Thoughts on Latif #2--From Peter Margulies

Peter Margulies of Roger Williams Law School writes in with the following critique of the Latif decision and praise of Judge David Tatel's dissent. While I don't agree with every aspect of this analysis, I agree with a great deal of it:

The majority opinion in Latif, which accords a virtually irrebuttable presumption of “regularity” to an intelligence report containing concededly unreliable information, destabilizes an evolving consensus on the D.C. Circuit. Under that consensus, a habeas court had to take a pragmatic, Bayesian view of evidence. When the government presented facts that tended to demonstrate a detainee had been part of Al Qaeda, and the detainee’s story lacked plausibility, a court would deny the detainee’s petition. However, when the government’s proof was lacking, the “meaningful review” guaranteed by the Supreme Court’s landmark decision in Boumediene preserved the possibility of relief for the detainee. As Judge Tatel’s incisive dissent makes clear, Latif appears to repudiate that consensus. In an even more troubling development, the majority strongly implies that its departure from this consensus stems from disagreement with the Supreme Court’s rationale in Boumediene. As Judge Tatel pointed out at p. 9 of his dissent, the intelligence report at issue in Latif, which linked the detainee to Al Qaeda and a known Qaeda recruiter, was assembled hastily in the field, featured “multiple layers of hearsay, depend[ed] on translators of unknown quality, and include[d] cautionary disclaimers” that alerted a reasonable reader that the entire composition process was “prone to significant errors.” These indicia of unreliability contrasted with documents, such as docket sheets, that typically enjoy a presumption of regularity because they contain data that is routinely collected and under the government’s sole control. When those conditions are absent, Judge Tatel noted, a court following Boumediene’s guidance should delve deeper to separate the wheat from the chaff. Indeed, Judge Tatel noted, the D.C. Circuit had previously engaged in careful scrutiny of the government’s evidence. The Latif majority failed in this task. The majority coupled excessive deference to the government’s evidence with a practiced skepticism regarding the narrative of the detainee, a Yemeni national. The detainee claimed that he had left Yemen in 2001 to obtain treatment for head injuries sustained in a 1994 car accident, and had been befriended by a fellow Yemeni, Ibrahim Al-Alawi, whom Latif followed to a religious institute in Kabul. The majority asserted that differences in spelling between the name of Latif’s benefactor and a known Al Qaeda facilitator, Ibrahim Ba’alawi, were simply inadvertent mistakes in translation. This view overlooks the different physical descriptions of the men provided by Latif and other detainees. It also conflicts with the majority’s own rationale for treating the government intelligence report as reliable: If mistakes in translation can occur in a detainee’s statements, they can also occur in intelligence reports. The majority viewed possible mistakes in the detainee’s translated evidence as supporting the government, while ignoring the equal risk of mistakes in the government’s evidence. The majority’s incredulity also extended to Latif’s travel route, which it viewed as evidence of his Al Qaeda ties. However, as Judge Tatel noted, a true Bayesian probability analysis would discount this evidence if many non-Al Qaeda fighters took the same route. Drivers on I-95, Judge Tatel sagely observed, may well include some drug traffickers; however, thousands of entirely innocent motorists take the same route. This was not a case like Uthman, Judge Tatel noted, where the detainee was apprehended with known members of Al Qaeda. That evidence would pass the Bayesian test; the evidence in Latif failed. Admittedly, Latif’s account also included inconsistencies. For example, Latif asserted in pleadings that a medical condition precluded his participation in combat, and supported this assertion by claiming that he had been hospitalized for three months prior to his capture. In testimony before his Combatant Status Review Tribunal, however, Latif said that he had been treated for five days, which suggests a less serious condition. (Maj. op. 43-44). The government conceded, however, that Latif had documentation of a medical condition dating back to 1994. To conclusively demonstrate that Latif lacked credibility on this score, the court would have had to find that all of the documentation was fake. The government did not suggest this, and the majority did not enter such a finding. Ironically, the Latif majority’s one-sided opinion echoes the tone, if not the outcome, of earlier habeas decisions that favored detainees. For example, the district court’s Al-Adahi opinion, justly criticized in Ben and Bobby’s pioneering habeas study, seemed blind to a series of facts, such as stays in a Qaeda safe-house and training camp, that suggested that the detainee was part of Al Qaeda. The manifest contradictions in the detainee’s own account supported this view. In reversing the district court, the D.C. Circuit birthed its Bayesian approach, which can be roughly summarized as, “If the detainee walks like [part of Al Qaeda], and talks like [ditto], he probably is part of Al Qaeda.” In Latif, however, the D.C. Circuit was blind to the flaws in the government’s evidence, and to the possibility that the detainee was telling the truth. On either side, “Heads I win, tails you lose” is not the probabilistic theory that Bayes had in mind. Even more seriously, the Latif majority’s savaging of Bayes has an agenda of its own: discrediting the Supreme Court’s decision in Boumediene. The majority condemns Boumediene’s “airy supposition[]” that judicial review is necessary to keep the political branches honest. (Maj. op. 52). It also accuses the Supreme Court of “fundamentally alter[ing] the calculus of war.” Id. This stance suggests that the D.C. Circuit wishes to undermine the purpose of habeas itself – a dire conclusion that supporters of the Bayesian turn had rejected when it was uttered by the D.C. Circuit’s critics among the habeas bar. In the face of Boumediene’s warning that habeas was crucial for preventing the “pendular swings” that stem from the political branches’ excesses, Latif represents a pendular swing of its own from the interventionist tenor of the early district court decisions like Al-Adahi. Unfortunately, the right-leaning judges of the Latif majority have followed left-leaning judges like the Ninth Circuit’s Stephen Reinhardt in forgetting what the average ninth grade civics student knows: circuit judges interpret Supreme Court precedent; they don’t get to rewrite it.