Moreover, the Hamdan appeal is the polar opposite of routine for at least two reasons. First, its issues are central to the much-disputed claims of broad presidential power in the war on terror. Second, the court's decision on the Geneva Conventions has a spillover effect on the legality of controversial interrogation techniques used by the government at Guantanamo and elsewhere. That is because the same provision of the Geneva Conventions that would protect Hamdan from unfair trials also protects detainees from cruel, humiliating, or degrading treatment. The D.C. Circuit's decision rejecting the Geneva Conventions' trial protections—a decision that hinged on Roberts' vote—also strips away an important legal safeguard against cruel and humiliating treatment that may fall just short of torture.
Given the case's importance, then, when Gonzales interviewed Roberts for a possible Supreme Court seat on April 1, the judge should have withdrawn from the Hamdan appeal. Or he and Gonzales, as the opposing lawyer, should have revealed the interview to Hamdan's lawyer, who could then have decided whether to make a formal recusal motion. The need to do one or the other became acute—indeed incontrovertible—when arrangements were made for the May 3 interview with six high government officials. (We don't know how long before May 3 the arrangements were made.)
We do not cite these events to raise questions about Roberts' fitness for the Supreme Court. In the rush of business, his oversight may be understandable. What is immediately at stake, however, is the appearance of justice in the Hamdan case and the proper resolution of an important legal question about the limits on presidential power. Although the procedural rules are murky, it may yet be possible for Judge Roberts to withdraw his vote retroactively. That would at least eliminate the precedential effect of the opinion on whether the Geneva Conventions grant minimum human rights to Hamdan and others in his position. Better yet, the Supreme Court can remove the opinion's precedential effect by taking the Hamdan case and reversing it.