President Obama has begun to address these protests and concerns, to some extent. (This may be why, as of late May, the United States had launched only 13 drone strikes in Pakistan in 2013.) Still, some of the protests are more valid—and some of Obama’s actions less responsive—than others.
An Arrogant Sort of Warfare
The most common criticism of drone strikes is that even when they’re aimed at military targets (terrorists, insurgent safe houses, etc.), they often wind up killing civilians. This is true, but it’s hardly unique to drones. In fact, drones cause far fewer civilian casualties than other kinds of air strikes. The weapons they carry are very small and accurate. The laser-guided Hellfire missile and GPS-guided Small Diameter Bomb land within a few feet of their targets and explode with the force of a mere 30 to 100 pounds of TNT. Aerial bombs in the past have been much larger and far less accurate.
Peter Bergen of the New America Foundation, who has made a thorough study of the publicly available data, estimates that from 2004 to mid-May of 2013, drone strikes killed between 258 and 307 civilians in Pakistan. That’s 7 to 15 percent of the total fatalities caused by drones in the country. Civilian fatalities in Yemen are harder to estimate, but they seem to make up about 8 percent of a much smaller total death toll. These are hardly numbers to wave away casually, but the weapons of a generation ago would have killed many more.*
And yet seen from a different angle, this comparison is nearly irrelevant, and the numbers appear to be quite high. For when we talk about accidental civilian deaths by drones in Pakistan and Yemen, we are talking about countries where the United States is not officially fighting wars. In other words, these are countries where the people killed—and their embittered friends and relatives—didn’t know that they were living in a war zone. Imagine that Mexican commanders launched an air strike on a border town in California because their enemies were hiding there and that, as a result of poor aim or bad intelligence or dumb luck, a few dozen American citizens were killed. The American people and the U.S. government would be outraged, and justifiably so.
Drone strikes are criticized as an arrogant sort of warfare. The whole idea of killing people from far away, invisibly and without risk of retaliation, seems somehow unfair. But the same was said when the British and Americans dropped bombs from airplanes in World War II. It was said when British archers used longbows against French knights. It’s natural for armies to find ways to maximize the enemy’s losses while minimizing their own.
It turns out that most of the people killed by drones are not al-Qaeda leaders. Often they’re not affiliated with al-Qaeda at all.
Still, these comparisons don’t quite fit. Drones are different, because of where they are used. Stanley McChrystal, a retired general who relied heavily on drone strikes when he was special-ops chief in Iraq and commander of all NATO forces in Afghanistan, put it this way in a recent interview with Reuters: “The resentment caused by American use of unmanned strikes … is much greater than the average American appreciates. They are hated on a visceral level, even by people who’ve never seen one or seen the effects of one.”
This isn’t a speculative matter. In April, at hearings before the Senate Judiciary Committee (the first public hearings on the consequences of drones), Farea al-Muslimi, a Yemeni activist and journalist, testified about a drone strike in his native village just a week earlier. Before the strike, al-Muslimi said, the villagers had a positive impression of the United States, drawn mainly from conversations with him about the year he’d spent here during high school, which he described as “one of the best years of my life.” But now, he went on, “when they think of America, they think of the terror they feel from the drones that hover over their heads, ready to fire missiles at any time.”
In a conventional war, this might be a regrettable side effect. But in the kinds of wars the United States has been fighting lately, in Yemen and elsewhere, it feeds into the main effect. These are wars against guerrillas, insurgents, terrorists, rogues, fought not only to kill the enemy but to influence the population (to “win hearts and minds,” as the old saying had it). If the most prominent weapon in this war alienates the people who live under its shadow—in some cases driving them into the arms of the enemy, either for protection or on the principle that the enemy of their enemy is their friend—then it is a lousy weapon. Retired general David Petraeus, in his 2006 U.S. Army field manual on counterinsurgency, made a similar point: “An operation that kills five insurgents is counterproductive if collateral damage leads to the recruitment of 50 more insurgents.”
Even so, as Petraeus noted, sometimes a commander has to fire the weapon regardless of the possible backlash; sometimes the target is too important, the threat too dangerous, to pass by. But here we come to another source of controversy about drones. As the strikes have evolved over the years, fewer and fewer of their targets have posed a genuine threat to the United States. In more and more instances, the targets of drone strikes are low-level militiamen, not terrorist leaders. In a striking number of cases, they are targeted for death even though their identities—their names, ranks, and the scope of their involvement in a terrorist organization—are unknown.
More and more, the drones are used for “signature strikes.” The officer or official approving a strike might not know who its targets are, but their behavior—as picked up by drone cameras, satellites, cell-phone intercepts, spies on the ground, or other “sources and methods” of intelligence agencies—strongly suggests that they’re active members of some organization whose leaders would be the natural targets of a drone strike. For instance, they might be moving in and out of a building that’s a known terrorist hangout, or they might be training at a known terrorist facility. In other words, their behavior bears the “signature” of a legitimate target.
Neither the Bush nor the Obama administration has ever confirmed the existence of signature strikes. (Like all CIA drone strikes, they are highly classified.) But one knowledgeable official told me that in Pakistan, the “vast majority” of drone strikes have been signature strikes—from the very beginning up until now.
There seems to be no formal list of the criteria that a suspected terrorist must meet before he can be targeted by a drone. Nor is there some quantitative technique for measuring an official’s degree of confidence in this signature. Those who pick the targets have a database of correlations between certain types of behavior and the presence of terrorist leaders. But it’s a judgment call, and there’s usually no way—or desire—to check afterward whether the judgment was good or bad. The practice evolved gradually from tactics in Iraq and Afghanistan. It made sense in a war zone. An officer sees a sniper on a rooftop, or someone planting an IED along a road, or armed men moving in and out of a known bomb factory. Almost certainly, they’re enemy combatants in a war. He doesn’t need to know their names; nor does it much matter whether they’re killed by a bullet, a mortar, a smart bomb from a helicopter, or a Hellfire missile from a drone.
But outside a war zone, such questions do matter. Attacks in those areas amount to assassinations—which, besides the political backlash they may inspire locally, are prohibited by U.S. and international law.
President Obama is aware of this; he was trained as a constitutional lawyer. In a speech on national security on May 23, he laid out three conditions that must be met before a drone strike can be approved. He said it must be determined that the target poses a “continuing, imminent threat” against the United States; that capturing the person alive is infeasible; and that there is “near certainty” that the strike will kill or injure no civilians.
These conditions were nothing new. They came from a 16-page Justice Department white paper that was leaked to the press in February. The white paper’s legal rationale was full of holes and evasions, and so was the speech it inspired.
The white paper’s main sleight of hand was to define the terms in such a way that the most basic fact about these attacks—that they’re conducted outside a war zone—is denied. To this end, it cites the Authorization for Use of Military Force, a joint resolution passed by Congress on September 14, 2001 (three days after the terrorist attacks on the World Trade Center and the Pentagon). Under the AUMF, the president may use all necessary and appropriate force against those nations, organizations, or persons he determines planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001, or harbored such organizations or persons, in order to prevent any future acts of international terrorism against the United States by such nations, organizations or persons.
This language is strikingly broad. Nothing is mentioned about geography. The premise is that al-Qaeda and its affiliates threaten U.S. security; so the president can attack its members, regardless of where they happen to be. Taken literally, the resolution turns the world into a free-fire zone.
The white paper then lays down the same three conditions that Obama later recited—ostensibly to impose restrictions on otherwise sweeping executive authority. In fact, they restrict nothing. Key to this legalistic gamesmanship is the paper’s definition of “imminent threat.” It states:
The condition that an operational leader [of al-Qaeda or an affiliated organization] presents an “imminent” threat of violent attack against the United States does not require the United States to have clear evidence that a specific attack … will take place in the immediate future.
In other words, “imminent,” in this context, does not mean imminent.
The paper’s logic is that leaders of al-Qaeda and its affiliates are “continually planning attacks” against the United States. “By its nature, therefore,” the threat demands “a broader concept of imminence.” That is to say, the threat of an attack is constant; it is always vaguely imminent, even if there are no signs of an actual attack. And so the first condition that must be met for a targeted assassination—an imminent threat of attack—is not a restriction in any real sense.
The second condition—that it must be infeasible to take the terrorist alive—is equally meaningless. Because the threat of attack is always imminent, the United States is likely to have “only a limited window of opportunity” for mobilizing a raid on the ground. By this standard, it is always infeasible to capture a terrorist. Therefore, once he is found, it is necessary to kill him with a drone strike. Again, it’s a test that, by design, cannot be failed.
Lax as these standards are, the United States has not lived up to them. For it turns out that most of the people killed by drones, in places like Yemen and Pakistan, are not al-Qaeda leaders. Often they’re not affiliated with al-Qaeda at all.
*Civilian casualties are a touchy subject, notoriously difficult to estimate. Other private groups, drawing on data similar to Bergen’s, come up with different numbers. (Bergen also notes that it’s “unknown” whether another 196 to 330 fatalities were civilian or military. Assuming they were all civilians, that would be 14 to 32 percent of total fatalities.) The London-based Bureau of Investigative Journalism puts civilian fatalities at 461 to 884, or 16 to 25 percent. The Long War Journal puts it at 153, or 5.7 percent. There are two very different perspectives in which to consider these estimates. On the one hand, Rosa Brooks, a law professor at Georgetown University, cites Red Cross data to show that on average, the wars of the 20th century produced 10 civilian fatalities for every combatant killed. By that standard, drone strikes are remarkably sparing. On the other hand, Bergen reports that only 55 militant leaders were killed in all the U.S. drone strikes in Pakistan in that nine-year period—accounting for barely 2 percent of all fatalities caused by drones. Therefore, by the standard of how drones are to be used outside of war zones—as set by U.S. and international law—the level of deaths is remarkably high.