Now let's consider how she might be restricted from hearing certain cases.
No federal judge can participate in a case in which she has
served in governmental employment and in such capacity participated as counsel [or] adviser … concerning the proceeding or expressed an opinion concerning the merits of the particular case in controversy. 28 U.S.C. § 455(b)(3).
Based on her answers to written questions from members of the Senate Judiciary Committee, there are several lgbt-related cases in which Justice Kagan would recuse herself and one case where she may be subject to recusal. Most will probably not be consequential, but recusal in the last case, if it occurs, would be significant.
The first is Smelt v. United States, a challenge to the constitutionality of DoMA. Kagan stated that she had participated in discussions of DoJ's briefs in this case and that she would recuse herself if it came before the Supreme Court. (These were the briefs defending DoMA that caused such heartburn among lgbt advocates.) However, the Smelt case was dismissed on technical grounds a year ago.
The second case is Witt v. Air Force, a DADT challenge. After the Ninth Circuit ruled that the military had to demonstrate on an individualized basis that the discharge of gay service members was justified by unit cohesion, the decision on whether to file a cert petition was squarely in the bailiwick of the SG's office. No cert petition was filed, and the case is now on remand to the district court. Again, there is no question that Kagan would have to recuse herself from this case. Hopefully, though, the DADT policy will have been repealed – and the case mooted – by the time it could re-enter the appellate process. Kagan testified that she also discussed discovery issues in Log Cabin Republicans v. U.S., another pending DADT challenge, which, like Witt, hopefully will be mooted by congressional action.
The most important case – and the one with the biggest question mark – is Gill v. OPM, in which a federal district court ruled that Section 3 of DoMA (barring federal recognition of same-sex marriages) is unconstitutional. This case is very likely to reach at least the front door of the Supreme Court, by which I mean that a cert petition will almost certainly be filed by whichever party loses in the Court of Appeals. If the First Circuit affirms the district court, the Supreme Court will probably grant review. And it's a virtual certainty that plaintiffs can't get five votes on this Court without one from Justice Kagan.
Kagan told the Senate Judiciary Committee that she would recuse herself from a case if she had "participated in formulating the government's litigating position or reviewed a draft pleading. In all other circumstances, I would consider recusal on a case-by-case basis."
With regard to Smelt, she said "I reviewed some briefs in the Smelt case and participated in discussions about the case shortly before the briefs were filed. My participation in the case was sufficiently substantial that I would recuse myself…"
She was then asked whether she reviewed briefs or participated in discussions about Gill. Her response: "Yes, I believe that discussions about Gill overlapped with discussions about Smelt." In other words, recusal in Gill is certainly a possibility. However, without knowing the extent and nature of the discussion specifically of Gill, it is not clear whether her participation was sufficiently substantial to require recusal. The outcome of the litigation could turn on the answer to that question.
There is a back-up scenario, however, even if Kagan is recused: if there is a four-four split among the Justices, the decision of the First Circuit would determine the outcome. Since it is unlikely that the Court would grant cert if the circuit court upholds the statute, this may turn out to be the fail safe option. In other words, if they win in the First Circuit, plaintiffs could win the case without any opinion from the Supreme Court. But the loss of Justice Kagan's vote, as well as of her participation in conferencing among the Justices, is likely to make that more difficult.