As I noted yesterday, both Dianne Feinstein and Carl Levin understand Section 1031 of the Defense Authorization to authorize the indefinite detention of American citizens. Levin says we don’t have to worry about that, though, because Americans would still have access to habeas corpus review.

Section 1031 makes no reference to habeas corpus, and places no limitation on habeas corpus review.  Nor could it.  Under the Constitution, habeas corpus review is available to any American citizen who is held in military custody, and to any non-citizen who is held in military custody inside the United States.

Even ignoring the case of Jose Padilla, which demonstrates how easily the government can make habeas unavailable to American citizens, there’s another problem with Levin’s assurances.

Habeas was gutted on October 14, when Janice Rogers Brown wrote a Circuit Court opinion holding that in habeas suits, judges must grant official government records the
presumption of regularity.

The habeas case of Adnan Farhan Abdul Latif largely focused on one report purporting to show that Latif fought with the Taliban. I suspect the report is an early 2002 CIA report, written during the period when the US was trying to sort through hundreds of detainees turned over (sometimes in exchange for a bounty) by the Pakistanis. The report I suspect is at issue summarizes the stories of at least 9 detainees, four of whom have already been transferred out of US custody. David Tatel’s dissent makes it clear that there were clear inaccuracies in the report, and he describes Judge Henry Kennedy’s judgment that this conditions under which this report was made–in the fog of war, the majority opinion agrees–increased the likelihood that the report was inaccurate. Of note, Latif’s Factual Return reveals the government believed him to be Bangladeshi until March 6, 2002 (see paragraph 4); they blame this misunderstanding on him lying, but seeing as how the language of an interrogation–whether Arabic or Bangladeshi–would either seem to make his Arab identity clear or beset the entire interrogation with language difficulties, it seems likely the misunderstanding came from the problem surrounding his early interrogations.

Beyond that report, the government relied on two things to claim that Latif had been appropriately detained: The claim that his travel facilitator, Ibrahim Alawi, is the same guy as an al Qaeda recruiter, Ibrahim Balawi (usually referred to as Abu Khulud), in spite of the fact that none of the 7 detainees recruited by Balawi have identified Latif. And the observation that Latif’s travel to Afghanistan from Yemen and then out of Afghanistan to Pakistan traveled the same path as that of al Qaeda fighters (here, too, none of the fighters who traveled that same path identified Latif as part of their group).

In other words, the government used one intelligence report of dubious reliability and uncorroborated pattern analysis to argue that Latif had fought with the Taliban and therefore is legally being held at Gitmo.

And in spite of the problem with the report (and therefore the government’s case), Judge Janice Rogers Brown held that unless Judge Kennedy finds Latif so credible as to rebut the government’s argument, he is properly held. More troubling, Rogers Brown held that judges must presume that government evidence gathering–intelligence reports–are accurate as a default.

When the detainee’s challenge is to the evidence-gathering process itself, should a presumption of regularity apply to the official government document that results ? We think the answer is yes.

Rogers Brown is arguing for a presumption of regularity, of course, for the same intelligence community that got us into Iraq on claims of WMD; the report in question almost certainly dates to around the same period that CIA went 6 months without noticing an obvious forgery.

Rogers Brown’s presumption of regularity is particularly troublesome given that raw intelligence is not meant to be definitive. It is the documentation of gossip and rumor that has not yet been vetted as to whether or not it is fact.

Here’s what Sabin Willett–the lawyer for two Uuighurs, Parhat and Kiyemba–says results from the Court’s decision that judges must accept such reports as definitive.

It is not hyperventilation to say, as so many have said, that Latif guts Boumediene, because — trust me —  every prisoner has an intelligence report.  Now the prisoner hasn’t just lost his judicial remedy to Kiyemba; if those reports control, factfinding is over, too.

[snip]

I tried Parhat.  He had an intelligence report too.  We picked it apart, as I’m sure Latif’s lawyers must have done with their report, and as Judge Garland did in the classified Parhat opinion.  No one could make a straight-faced argument for a presumption after that was done.  You have to–I can’t say this any other way, because Parhat’s documents remain classified–but you have to see an “intelligence report”  to appreciate just how surreal the proposition is.

The trial lawyer would think this way: if this tissue of hearsay, speculation, and gossip comes in evidence at all, the trial court must at least be allowed to weigh it.  But when the circuit lays the thumb of presumption on the scale, there’s no more judicial review — not even in the court of appeals.  “Review” is in the anonymous DoD analyst who wrote the report.

Review was Judge Kennedy’s job, and he did his job.  Whether we agree or disagree with his weighing, the scale had always been his before.  This idea, I think, lies at the bottom of Judge Tatel’s thoughtful dissent.  Can the jailer’s report trump the judicial officer, in civil cases that are supposed to be a check on the jailer itself?   There’s not much evidence that anybody up at SCOTUS cares about the GTMO prisoners any more (whose imprisonments now treble WW2 detentions), but there may still be four of them who worry about trial judges.

[snip]

Pause a moment.  A man sits in government prison for ten years and counting, on the strength of a secret document created by the jailer, in haste, from hearsay, which didn’t persuade an experienced trial judge. Does that sound like the stuff of regimes we are prone to condemn?

And now with some version of 1031 set to pass Congress, this is the standard that courts will use not just with UIghurs and Yemenis picked up in Afghanistan, but potentially with young Muslim American men who sound off in chat rooms. With the presumption of regularity, intelligence reports based on paid informants’ claims about what got said at a mosque will be enough to hold an American citizen indefinitely.

And it’s not just the report. Rogers Brown accepts pattern analysis–which in Latif consisted of travel patterns but which in US-based counterterrorism usually tracks the patterns of the kinds of calls you make, your geolocation, which falafel joint you frequent–as the sole corroboration for the dicey intelligence report.

The way Rogers Brown treats such pattern analysis, in lieu of any real witnesses, as corroboration bodes particularly poorly for the US given how much pattern analysis the government is already doing on innocent Americans.

Carl Levin may well believe his compromise language carries no risk to Americans given the guarantee of habeas, but with Latif as precedent in war on terror habeas cases, he’s wrong. As the senator representing one of the largest communities of Arab-Americans and Muslims in the country, his carelessness on this point is particularly troubling.

While it’s not the primary goal, Levin’s “compromise” language could put some of his constituents–guilty of nothing more than religion, proximity, and gossip–in indefinite detention, with little recourse. And he doesn’t seem all that bothered by the possibility.