The Obama administration opened up significantly this summer about its efforts to kill suspected terrorists outside what it calls “areas of active hostilities,” like Afghanistan, Iraq, and Syria. It released information on the number of civilians accidentally killed in drone strikes and unveiled its framework for making targeting decisions. Coming in the waning days of the administration, these measures presumably represent the president’s last major maneuver in his efforts to defend a program that will help define his legacy—and that his successor will inherit.
The administration’s defense of the program has largely focused on its legality. Top government lawyers have repeatedly made the case that in countries like Pakistan, Yemen, and Somalia, the program not only meets but exceeds the legal constraints on targeting those the Obama administration deems threats. But is Obama’s drone war moral?
There are two parts to this question: First, can targeted killings outside of conflict zones ever be justified? Second, is the structure that Obama has put in place over the years actually upholding those standards? The answer to the first question, according to philosophers of war, is yes, and the standards for “moral” killing allow for more civilian casualties than one might expect. The answer to the second question is harder. The intense secrecy surrounding the program not only makes it difficult for outsiders to assess the morality of targeted killings, it means the government is failing to uphold its basic moral duties to the public and the people it is targeting.
When Obama has made the moral case for the drone war, he has claimed to uphold the highest standards in defending the nation’s security. “That means taking strikes only when we face a continuing, imminent threat, and only where there is … near certainty of no civilian casualties,” he told an audience at West Point in 2014. Last month, the administration gave its first self-evaluation of how well it had succeeded, claiming to have killed at most 116 civilians in drone strikes outside of war zones, compared with up to 2,581 “combatants.” (Independent estimates put the number of civilian casualties far higher.) But what do these numbers really signify? And can they help Americans grade the morality of the drone campaign as a whole?
The standards Obama articulated track reasonably well with the traditional requirements laid out by moral philosophers for carrying out violence in self-defense. To be morally justified, defensive killing must be necessary, meaning only people who pose serious threats can be targeted, and taking a life (as opposed to arresting the threatening party) must be the only possible means to end those threats. The defender must strictly distinguish between combatants and innocent bystanders, intentionally targeting only combatants, and must keep any damage to those bystanders in careful proportion to the initial threat.
Imagine that a government interrupted a Paris-style attack: A terrorist cell sets off suicide bombs at a stadium, and other cell members are racing to a concert hall to gun down civilians. Assuming it followed those moral guidelines, the government could blow up the terrorists’ car with a drone-fired missile. There is clearly a real threat, and, as the terrorists have made clear that they are willing to blow themselves up and will not be taken alive, killing them is necessary to stop them. To make killing the terrorists fully justified, the government must also show that it is not deliberately targeting any innocents, and that any harm to civilians is proportionally less than the damage the terrorists are expected to cause. In other words, if the terrorists have a lone hostage in the car, the threat to the many concertgoers may outweigh that hostage’s right to life, but if a bus full of schoolchildren is passing by, the moral calculus is different, and it may be necessary to hold off.
That’s the abstract case. But the principles are much more difficult to apply systematically to ongoing operations in the real world. It is rare for law enforcement to encounter terrorists who have shown their intentions but not yet completed their plots. The real Paris attackers weren’t stopped until they had already committed atrocities, nor were the 9/11 hijackers. If a government wants to avoid waiting until attacks are underway to move on terrorists, it has to find a way to preempt the violence.
That need to move defensive action back in time, before attacks occur, explains Obama’s reference to “continuing, imminent” threats. An imminent threat, to most people, means that an attack is just moments away. In the Old West, the good sheriff would wait for the murderous cowboy to start drawing his gun before shooting. But the Obama administration has argued that this idea is outmoded in an era when terrorists can plot to attack American civilians from safe havens overseas. Most famously, a leaked legal memo argued for a “broader concept of imminence” that permits defensive action even when a terrorist plot won’t unfold “in the immediate future.” There’s no reason, the thinking goes, to wait until the terrorist has his finger on the detonator of his suicide vest to take him out.
The U.S. government is on good moral ground here. “The imminence requirement, strictly speaking, is a temporal condition. And that’s just irrelevant,” says Jeff McMahan, an Oxford philosopher who has written extensively on when defensive violence is justified. “It’s just that the closer you get to the time of the attacks, the greater the certainty you have about the attack that you’re trying to prevent.”
What’s morally difficult about preventive violence, though, is it that it involves great uncertainty about whether a possible terrorist really will commit harm. “In the case of preventive killing, when somebody is just in the desert having dinner,” says McMahan, “they’re not engaging in an attack, one doesn’t have that vital piece of evidence.” If the sheriff truly knows the cowboy is about to murder someone, morality doesn’t demand he wait for the cowboy to draw before shooting him down.
Drones don’t use six-shooters, of course. Even if they do target the right people, their missiles may also kill innocent bystanders. Obama’s claim that the targeted killing program goes out of its way to minimize harm to civilians has been heavily contested. But military ethicists have long acknowledged that civilian deaths are permissible if they are kept in careful proportion to the lives saved by a particular operation.
“Suppose there’s a terrorist who we know has killed a number of people in the past, and is clearly planning or preparing for other operations, so we can predict that killing him will save the lives of hundreds of people,” says McMahan. “And if we kill him, we’ll kill 15 innocent people, and there’s no other way to eliminate his causal contribution [to the deaths of hundreds]. Then it looks like it’s going to be proportionate.” In other words, Obama’s drone program is not an automatic moral failure if it kills civilians along with terrorists. This is the same logic that permitted the police in Dallas to use an explosive to kill the sniper Micah Johnson after he killed five police officers. Explosives risk harming bystanders in addition to the intended target, but the police assessed, based on his actions, that Johnson threatened many more people than the explosives might have accidentally harmed.
In releasing data on civilian casualties this summer, the Obama administration had an opportunity to make the case that its targeted killing program is proportionate. Instead, the data shows where the ideal moral case breaks down as it meets reality. Assuming the accuracy of the statistics the government released in July—which are low compared to other counts—473 strikes killed from 64 to 116 civilians and 2,372 to 2,581 “combatants,” or as many as 40 terrorists for each civilian. But that kind of calculus is morally meaningless. It would in principle be acceptable for all 116 civilians to have died in a single strike that killed only one terrorist, if that strike prevented an attack on the scale of September 11. On the other hand, even the low end of 64 civilian deaths may have all been unjustified, if those individuals were killed in the process of targeting someone who posed only a minor threat to other innocents. But the Obama administration has released only aggregate numbers, refusing to detail individual strikes. As a result, it’s impossible for the public to assess the morality of the program.
To be sure, the government claims it does the moral math. According to CIA Director John Brennan, any proposal to target an individual “will go through a careful review and, as appropriate, will be evaluated by the very most senior officials in our government for decision.” According to the policy framework made public this month, which sets out an elaborate structure for making targeting decisions, these officials include the president himself, who bears the ultimate responsibility for the strikes.
Yet the administration also reserves the right to conduct its targeted killing program in secret. The president and his advisers may know why they have accepted the deaths of certain civilians or risked killing the wrong person, but they do not share that knowledge with the public, relying instead an abstract ideal of national self-defense. “There’s this idea that people have that if we get the ethics of self-defense just right, then the context in which we make these decisions will work itself out,” says Sari Kisilevsky, a philosopher of law and ethics at Queens College, CUNY. The government can justifiably classify particular pieces of information if their release would do harm. The problem is with treating national security as a permanent structural factor that balances against individual rights and liberties—even the rights of suspected terrorists.
It doesn’t matter that attacks might emanate from other countries where law and order is weak. “Justice resides in upholding people’s legal rights,” says Kisilevsky. “You can’t balance that against security. That’s just what justice is there to protect against.”
There is good reason to worry about the government’s ability to proactively identify suspected terrorists. The FBI twice investigated the Orlando attacker, Omar Mateen, but was incapable of classifying him a threat within the law. The process for identifying individuals for targeted killing is inherently weaker; it operates entirely within one branch of government, in secret, without advocates for the opposition. When threatened attacks may not take place for months, the process of gathering and weighing evidence is vitally important. This, after all, is why courts exist. “People get it wrong all the time,” cautions McMahan. “And that’s a serious objection to the whole practice.”
Part of the problem is in continuing to treat the “war on terror” as a military engagement. “We shouldn’t think of this as war,” says McMahan. “These are murderers who are contemplating murder on a massive scale. If it’s impossible to arrest them and put them on trial, we have to think of them as being like rampaging murderers.”
The police response in Dallas shows clearly that the government does not need secret judicial processes to deal with rampaging murderers at home. The burden is on the Obama administration to prove why, well more than a decade into the war on terror, it needs extreme secrecy to defend against terrorists outside its borders. Treating targeted killing as a law-enforcement problem would mean requiring the government to tell a story, in public, about why it targets particular individuals. And that would involve giving the public much more information than aggregate data about collateral damage.
“We don’t do legal things in secret,” says Kisilevsky. “The legal structure is in part, intrinsically, necessarily a public structure.” For a state to uphold its moral obligations—to be just—it needs to operate in public unless exigent circumstances override it. Some of these situations are familiar: There are good reasons to use undercover cops, for instance, and sealed indictments. But these are exceptions to the rule in a system that is otherwise largely open to public scrutiny.
“Is it morally justified for a government to carry out a policy of targeted killings against proposed threats where we don’t know what the policy is, we don’t know who the targets are, we don’t know what the criteria for the targets are, what evidence they have to carry it out, and what their standards are?” asks Kisilevsky. “Can that be justified for the sake of national security? Even if it makes us safer, it’s hard to see why that would be a more just society where the government has that power.”
Nearly 15 years after September 11, terrorism poses real threats to American public safety but cannot be said to be a permanent existential threat to the nation in a way that might justify waiving some of the government’s fundamental duties. This is in part because the U.S. government takes terrorist threats seriously. From the public picture journalists have provided of the drone program, the administration is careful and deliberative in choosing its targets. To see the alternative, look no further than the other supposed counterterror campaign in Yemen, where the Saudi-led coalition has bombed at least four hospitals through malice or inattention in the past year.
Obama has clearly grappled with the challenge of justifying preemptive strikes on terrorists. The targeted-killing rulebook is “a reminder of what it means to be a morally serious president,” wrote Lawfare’s Benjamin Wittes. But the man’s moral concern has failed to translate to the state. Obama’s drone policy lives in a gray zone between war and peace. Were the “war on terrorism” truly a war, the government would not need intensive internal reviews to target combatants. Were the government enforcing the law, it would need to press public cases against individuals deemed threats. The drone rulebook is only public because of an ACLU Freedom of Information Act request. If the government can’t make a full accounting to its people why it is killing on their behalf, it cannot kill morally.