The Real D.C. Circuit Habeas Record

A few days ago, the New York Times editorial page made a remarkable claim: "In the 19 [Guantanamo habeas] appeals [the D.C. Circuit] has decided, the court has never allowed a prisoner to prevail." As I pointed out at the time, the claim is wrong. The government's win-loss record before the appeals court cannot reasonably be described as perfect without grossly manipulating numbers. But the Times figures raise two interesting questions: Where does this 19-to-0 figure come from? And what would a reasonable accounting of habeas wins and losses look like? Let's start with the question of how one could derive a 19-to-0 record and why it is a dishonest way to tabulate the government's performance in habeas cases before the D.C. Circuit. Broadly speaking, a habeas merits case will come to the D.C. Circuit in one of two postures: with the writ having been granted by the district court (the government having lost), or with the writ having been denied (the government having won). And broadly speaking, the appeals court can do one of several things with that case. It can affirm the judgment. It can reverse the judgment. Or it can reverse or vacate the judgment but also remand the case for further proceedings subject to some new standard that it sets. Thus, the government can win by having a writ denial affirmed; or it can win by having a writ grant reversed. And more ambiguously, it can win by having a writ grant reversed but remanded. This is not a full win, since the detainee in this situation lives to fight another day in the district court. Conversely, the government can lose by having a writ denial reversed, by having a writ grant affirmed, or--more ambiguously, by having a writ denial reversed but  remanded, but in this latter situation, the loss is not complete as the government gets to fight another day in the district court. The only way to count 19 wins and zero losses for the government, at least as I look at the data, would be to count all remands in which a habeas grant is reversed and remanded as a win for the government but not to count government wins reversed and remanded as losses. As I count, the government has had 12 district court denials of the writ affirmed by the Court of Appeals. It has also gotten three writ grants reversed outright. In three additional cases, writ grants have been reversed and remanded for further proceedings. And in one case, the D.C. Circuit affirmed a denial of the writ in part but remanded the case for further fact-finding on a key point (the district court later ruled in the government's favor on that point). If we count all of these cases as simple government wins, we get 19. The trouble for the Times, is that if we are going to count government wins that way, we should count government losses that way too. In other words, we should count as a government loss any case in which the D.C. Circuit reverses a habeas denial by the district court, even if the reversal is accompanied by a remand in which the government lives to fight another day in district court. There is such a case, as I pointed out in my earlier post: Bensayah. That would make the score 19-to-1. And if we are not counting remand cases like Bensayah in the government's loss column, we shouldn't count remand cases in the win column either. If we exclude remand cases, the government's win total would fall to 15. This may sound like a pretty minor adjustment, but I'm only getting started. Because the problems with the 19-to-0 figure go deeper. Let's consider some other things it leaves out. First, there are those 17 Uighurs. Their victory in Parhat was not quite a habeas case, but that didn't stop people at the time the so-called "scorecard" was so embarrassing for the government from considering Uighurs among the dozens of government habeas losses. Why then doesn't Parhat count as a government loss now? Indeed, as it controls 17 separate dispositions, why doesn't it count as 17 losses? Second, there's a large group of other cases where the government chose not to appeal district court losses. Some of those cases the government probably could have won at the court of appeals, but it's worth remembering that the D.C. Circuit has seen only those habeas cases in which the government either won at the district court level (and the appeal was thus brought by the petitioner) or thought it had a strong chance of prevailing. Both of these groups of cases strongly bias the sample of D.C. Circuit cases toward those in which the government would tend to win. So how should we count the government's record in habeas cases? I would do it as follows--a structure that is loosely based on our earlier alternative to the scorecard metrics but adapted to focus a little more precisely on the way the D.C. Circuit is handling merits cases:

• Uighur cases controlled directly or indirectly by Parhart and requiring detainee release: 17
• District court cases where the writ was granted and the detainee transferred without appeal: 14 (Lahmar, Nechla, Idir, Boumediene, Boudella, Gharani, Basardh, Ali Ahmed, Al Janko, Al Mutairi, Jawad, Al Rabiah, Mohammed, Odaini)
• District court cases where the writ was granted and the appeals court affirmed: 0
• District court cases where the writ was granted and the appeals court reversed outright: 3 (Al Adahi, Uthman, Almerfedi)
• District court cases where the writ was granted and the appeals court reversed but remanded: 3 (Salahi, Hatim, Latif)
• District court cases where the writ was granted and appeal is pending: 1 (Mingazov)
• District court cases where the writ was denied and detainee was released or dropped appeal: 4 (Hammamy, Nahdi, Assani, Khalifh)
• District court cases where the writ was denied and the appeals court affirmed: 12 (Al Bihani, Awad, Barhoumi, Al Odah, Al Bihani, Esmail, Madhwani, Al Alwi, Khan, Al Kandari, Sulayman, Al Sabri)
• District court cases where the writ was denied and the appeals court affirmed but remanded: 1 (Warafi)
• District court cases where the writ was denied and the appeals court reversed outright: 0
• District court cases where the writ  was denied and the appeals court reversed but remanded: 1 (Bensayah)
• District court cases where the writ was denied and the appeal is pending: 5 (Sliti, Obaidullah, Razak Ali, Khairkhwa, Hassain)
• Cert petitions from detainee petitioners granted: 0
• Cert petitions from detainee petitioners denied: 11 (Adahi, Awad, Al Bihani, Al Odah, Uthman, Al Alwi, Latif, Al Kandari, Al Madhwani, Al Bihani, Almerfedi)

When you look at the data this way, one fact stands out very strongly: We are not really arguing over 19 cases. Surely, for example, we are not blaming the D.C. Circuit for, say, the 12 cases in which all it did was affirm a lower court decision to deny the writ. The issue, rather, is really about the six cases in which the lower court would have granted the writ and the D.C. Circuit reversed. While I don't disagree that those six cases represent a pretty dramatic reorientation of the lower court on Guantanamo habeas matters, the merits of those six cases differ rather significantly. Putting aside certain tonal issues in Al Adahi, for example, I have no problem with Salahi, Al Adahi, Hatim, and Uthman. Almerfedi is harder. And Latif seems to me wrongheaded. But all of a sudden we're back to talking about law and cases, not numbers. This is suddenly not a discussion of 19-to-0 but a discussion of which of a small group of truly important cases one thinks was rightly decided and which one thinks was wrong. That's a story the numbers can't tell.