Thinking about Recidivism

By Benjamin Wittes
Friday, December 10, 2010, 11:28 AM

Sen. Lindsey Graham has called for a halt to transfers from Guantanamo in light of the recidivism rate reported by the Director of National Intelligence. A few thoughts on this general subject:

First, we have a wholly underdeveloped political vocabulary for discussing Gitmo recidivism. We have no social agreement--and very little serious discussion--about what constitutes an acceptable rate of recidivism. I will go further: As Eric Holder might say (in a different context), we are a nation of cowards who lack the guts to have this conversation, which would force us to confront some very hard questions. For example, if someone is probably--but not certainly--a combatant and thus lawfully detainable and he may well--but will not certainly--kill civilians at some point following his release, how do we value his liberty interests against civilian lives? What about if we are reasonably confident that he won't kill Americans but less confident that he won't kill Afghan or Iraqi civilians? We have no useful vocabulary for such questions. The human rights community, to its undying shame, has a relentlessly dishonest way of calculating the human rights cost-benefit ratio of releases; it is crude, simple, and endlessly favoring of the freedom of combatants over the lives of civilians. Detention is a human rights bad, in this worldview, a government action that impinges on liberty. Releases, by contrast, are a human rights good. And when they result in human death, they do not get evaluated as government actions that impinge on liberty. In other words, the human rights community calculates only the benefits of releases and only the costs of detention. In fact, when people released from Guantanamo go on to kill people, the human rights community assumes as a default matter that their radicalization took place as a result of their detention, rather than the other way around. (This is an overgeneralization, but not much of one.)

By contrast, the political right opposes releases. It doesn't matter that some people are being ordered released by the courts or that other people can be freed to the supervision of governments with a fairly good track record of watching their people. It doesn't matter that some detainees are very low-grade. It doesn't matter that some recidivism is inevitable. It doesn't matter that, as the Bush administration concluded quite early on, it just isn't plausible to detain everyone forever. (This is an overgeneralization, but not much of one.)

The result is that we have a debate that willfully blinds itself to the simple reality with which the executive branch has to grapple every day: We are going to make mistakes in detention, and some of them are going to be tragic, and we are going to make mistakes in releasing people from detention, and some of them are going to be tragic.

Second, whether the reported rate of recidivism is acceptable or not depends in large measure on the average and outlying manner of detainee reengagement with the enemy. The fact that 25 percent of released Guantanamo detainees are suspected of reengaging (a figure that includes 13.5 percent who are reportedly confirmed to have reengaged and 11.5 percent who are suspected of having reengaged) by itself conveys relatively little. If, by and large, this group has rejoined enemy insurgent forces as foot soldiers (who are a dime a dozen, anyway), the marginal impact is very different than if it has, by and large, become leadership cadre. If released detainees are, by and large, organizing major operations, that's different from if they are joining up with existing militias. From the perspective of U.S. policymakers, though they are unlikely to say so, it probably also matters whom the renewed activities of detainees threatens. If they are linking up with AQAP and trying to bomb airplanes headed our way, that's rather different from if they are strapping suicide vests onto themselves in Iraq, and both different from if they're becoming Taliban foot soldiers. As a practical matter, we know that released detainees have gone on to a range of unfortunate activities, encompassing all of these things, but the balance matters a lot.

Third, it is a mistake for people on the left to focus on the difference between the rate of recidivism for the Bush administration's releases and the rate for the Obama administration's releases. Adam Serwer writes:

Here are the recidivism rates for detainees released during each administration:

Bush: Confirmed and suspected: 27.2% - Confirmed 14.8% (79 of 532); Suspected 12.8% (68 of 532)

Obama: Confirmed and suspected: 7.5% - Confirmed 3% (2 of 66); Suspected 4.5% (3 of 66)

Bush released a much larger number of detainees than Obama, but given that the Bush administration did an exceptionally poor job of handling the information it had gathered on those it was detaining while publicly declaring them all to be the "worst of the worst," it's not surprising that Obama's numbers are lower. He did, after all, create an entire task force to evaluate the evidence against each detainee as part of his effort to close Gitmo.

The Obama administration's numbers will go up. I feel altogether confident of that. They may not go up quite as high as those for the earlier releases, but I doubt that this is chiefly for the reason Adam identifies. Rather, it is because the Bush administration engaged in large bulk releases of detainees based on their countries of origin--most importantly, to Saudi Arabia. These created a lot of opportunity for system failures. Had the Obama administration's attempt at something similar for Yemen succeeded, its recidivism rate would be a lot higher too. I think the bulk releases were the right thing to do, but they will tend, I suspect, to jack up the recidivism rate. The other important reason for the differential, I suspect, is that the people the Bush administration released have had longer periods of time to reengage, and our intelligence has had more time to find out about it too. As the DNI's report puts it,

The Intelligence Community assesses that the number of former detainees identified as reengaged in terrorist or insurgent activity will increase. A February 2010 review of GTMO detainees’ release dates compared to first reporting of confirmed or suspected reengagement shows about 2.5 years between leaving GTMO and the first identified reengagement reports. Based on trends identified during the past 6 years, the Intelligence Community further assesses that if additional detainees are transferred from GTMO, some of them will reengage in terrorist or insurgent activities.
In other words, look for the gap to close as more recently-released detainees show their true colors over time.
Fourth, it is impossible to have a serious discussion of releases without a serious conversation of detention itself--and both are impossible as long as we remain in denial about detention. This is the major theme of my new book, Detention and Denial: The Case for Candor After Guantanamo, which should be available any day. The concluding paragraphs read as follows:
In its last few years in office, the Bush administration made a great investment in a Saudi rehabilitation program for jihadists, sending home the overwhelming majority of Saudi detainees. A few years later, the percentage of former detainees who have returned to the fight seems relatively low, though the exact numbers are fuzzy. That comparatively small number, however, includes a few who have gone on to leadership positions in a new terrorist group, al Qaeda in the Arabian Peninsula—the group responsible for the Christmas Day bombing attempt. So stunted is our debate over detention that we have no useful vocabulary for discussing whether that pattern represents a success or a failure. Assuming for a moment (probably counterfactually) that all of the Saudis that we released were properly detained, is it reasonable to keep 120 combatants locked up indefinitely to prevent two  or three from doing something horrible that creates a new strategic threat for the United States? Conversely, is it irresponsible to take risks with civilian lives by releasing combatants who have some small percent chance of killing civilians? To avoid a serious debate about detention means to refuse to address such questions. It is not a morally serious position.

In addition, it leaves us especially vulnerable to recriminations after the fact. At the end of the film The Usual Suspects, to end where I began, the master criminal—who has pretended throughout the movie to be a cripple—limps out of the police station, his bad arm and hand stiff and useless. Down the street, the limp fades and the hand relaxes to normalcy, and he gets into his  lawyer’s car and drives off. Inside the police station, Agent Kujan comes to the blinding realization that he has been duped by a man whom he thought was stupid but who had been selling him a bill of goods since they met. He races outside, but it is too late; the agent is left with only self-recrimination, frustration, and anger.

That is us, too, when former Guantánamo inmates turn out to be among the senior leadership of al Qaeda in the Arabian Peninsula. Yet unlike in the movie, which portrays Agent Kujan as the fool, we have no accountability when our system fails. Were these releases the fault of the courts (whose threats of review spurred them), the Bush administration (which carried them out), the Saudi government (which didn’t keep track of former detainees adequately), or the left and the international community (which relentlessly pushed for them)? The less responsibility we take for detention, the less accountability there is when it goes wrong, as it most certainly will—when we lock up the wrong guys or release the wrong guys, when we jail Chinese Uighurs or release suicide bombers. It’s long past time for us to face the reality of the project in which we are engaged.

Fifth, for those who have not seen The Usual Suspects, the relevant scene is here: