Let Them Go

Why John Brennan should give a passport and $10K to every Gitmo prisoner.

Why not just let them go?

The fear that detainees at Guantánamo or Afghanistan might "return to the battlefield" if released remains an ongoing feature of debates about U.S. detention policy. The concern isn't entirely frivolous: Some former detainees have taken up arms against the United States after gaining their freedom. But here's my heretical thought of the week: So what? Or, more precisely: Isn't it time to recognize that the dangers associated with releasing detainees might be outweighed by the dangers associated with continuing to hold them?

Under the law of war, states can detain enemy combatants as prisoners of war for the duration of the conflict, in order to keep them off the battlefield. But historically, many states -- including the United States -- have engaged in routine prisoner exchanges during armed conflicts, freeing enemy prisoners in exchange for the return of our own prisoners. (During World War II, we even exchanged some POWs with the Nazis). The long-standing practice of prisoner exchanges implies something we often seem to forget these days: Sometimes, letting bad guys go is much more useful than hanging on to them.

We're reaching that point in the Afghanistan conflict. On Monday, U.S. military authorities formally transferred control of Afghanistan's Parwan detention facility to the government of Afghanistan. The transfer of the facility, which holds some 4,000 people detained by coalition forces during a decade of war, had been delayed and delayed again due to U.S. fears that Afghan authorities would simply release many of the detainees -- who would end up "returning to the battlefield." But as an unnamed U.S. official told the New York Times, there's "a shift that's going on in how the U.S. is looking at what's important.... We have to look at the larger picture: What's the U.S. strategic interest here?"

Right. What's better for the United States after a dozen years of war, and with plans for a large-scale troop withdrawal in 2014: holding on to every last Taliban detainee "just in case," or letting the Afghans figure out what to do with the detainees -- even if it means that some are released and rejoin the Taliban?

Whether the U.S. effort in Afghanistan succeeds or fails surely does not depend on whether a few thousand Taliban detainees return to being Taliban fighters. No one knows for sure how many fighters the Taliban has in the first place, but estimates range from 25,000 to 40,000. Arrayed against roughly 350,000 members of the Afghan National Security Forces (not to mention the remaining U.S. troops), it's hard to imagine that tossing a few thousand Taliban fighters back into the mix will be a decisive factor, even in the unlikely event that Afghan authorities engage in a wholesale prisoner release.

In part, this is because Taliban fighters appear to be a renewable resource: With or without detainees released from Parwan or other facilities, the Taliban seem quite effective in recruiting new young men to serve as cannon fodder. (Most will occupy only low-level roles, and most will not live long.) And the continued detention of thousands of Afghans by the United States plays a role in ensuring a steady stream of Taliban recruits.

That role is impossible to quantify, but difficult to doubt: Afghan President Hamid Karzai has frequently made clear how bitterly his government -- our putative partner against the Taliban -- resented U.S. control over Parwan. Perhaps more tellingly, ordinary Afghans express striking ambivalence about the presence of international forces: Two years running, for instance, three-quarters of Afghans surveyed by the Asia Foundation say they have "some" or "a lot" of fear about encountering international forces.

It's impossible to determine the degree to which fear and resentment of U.S. detention policies drives Taliban recruiting efforts; night raids, air strikes, frightening checkpoint encounters, and the U.S. role in enabling Karzai's corrupt government undoubtedly also play a role in inspiring armed resistance to what many Afghans view as foreign occupation. But it seems reasonable to assume that Afghan unhappiness with U.S. detention policy is part of that picture.

We thus have to weigh the potential costs associated with releasing Afghan prisoners -- some of whom will likely "return to the battlefield" -- against the potential costs of not releasing them. These costs include the distinct possibility that our continued detention of thousands of Afghans could inspire just as many new Taliban recruits.

This logic seems to have finally won out, as evidenced by this week's transfer of authority for the Parwan detention facility to the Afghans. But the fear of recidivism hasn't fully receded: According to the New York Times, the long impasse over Parwan was resolved only when U.S. officials received "private assurances" that Afghanistan would continue to detain those prisoners viewed as most dangerous by the United States.

"As of today, we don't have prisoners," Secretary of State John Kerry said on Monday. "Whatever is occurring here is under the control of the Afghan people." But Kerry forgot to mention that we didn't turn all detainees over to the Afghans: the Washington Post reports that even with the nominal transfer of Parwan to Afghan control, the United States continues to detain several dozen Afghan nationals deemed to pose "enduring security threats," along with a similar number of non-Afghan detainees (from Yemen, Pakistan, and elsewhere).

What will we do with the hundred or so Parwan detainees we're not willing to hand over to the Afghans? No one knows -- any more than anyone knows what we'll do with the more than 150 men who remain in detention at Guantánamo.

As with the remaining Parwan detainees, it's hard to pin down the precise numbers, identities, or status of men still held at Guantánamo, but they are divided into at least three general categories. First, there is a small number of detainees held pending military commission trials. Second, there are several dozen detainees who have been "cleared for transfer" or release, but who continue to languish at Guantánamo either because no country, including the United States, is willing to accept them inside its borders, or because the United States is not satisfied that -- you guessed it -- detainees won't "return to the fight" if transferred, say, to Yemen. Finally, there are several dozen detainees who are being held indefinitely, on the grounds that "evidentiary problems" (read: past torture and/or U.S. anxiety about revealing intelligence sources and methods) make it impossible for them to be put on trial, but they're just "too dangerous" to release.

Just as in the Afghan context, however, fears about detainee recidivism are almost certainly overblown. For one thing, most previously released Guantánamo detainees have not "returned to the battlefield" -- and of those who have, few appear to have posed a direct or severe threat to the United States. In a 2011 analysis, for instance, Peter Bergen, Katherine Tiedmann, and Andrew Lebovich found that U.S. government claims about Guantánamo recidivism rates often lumped together anti-U.S. activities with militant activities not directed at the United States.

As a result, they argue, the true rate of anti-U.S. recidivism was probably one-in-seventeen, not one-in-four as claimed at the time by U.S. government sources. And Bergen and his colleagues don't differentiate between abstract "dangerousness" of post-Guantánamo activities and actual harm caused by recidivist detainees -- solid information is impossible to obtain, but logically, not all the former detainees who seek to harm the United States will in fact achieve their goals. (Bergen and company do note, dryly, that the Guantánamo recidivism rate looks, by any measure, a whole lot better than the recidivism rate for criminals incarcerated in U.S. prisons: One U.S. study found that nearly two-thirds of released prisoners were ultimately rearrested.)

If we were to think rationally about closing Guantánamo -- I know, not likely -- we would evaluate several factors with regard to potential detainee releases. First, we'd ask how likely it is that a detainee would seek to "rejoin the fight" in one way or another. Second, we'd ask what level of hazard would likely be posed by recidivism: Would we simply have created yet another low-level Taliban or al Qaeda operative, or do we believe a detainee, if released, would be in a position to cause truly grave harm to the United States? Third, we'd ask whether we can mitigate any risk of harm (more on this in a moment). And fourth, we'd ask ourselves some tough questions about the dangers of holding onto detainees indefinitely.

Here again, perhaps even more than in the Afghan context, we should weigh the potential dangers of releasing detainees against the potential long-term threat posed to the United States by our own detention policies. As in Afghanistan, it's hard to entirely unravel overseas anger at U.S. detention policy from anger at other U.S. policies: globally, these likely include U.S. drone strikes, past interrogation policy, or other issues. Nonetheless, there's ample reason to believe that U.S. detention policies have incited anti-American sentiment around the globe.

Our government seems generally averse to engaging in the serious cost-benefit analysis of our detention policies I have suggested here. (It seems similarly averse to engaging in such a cost-benefit analysis of targeted killings.) Congress and the executive branch share the blame for this, but we shouldn't be terribly surprised: For a nation that prides itself on hard-headed capitalist realism, we Americans often seem wholly unable to conduct the most basic risk assessments. In other words: We're routinely irrational when it comes to risk perception.

There's a vast literature on risk perception and risk management. In theory, people should compare risks by evaluating both the probability of a negative event and the potential magnitude of its consequences. In practice, people have trouble accurately assessing either probabilities or consequences: we are often overly influenced by the perceived novelty of dangers, for instance. Similarly, we tend to be less troubled by risks that we believe we can control: Thus, we worry more about plane crashes (low probability, but once you're buckled into your miserable coach class seat, you have zero control) than car crashes (high probability, but if you're at the wheel, you're likely to place excessive faith in your ability to avoid dangers).

This suggests, however, another potential basis for reconsidering our collective fear that released detainees will return to the battlefield. Although we tend to overlook it, we have the ability to significantly control and mitigate the risk posed by released detainees. For one thing, we have the ability to closely monitor released detainees, using a wide range of surveillance technologies -- thus drastically reducing the likelihood of nasty surprises.

We also have the ability to dramatically reduce the likelihood that a released detainee will be welcomed back into the fold by his former comrades. The credibility of released detainees is already low, since their former colleagues are apt to assume they've been compromised. We can make their credibility lower still.

So here's my idea: Have CIA Director John Brennan fly down to Guantánamo with a retinue of news media from all over the world. As the cameras roll, Brennan should hand every last Guantánamo detainee a U.S. passport and ten thousand bucks. (For Khalid Sheik Mohammed, and other high-value detainees, double or triple that figure.) Brennan should hug the detainees, apologize for the inconvenience caused by 10 years in detention, and thank them profusely for everything they've done to help the United States eliminate al Qaeda and its associates.

And then...we should let them go wherever they want. Yemen? Pakistan? Sure, we'll fly them there first class (and monitor every breath they take, every step they take, every call they make, and so on -- wouldn't it be useful to see who they contact?).

With a send-off like that, we can be pretty sure of one thing. Some detainees may want to return to the battlefield...but the battlefield won't be wanting them back.


National Security

A Drone of One's Own

Predators could be useful, but I've settled for a Roomba.

Although I have been a critic of U.S. "targeted killings" policy, I have nothing against drones as such. In fact, I am the proud owner of several drones.

The first drone -- a small helicopter -- entered my life fortuitously. One day, I opened the front door to see my neighbor across the street merrily piloting a helicopter around the roof of his house. My eight-year-old and I oohed and aahed as he skillfully steered the tiny rotary wing drone with a small remote control, and a few days later, he knocked on the door and presented my daughter with a drone of her own.

Not a weaponized drone, I hasten to add -- the little chopper flies around, but that's it. No payload. And that's a good thing, since my two children took to the new drone instantly, soon using it to buzz the windows of nearby houses. When I took the controls, the drone made an immediate crash landing, causing significant collateral damage to a flowerpot. The children revoked my drone piloting license.

I ordered the second drone myself, since once one child has a drone, every child needs one. The second drone -- purchased from Amazon.com -- matched the first, minus the flower-pot induced damage. 

In a retro move, we went on to purchase a fleet of remote control cars to keep the unmanned aerial vehicles company. Then we bought a giant remote-controlled flying goldfish, because why not? Soon the cars, the goldfish, and the helicopters were all active at once, making it difficult to maneuver through the dining room.

But we weren't done yet. "A woman must have money and a room of her own if she is to write fiction," wrote Virginia Woolf in A Room of One's Own. A room of one's own is all very well, but when it comes to warding off the importunate demands of domesticity, a drone of one's own is manifestly better.

Think how pleasant it might be to have an obedient drone (preferably weaponized) hovering nearby at all times! Imagine how this would alter ordinary domestic conversations:

"Honey, I need you to take care of the kids while I work on my column."

"But I was going to watch the game on TV."

"Honey, my Predator drone and I need you to take care of the kids while I work on my column."

Weaponized Predator drones are currently impractical for household use: Buying one would exhaust the children's college fund, and anyway, my backyard's not large enough for a runway. For now, I've settled for a Roomba. It doesn't fundamentally alter the domestic balance of power, but at least it does the vacuuming.

As my one-household drone arms race escalates, I started to ask the important question: Tell me how this ends! Leave the remote-controlled cars and the Roomba aside: how many drones can one household own, and at what level of menace, before some tedious law gets in the way? What legal regulations govern the use of unmanned aerial vehicles by private citizens?

Not many, it turns out. The general category of "things that fly but are not birds" is regulated by the Federal Aviation Administration. Thus, just like grown-up airplanes, private UAVs are regulated by the FAA. At the moment, however, the FAA is mostly MIA when it comes to personal UAVs. The only operative guidelines are the FAA's Model Aircraft Operating Standards, which date back to 1981 and are purely "advisory" in nature. In and of themselves, the standards don't place any legally enforceable limits on the size or operation of personal drones, but merely suggest that in the interest of "a good neighbor environment," UAV pilots should keep all drones under an altitude of 400 feet, avoid flight in "noise sensitive areas" such as schools, hospitals, and churches, and give right of way to "full scale aircraft."

These standards were designed for hobbyists flying model planes, and despite dramatic technological change since they were first developed, they remain, as best I can tell, the only standards specifically applicable to private, non-commercial UAVs.

There are far more stringent rules governing both government agencies wishing to use UAVs and those who wish to use UAVs for commercial purposes. I had hoped to persuade Foreign Policy magazine to purchase its own large UAV and fly it around Washington trailing a large advertising banner with my name on it, but the FAA prohibits the unauthorized business use of drones. If you want your drone to be part of a money-making enterprise, you'll need to go through a cumbersome process to gain FAA approval. (This requirement has already been the downfall of some entrepreneurial real estate photographers who used drones to take aerial photos of properties on the market.) Police, the military, and other public agencies must also get FAA approval before using UAVs inside the United States. But you and me and the guy across the street? Our private drone use is currently exempt.

This isn't to say that you can buy any drone at all: In addition to the limits imposed by your bank account, certain specific technologies and products were developed solely for specific government clients and are currently restricted. The same is true of certain weapons and imaging technologies, which cannot be privately purchased.

But this doesn't pose much of a limit to the size or potential lethality of your personal drone. The technologies relating to UAVs are no easier to put back into bottles than any other genies, and there's currently nothing to stop a mayhem-minded citizen from creating his own fairly sophisticated weaponized UAV. You can buy toy drones for under a hundred bucks. Attach a tiny camera and a small remotely controlled explosive charge and you can, at a minimum, cause havoc in your immediate neighborhood. If you're technologically savvy and you have a bigger budget, you can escalate: Buy a bigger drone, or an even bigger one, and attach legally available weapons, and you're in excellent shape to start a small armed conflict.

As long as you stick to components that are legal to purchase, there are no apparent legal barriers to creating your own weaponized UAV (though it remains an idiotic and antisocial idea). Using your weaponized drone is another story: All the usual criminal laws apply. So don't even think about it.

And except in uncontested airspace, small drone wars are not easy to sustain; even sophisticated military drones remain fairly vulnerable to basic anti-aircraft systems. They can be shot down; their computerized control systems can be hacked. This is still more true for do-it-yourself drones manufactured by even the most malevolent hobbyists.

In this sense, UAVs don't pose unique new problems. They just offer additional (and often cheaper) ways to pose common old problems. Neighbors who stockpile assault weapons are scary, whether they add drones to the arsenal or not. Those who unlawfully use weapons to intimidate or kill are criminals, whether they do their killing with slingshots or hand-made explosives dropped from personal UAVs. By and large, the same is true of UAVs equipped with surveillance cameras and electronic eavesdropping devices: If the paparazzi are after you, there are already plenty of legal ways for them to make your life miserable. The same goes for your nosy neighbor and for paranoid moms with nanny-cams.

None of this keeps most of us from drone-related freak-outs, of course. In a recent poll, 60 percent of Americans reported concerns that government use of drones would invade their privacy. Incredibly, 47 percent of Americans believe they should have a right to shoot down a drone flying over their property and taking pictures. (My Roomba shows no inclination to go airborne, but just to be safe, I've ordered it to stay indoors).

Widespread angst about drones will almost certainly lead to tighter regulation in the next few years. The FAA has been charged by Congress with finding a way to integrate both privately owned and government-owned UAVs into the national airspace by 2015, which will require the agency to develop rules relating to the licensing and use of UAVs. There will likely be more stringent rules for hobbyists as well.

Meanwhile, several states are already considering drone-related legislation, though given FAA preeminence when it comes to regulating airspace, it's an open question whether states have the legal authority to regulate drones themselves. (Does the airspace at knee-level count as airspace, wonders US News?) Some scholars, such as the University of Washington's Ryan Calo, even argue that drones may end up being the best thing that ever happened to privacy law. Until recently, it was mostly the left and libertarians who worried about the rise of the surveillance state. Thanks to drones, these anxieties have gone mainstream.

Not all drone uses are nefarious, of course: UAVs also have great potential to be used "for good" -- finding missing persons, spotting early signs of impending atrocities, assisting during natural disasters, and so on. Ironically, some journalists are using personally owned surveillance drones to keep an eye on government drone bases. As we move towards a more coherent legal and regulatory regime for UAVs and the behaviors they enable, we'll need to make sure we preserve the ability to use drones for the public good.

Drones are the new "in thing" in the world of policy-wonkery, too. Consider political science pundit extraordinaire Francis Fukuyama, who is building his own drone (and blogging about it). This is the man who once prophesized "the end of history."

Really, what more do you need to know?