organizations such as New America Foundation to its own allegedly superior access to intelligence, and to the independent organizations’ reliance on news accounts that may be susceptible to terrorist propaganda. Yet in 2011, John Brennan, then Obama’s top national security adviser, claimed that not a single civilian had been killed in a US drone strike in the previous year, illustrating that it’s not only terrorist groups that seek to spread propaganda.
At the same time, an important question with respect to civilian casualties from drone strikes is, compared to what? In one sense, the frequency of civilian casualties seems a curious charge to level at drones. As University of Massachusetts at Dartmouth professor Avery Plaw has convincingly argued, of the various mechanisms of warfare, drones are the least likely to result in civilian deaths.2 Manned aircraft cannot be nearly as discriminating, because they lack the drone’s ability to hover for hours in order to find a propitious time to strike. And ground troops will almost inevitably do more “collateral damage” than a drone strike.
In another sense, though, it is the drone’s very promise of precision that gives allegations of causing civilian casualty their force. If this weapon has the capacity to kill combatants without harming civilians, why have so many civilians died? Civilian casualties that in prior times seemed tragic but inevitable are all the more suspect when they are avoidable.
The civilian deaths are attributable to many factors, the most important being the difficulty of identifying “the enemy” when he hides among civilian populations and does not wear a uniform. Faulty intelligence is another cause, as is a reliance on metadata associated with cell phones that may or may not be in the possession of the intended target. As one former drone operator told The Intercept, “We’re not going after people—we’re going after their phones, in the hopes that the person on the other end of that missile is the bad guy.” Or as Michael Hayden said in a 2014 debate with me at Johns Hopkins University, “We kill people based on metadata.”
Still, civilian casualties from drone strikes outside war zones have dropped precipitously in recent years. The New America Foundation, for example, reported between forty-nine and sixty-three civilian deaths in Pakistan in 2011 alone, yet from 2014 through 2016 it identifies only two. In Yemen, it reported sixteen civilian casualties in 2012, yet only five in 2014, and none in 2015 and 2016. The Bureau of Investigative Journalism’s numbers are generally higher, but it, too, shows a decline, and reports very few civilian deaths from drones in recent years.
One reason for the improved record may be the new standard that Obama announced in May 2013 for drone strikes outside zones of armed conflict. In a speech to the National Defense University, and in a classified Presidential Policy Guidance issued simultaneously, he stated that he would authorize targeted killings beyond war zones only when (1) an individual poses a continuing and imminent threat to US persons; (2) capture is not feasible and no reasonable alternative is available to address the threat; and (3) there is a near certainty that no civilians will be injured or killed. I and others have raised questions about the administration’s elongated definition of “imminence,” and about what makes capture not “feasible,” but if these terms were interpreted strictly, there would be little to criticize, and one would expect to see few strikes and fewer civilian deaths.3
Thus, Obama’s actual record on drones is mixed. While he employed the tactic aggressively in his first term, he has become much more discriminating since then. There are likely two principal reasons for this. First, over the course of his tenure, and in response to widespread criticism, Obama has become increasingly transparent about the program, albeit fitfully. At the outset, the administration refused even to acknowledge that the targeted killing program existed. Such a policy of secret, unacknowledged killing is not only legally and ethically illegitimate, but counterproductive, because in the face of secrecy people will fear the worst, and evidence of civilian killing is likely to emerge in any case.
Beginning with a speech by then State Department legal adviser Harold Koh in March 2010, however, the Obama administration began to defend the program in public, in increasingly detailed speeches. In May 2013, President Obama himself addressed the subject, admitted for the first time that the US had killed several American citizens with drones, and issued the restrictive guidelines noted above.
This March, in response to a lawsuit filed by the ACLU, the administration agreed to release a redacted version of the heretofore classified Presidential Policy Guidance. In a sharp break from past practice, the administration has publicly acknowledged all nine of its drone strikes this year in Yemen. And on July 1, when the administration released its accounting of casualties from drone strikes beyond active war zones, it also issued an executive order on minimizing civilian casualties with respect to all uses of force. With transparency comes accountability; it is no accident that greater openness has resulted in a more restrictive policy, and more restrained deployment of drones.
Second, the administration may have come to realize the strategic shortcomings of its initial heavy reliance on drones. On the one hand, as Peter Bergen and Jennifer Rowland note in the collection Drone Wars, drone strikes are effective at eliminating dangerous individuals in hard-to-reach places—assuming they have been properly identified. In May, for example, the Obama administration used a drone to kill the Taliban’s leader, Mullah Akhtar Muhammad Mansour, in Baluchistan, Pakistan. Killing off leaders can undermine and demoralize the enemy, and is fair game in an armed conflict. As Audrey Cronin notes in the collection Drones and the Future of Armed Conflict:
The drone threat has disrupted and altered terrorist operations. It has forced al-Qaeda and its associates to dramatically change their behavior, keeping them preoccupied with survival and hindering their ability to move, plan, and carry out operations.
On the other hand, Cronin writes, drone strikes “have not thwarted leadership succession, ongoing propaganda operations, or local attacks.” And if drones inspire resentment and promote support for our enemies, they may be counterproductive. Most accounts suggest that they do just that. Imagine how Americans would feel if another country was executing individuals living among us by dropping bombs from unmanned aircraft hovering overhead—and refusing even to acknowledge that it was doing so. General Stanley McChrystal, who commanded US troops in Afghanistan, told The Huffington Post in 2013 that drone strikes create “a perception of American arrogance” and generate hatred “at a visceral level.” Admiral Dennis Blair, Obama’s first director of national intelligence, has similarly argued that drones are counterproductive because they cause so much anti-American resentment in targeted countries.
In 2012, opinion polls reported that as many as 90 percent of Pakistanis opposed drone strikes and 74 percent considered the US an enemy, despite the fact that Pakistan is second only to Afghanistan (and ahead of Israel) in the amount of US foreign aid it receives. Saba Imtiaz, a Pakistani journalist, suggests in the collection Drone Wars that this and other polls generally do not capture the views of those who actually live in the Pakistani border regions that have been most targeted. A recent study focused on the targeted Pakistani border regions found that residents living there are in fact less opposed to drones than the Pakistani population as a whole.4 But even so, it cannot be good for American security interests if three quarters of Pakistanis consider the US the enemy.
Terrorist groups, Cronin and others have argued, have generally been defeated not through military means, but by isolating them from their potential communities of support. One of the strongest grounds for isolating groups such as al-Qaeda and ISIS is that as many as 85 percent of the victims of their attacks are themselves Muslim. But “drones are not helping to discredit al-Qaeda tactics of indiscriminately killing Muslims,” Cronin writes. “Our ‘collateral’ damage obscures which actor is doing more of that.”
Precisely because of their special capabilities, drones make the resort to force too tempting. They are cheaper than other forms of force; they put no American lives at risk; and they allow for a kind of superficial plausible deniability. As Stephen Wrage, a US Naval Academy professor, argues, “if there are unusually useable weapons in the arsenal, there will be unusual pressures to use them.” As a result, Gusterson warns, drones enable “a kind of permanent, low-level military action that threatens to erase the boundary between war and peace.”
The question for President Obama is whether he wants to be remembered as the leader who ushered in the era of permanent, low-level drone warfare. His actions will be looked to for justification by those that follow, here and abroad. As Daniel Reisner, former head of the Israel Defense Forces legal department, has said, “If you do something for long enough, the world will accept it…. International law progresses through violations.”
It is fully within Obama’s power—and therefore responsibility—to leave office with a less checkered record on drones. But to do so, he needs to commit to several further reforms. As one knowledgeable former Obama official told me on background, it is not enough to make the Presidential Policy Guidance public; he should formally issue it as an executive order, which would make it more difficult for a successor to abandon. And he should make the restrictive conditions in that policy the basis for launching discussions with our NATO allies on a commonly accepted set of standards for the use of drones outside of war zones. Whether we like it or not, drone warfare is the way of the future, and it is in our, and the world’s, interest to establish at an international level a high threshold for resort to such force.
The administration’s July 1 report on civilian casualties is an important step forward. It marks the first effort to offer a public accounting of the drone war’s results. The executive order issued the same day commits to similar reports on an annual basis going forward (although a future president could rescind the order, it might be politically difficult to do so). And it imposes obligations on all agencies involved in the use of force, within or outside zones of armed conflict, to minimize civilian casualties, and, “as appropriate and consistent with mission objectives,” to acknowledge and make condolence payments to civilian victims or their survivors. The administration deserves substantial credit for all of this.
At the same time, the disclosure and executive order fall short in disappointing ways. By providing data only in the aggregate for a seven-year period, the administration makes it impossible to check the accuracy of its figures against the incident-by-incident reports of nongovernmental organizations. The White House claimed that its report reflected Obama’s desire to be “as transparent as possible with the American public.” But he certainly could have been more forthcoming. Why not, for example, present the casualty figures from 2009 to 2015 by year, as he has committed his successors to do, rather than lumping them into one figure? Such a breakdown might tell us whether the more restrictive strike policies he announced in April 2013 have made a difference. Or why not report casualties on a strike-by-strike basis, as US Central Command sometimes does with respect to civilian casualties in war zones? No explanation was offered. Without more specificity, the numbers serve little purpose other than as propaganda for the administration to counter what it considers terrorist propaganda.
Moreover, while most of the executive order on civilian casualties applies to all uses of force, wherever they take place, the annual reporting requirement applies only to civilian casualties outside zones of active conflict. But surely civilian casualties deserve public acknowledgment no matter where they occur. The administration offers no explanation for its limited reporting requirement.
President Obama should also ensure that before he leaves office, the authority to conduct drone strikes is transferred entirely to the Department of Defense. At the moment, the CIA conducts some strikes, and the military carries out others. Both are presumably subject to the same substantive standards. But the culture of the CIA does not encourage fidelity to law; the military’s culture does. The military is familiar with law-of-war questions of necessity, proportionality, and the like; the CIA is not. And the CIA’s default position is secrecy, so leaving drone strikes in its hands will impede efforts at greater transparency.
Most importantly, President Obama needs to commit to justifying not only the program’s general guidelines, but its actual implementation: the authority to take another human being’s life must be subject to strict legal limits. Accountability requires at least some degree of formal oversight. In Israel, all targeted killings must be reviewed by the courts after the fact. The European Court of Human Rights is moving in the direction of requiring review of killing even on the battlefield.5
Yet under President Obama, the US has executed thousands of individuals, far from any battlefield, and has yet to offer any specific accounting of the basis for its actions, with the lone exception of the September 2011 killing in Yemen of Anwar al-Awlaki, an American citizen. In particular, the administration has not explained who it has killed, the basis for the decision to kill, or the actual result of specific drone strikes. Killing enemy fighters in wartime on the battlefield is an accepted practice, and generally does not require individualized justifications; enemy status is sufficient. But when a nation asserts the power to kill specific individuals, outside of war zones, based on their alleged misconduct, it must justify its actions—and must do so publicly, to the extent possible. That justification has been lacking. Secret executions cannot be squared with the rule of law. They are the stuff of death squads, not democracies.
As President Obama said in his 2013 NDU speech, “The same human progress that gives us the technology to strike half a world away also demands the discipline to constrain that power—or risk abusing it.” He now has the opportunity to impose constraint in a meaningful way. If he fails to take it, his legacy will be as the Nobel Peace Prize winner who pioneered a dramatically dangerous and ethically dubious form of warfare. That’s not the Obama I want to remember.
For a critical discussion of the specific legal justification advanced for the killing of the US citizen Anwar al-Awalaki, see Karen J. Greenberg, Rogue Justice: The Making of the Security State (Crown, 2016), pp. 205–222, and H. Jefferson Powell, Targeting Americans: The Constitutionality of the US Drone War (Oxford University Press, 2016). The courts have declared the issue a “political question” not fit for judicial resolution. ↩
Avery Plaw, “Counting the Dead: The Proportionality of Predation in Pakistan,” in Killing by Remote Control: The Ethics of an Unmanned Military, edited by Bradley Jay Strawser (Oxford University Press, 2013). ↩
Aqil Shah, “Drone Blowback in Pakistan Is a Myth. Here’s Why,” The Washington Post, May 17, 2016. ↩
Public Committee Against Torture in Israel v. Government of Israel, Case No. HCJ 769/02, December 13, 2006; Al-Skeini v. United Kingdom, App. No. 55721/07, 53 Eur. H.R. Rep. 589 (European Court of Human Rights, 2011). ↩