Take Two Drones and Call Me in the Morning

The perils of our addiction to remote-controlled war.

Last week, I argued that many common objections to U.S. drone strikes don't hold up well under scrutiny. For the most part, complaints that "drones kill civilians" or "drones kill from a distance" are red herrings -- every weapons system can cause civilian casualties, and planes, Tomahawk missiles, and snipers all enable killing from a distance. We should worry about armed drone technologies for a different set of reasons: By lowering or disguising the costs of lethal force, their availability can blind us to the potentially dangerous longer-term consequences of our strategic choices.

Armed drones lower the perceived costs of using lethal force in at least three ways.

First, drones reduce the dollar cost of using lethal force inside foreign countries. Most drones are a bargain compared with the available alternatives. Manned aircraft, for instance, are quite expensive: Lockheed Martin's F-22 fighter jets cost about $150 million each; F-35s are $90 million; and F-16s are $55 million. But the 2011 price of a Reaper drone was $28.4 million, while Predator drones cost only about $5 million to make. (And Hellfire missiles are a steal at less than $60,000 each; you could buy one with a home equity loan.)

Second, relying on drone attacks reduces the domestic political costs of using lethal force. Sending special operations forces after a suspected terrorist places the lives of U.S. personnel at risk, and full-scale invasions and occupations endanger even more American lives. In contrast, using armed drones eliminates all short-term risks to the lives of U.S. personnel involved in the operations. (And because drone attacks don't involve "sustained fighting … active exchanges of fire … [or] U.S. ground troops," any need for congressional notification and approval under the War Powers Resolution can conveniently be avoided.)

Third, by reducing accidental civilian casualties, precision drone technologies reduce the perceived moral and reputational costs of using lethal force. Contrary to popular belief, most U.S. officials care greatly about avoiding civilian casualties. (Yes, Virginia, there is such a thing as a conscience, even in CIA officials.) Even those willing to discount the moral cost of civilian deaths understand the reputational costs: Dead civilians upset local populations and host-country governments, alienate the international community, and sometimes even disturb the sleep of American voters.

I don't believe that humans can be reduced to Homo economicus, but as a group, government officials are remarkably sensitive to financial, political, and reputational costs. Thus, when new technologies appear to reduce the costs of using lethal force, their threshold for deciding to use lethal force correspondingly drops.

If killing a suspected terrorist based in Yemen or Somalia will endanger expensive manned aircraft, the lives of U.S. troops, and/or the lives of many innocent civilians, U.S. officials will reserve such killings for situations of extreme urgency and gravity (stopping another 9/11, finally getting Osama bin Laden). But if all that appears to be at risk is an easily replaceable drone, officials will be tempted to use lethal force more and more casually.

And this, of course, is exactly what has been happening over the last four years. Increasingly, drones strikes have targeted militants who are lower and lower down the terrorist food chain, rather than terrorist masterminds. Strikes increasingly target individuals who pose speculative, distant future threats rather than only those posing urgent or catastrophic threats. And drone strikes have spread ever further from "hot" battlefields, migrating from Pakistan to Yemen to Somalia (and perhaps to Mali and the Philippines as well). Although drone strikes are believed to have killed more than 3,000 people since 2004, only a tiny fraction of the dead appear to have been so-called "high-value targets."

These numbers are just estimates culled from news stories, NGO reports, and rumors. Although people tend to notice several-thousand dead bodies, Barack Obama's administration still refuses to openly acknowledge that the CIA uses drone strikes anywhere other than Pakistan (and this was acknowledged only recently and grudgingly).

But though drone technologies enable the United States to reduce some of the costs of using lethal force inside the borders of other states, overreliance on drones may have potentially devastating costs of its own.

For one thing, overreliance on drones reflects a dangerous blurring of the boundaries between "war" and "non-war," with grave consequences for the rule of law. As I wrote in an earlier column, whether one regards drone strikes as lawful acts of war or as extrajudicial killings (murders, in plain English) depends wholly on how far one is willing to stretch the law of war in efforts to make it fit an increasingly unprecedented situation.

Defenders of the administration's increasing reliance on drone strikes outside "hot" battlefields assert that the law of war is applicable -- in any place and at any time -- with regard to any person the administration deems a combatant. Such an assertion wouldn't be troubling if the United States were in a conventional conflict with the uniformed forces of an enemy state. If that were the case, it would be fairly easy for journalists or moderately intrepid citizens to confirm the basic facts justifying government claims about the applicability of the law of war: The presence of thousands of uniformed troops shooting at one another is hard to misconstrue.

But outside Afghanistan, the United States is not in a conventional war. It's in an open-ended conflict with an inchoate, undefined adversary -- and administration assertions about who is a combatant and what constitutes a threat are entirely non-falsifiable because they're based wholly on undisclosed evidence.

In this murky context, it's facile to assert that the law of war "obviously" applies to all U.S. drone strikes and leave it at that. As I wrote on Aug. 29, that amounts, in practice, to a claim that the executive branch has the unreviewable power to kill anyone, anywhere, at any time, based on secret criteria and secret information discussed in a secret process by largely anonymous individuals.

Law exists to restrain untrammeled power. Sure, it's possible to make a plausible legal argument justifying each and every U.S. drone strike -- but this merely suggests that we're working with a legal framework that has begun to outlive its usefulness. The real question isn't whether U.S. drone strikes are "legal." The real question is: Do we really want to live in a world in which the U.S. government's justification for killing is so malleable?

Defenders of administration policy argue that these criticisms miss the mark because -- insiders insist -- executive branch officials go through an elaborate process in which they carefully consider every possible issue before determining that a drone strike is lawful. No doubt they do. But formal processes tend to further normalize once-exceptional activities -- and "trust us" is a pretty shaky foundation for the rule of law.

Here's another reason to worry about the U.S. overreliance on drone strikes: Other states -- and ultimately, nonstate actors -- will follow America's example, and the results won't be pretty.

Right now, the United States has a decided technological advantage when it comes to armed drones, but that won't last long. The country should use this window to advance a robust legal and normative framework that will help protect against abuses by those states whose leaders can rarely be trusted. Right now, the country is doing the exact opposite: Instead of articulating norms about transparency and accountability, the United States is effectively handing China, Russia, and every other repressive state a playbook for how to get away with murder.

Consider Vladimir Putin and Dmitry Medvedev's Russia, in which the life expectancy of dissidents, inquisitive journalists, and unwanted political rivals is already quite short. At the moment, the Russian government at least feels constrained to disclaim responsibility when a troublesome citizen is conveniently murdered. But with the United States putting forward an infinitely flexible interpretation of the law of war, why should Russia bother to cover its tracks in the future?

Far simpler just to shrug off the next dissident's death (whether by drone strike in Chechnya or radioactive sushi in London) with a dignified news release. The dead "dissident"? A combatant in Russia's war with terrorists. The evidence? That's classified, but all actions taken are lawful and subject to a rigorous internal Kremlin review process. You got a problem with that? To quote Obama, "There are classified issues, and a lot of what you read in the press … isn't always accurate.… My most sacred duty … is to keep the … people safe."

Here's one final reason to worry about drone technologies: They enable a "short-term fix" approach to counterterrorism, one that relies excessively on eliminating specific individuals deemed to be a threat, without much discussion of whether this strategy is likely to produce long-term security gains.

Drone strikes -- lawful or not, justifiable or not -- sow fear among the "guilty" and the innocent alike. What impact will they ultimately have on the stability of those societies? To what degree -- especially as we reach further and further down the terrorist food chain -- are we actually creating new grievances? As Defense Secretary Donald Rumsfeld asked during the Iraq war, are we creating terrorists faster than we kill them? And will our increasing use of cross-border strikes have an impact on global stability, undoing fragile post-World War II bargains about sovereignty and the use of force?

As far as I can tell, none of these questions is being discussed within the Obama administration in any structured or systematic way. Meanwhile, U.S. reliance on drone strikes continues to increase. And because so much about U.S. drone strikes is classified, it's almost impossible for journalists, regional experts, human rights groups, or the general public to weigh in with informed views.

There's nothing preordained about how we use new technologies, but by lowering the perceived costs of using lethal force, drone technologies enable a particularly invidious sort of mission creep. When covert killings are the rare exception, they don't pose a fundamental challenge to the legal, moral, and political framework in which we live. But when covert killings become a routine and ubiquitous tool of U.S. foreign policy, everything is up for grabs.


National Security

What's Not Wrong With Drones?

The wildly overblown case against remote-controlled war. 

For many on the political left, and more than a few in the middle, drone strikes are the paradigmatic example of U.S. militarism run amok. I'm not crazy about the way the United States has been using drone strikes myself, but many of the most common objections to drones don't hold up well under scrutiny.

Let's review the case against the drones.

1. Drone strikes kill innocent civilians.

This is undoubtedly true, but it's not an argument against drone strikes as such. War kills innocent civilians, period. But some means and methods of warfare tend to cause more unintended civilian deaths than others.

"Drones scout over [Afghanistan and Pakistan] launching Hellfire missiles into the region missing their intended targets, resulting in the deaths of many innocent people," trumpets the website for Code Pink, a women's peace group. Similarly, the Anti-War Committee asserts that "the physical distance between the drone and its shooter makes lack of precision unavoidable."

But to paraphrase the NRA, "Drones don't kill people, people kill people." At any rate, drone strikes kill civilians at no higher a rate, and almost certainly at a lower rate, than most other common means of warfare.  Drones actually permit far greater precision in targeting. Today's unmanned aerial vehicles (UAVs) can carry small bombs that do less widespread damage, and there's no human pilot whose fatigue might limit flight time. Their low profile and relative fuel efficiency combines with this to permit them to spend more "time on target" than any manned aircraft.

Drones can engage in "persistent surveillance.­" That means they don't just swoop in, fire missiles and swoop out: they may spend hours, days, or even months monitoring a potential target. Equipped with imaging technologies that enable operators even thousands of miles away to see details as fine as individual faces, modern drone technologies allow their operators to distinguish between civilians and combatants far more effectively than most other weapons systems.

That doesn't mean civilians don't get killed in drone strikes. They do.

How many civilians? It depends how you count. The British Bureau of Investigative Journalism analyzed reports by "government, military and intelligence officials, and by credible media, academic and other sources" and came up with a range: the 344 known drone strikes in Pakistan between 2004 and 2012 killed between 2,562 and 3,325 people, of whom between 474 and 881 were civilians. (The numbers for Yemen and Somalia are much squishier.) The New America Foundation, at which I'm a fellow, came up with slightly lower numbers: somewhere between 1,873 and 3,171 people killed overall in Pakistan, of whom between 282 and 459 were civilians.

That means somewhere between 8 percent and 47 percent of Pakistan drone strike victims were probably civilians. Work out the civilian deaths per drone strike ratio for the last eight years, and on average, each drone strike seems to have killed between 0.8 and 2.5 civilians.

These are gruesome calculations: behind the numbers, regardless of which data set is right, lie the mangled bodies of human beings. But whether drones strikes cause "a lot" or "only a few" civilian casualties depends what we regard as the right point of comparison. A study by the International Committee for the Red Cross found that on average, 10 civilians died for every combatant killed during the armed conflicts of the 20th century. For the Iraq War, estimates vary widely; different studies place the ratio of civilian deaths to combatant deaths anywhere between 10 to 1 and 2 to 1.

Compared to the atom bombs dropped on Hiroshima and Nagasaki, drone strikes look pretty good. Compared to world peace, not so much.

The most meaningful point of comparison is probably manned aircraft. It's difficult to get solid numbers here, but one analysis published in the Small Wars Journal suggested that in 2007 the ratio of civilian deaths due to coalition air attacks in Afghanistan may have been as high as 15 to 1. More recent UN figures suggest a far lower rate, with as few as one civilian killed for every ten airstrikes in Afghanistan.

But drone strikes have also gotten far less lethal for civilians in the last few years: the New America Foundation concludes that only three to nine civilians were killed during 72 U.S. drone strikes in Pakistan in 2011, and the 2012 number -- so far -- is zero civilians killed in 36 strikes.  In part, this is due to technological advances over the last decade, but it's also due to far more stringent rules for when drones can release weapons.

2. Drones strikes are bad because killing at a distance is unsavory.

Really? If killing from a safe distance (say, Creech Air Force Base in Nevada) is somehow "wrong," what should be our preferred alternative -- stripping troops of body armor, or taking away their guns and requiring them to engage in hand-to-hand combat? If drone strikes enable us to kill enemies without exposing our own personnel, this is presumably a good thing, not a bad thing. Maybe we shouldn't kill anyone, or maybe we're killing the wrong people -- but these are assertions about ethics, intelligence and strategy, not about drones.

Drones don't present any "new" issues not already presented by aerial bombing -- or by any previous historical method of killing from a distance. In the early 1600s, Cervantes called artillery a "devilish invention" allowing "a base cowardly hand to take the life of the bravest gentleman," with bullets "coming nobody knows how or from whence." (Much like drones.)

The longbow and cross bow were also once considered immoral, for that matter: in 1139, the Second Lateran Council of Pope Innocent II is said to have "prohibit[ed] under anathema that murderous art of crossbowmen and archers, which is hateful to God" -- at least when used against Christians.

3. Drones Turn Killing into a Video Game.

Writing in the Guardian, Phillip Allston (the United Nations special rapporteur on extrajudicial, summary or arbitrary executions) and Hina Shamsi of the ACLU decry "the PlayStation mentality" created by drone technologies. "Young military personnel raised on a diet of video games now kill real people remotely using joysticks. Far removed from the human consequences of their actions, how will this generation of fighters value the right to life?"

But are drones more "video game-like" than, say, having cameras in the noses of cruise missiles?  Those old enough to remember the first Gulf War will recall the shocking novelty of images taken by cameras inside U.S. Tomahawk missiles, the jolting, grainy images in the crosshairs before everything went ominously black.

Regardless, there's little evidence that drone technologies "reduce" their operators' awareness of human suffering. If anything, drone operators may be far more keenly aware of the suffering they help inflict than any distant sniper or bomber pilot could be.

Journalist Daniel Klaidman reports the words of one CIA drone operator, a former Air Force pilot: "I used to fly my own air missions.... I dropped bombs, hit my target load, but had no idea who I hit. [With drones], I can look at their faces... see these guys playing with their kids and wives.... After the strike, I see the bodies being carried out of the house. I see the women weeping and in positions of mourning. That's not PlayStation; that's real."

Increasingly, there's evidence that drone pilots, just like combat troops, can suffer from post-traumatic stress disorder: watching a man play with his children, then seeing his mangled body takes a psychological toll. A recent Air Force study found that 29 percent of drone pilots suffered from "burnout," with 17 percent "clinically distressed."

4. Targeted killings are creepy.

Many critics of drone strikes are also broadly uncomfortable with targeted killings, viewing them as little more than assassinations or simple murder. In targeted killings, lethal force is aimed at specific, named individuals. (Not all targeted killings involve drone strikes, just as not all drone strikes are targeted killings.)

Last week, I wrote that whether targeted killings are lawful depends entirely on whether you think the law of war applies in a given situation. But assuming it does apply -- which is surely true in Afghanistan, at least-- it's hard to see the problem with targeted killing. Should we prefer untargeted killing?

Going after a named, specifically identifiable individual we know to be a bad guy may make us uncomfortable, but it's surely better -- assuming our intelligence is solid -- than, say, lobbing grenades into a compound filled with unnamed probable bad guys.

None of this means we should feel sanguine about the way drone strikes are used by the United States. In next week's column, I'll focus on reasons we should worry about drone strikes. Current U.S. practice presents glaring rule of law problems, and unmanned technologies also provide the executive branch with a new means of circumventing the War Powers Act, increasing the temptation to use force. Perhaps even more ominously, the increased use of drones by the U.S. also risks completely unsettling the precarious collective security structures created by the U.N. Charter.

There's plenty not to like about drone warfare. But if we're going to critique it, we should do so for the right reasons, not the wrong ones.

Ethan Miller/Getty Images