By Michael Dorf
Justice Kagan is recused in Fisher v. Univ. of Texas at Austin (Fisher II), the SCOTUS affirmative action case that was argued on Wednesday. Consequently, going into the argument, the best that the university could realistically hope for was a tie: a 4-4 ruling with Justice Kennedy joining the three remaining relatively liberal Justices to affirm the Fifth Circuit's grant of summary judgment in favor of the defendant UT tying the four most conservative Justices to reverse that decision and instead grant summary judgment for Fisher. Under a rule dating to the early nineteenth century, the result of a tie would be that the lower court decision would be "affirmed by an equally divided court." The case would set no precedent but the Fifth Circuit ruling would remain.
During the oral argument, however, a third possibility emerged. At a number of points, Justice Kennedy suggested that perhaps the right result would be to remand to the district court either to supplement the summary judgment record or to hold a trial. To be sure, by the end of the argument, it appeared that Justice Kennedy's enthusiasm for a remand was dampened by UT attorney Greg Garre's vague answers about what evidence would be introduced if the record were reopened, but it remains possible that Justice Kennedy could conclude that a remand to the district court to take evidence is the right result. In any event, I'm going to assume that he favors that result for purposes of this post. On that assumption, the most likely initial vote at the Justices' conference on the case (which will happen in private today) would be:
Reverse and grant summary judgment for Fisher: 4 votes (Roberts, Scalia, Thomas, Alito)
Affirm: 3 votes (Ginsburg, Breyer, Sotomayor)
Vacate and remand for further district court proceedings: 1 vote (Kennedy)
What would happen in that situation? There would be no majority for any one outcome. Note that this is different from the typical case in which there is majority support for a result but not for a rationale, and thus no Opinion of the Court. To stick to the subject of affirmative action, we could consider Regents of the Univ of California v. Bakke for a standard example of the latter, more common kind of case. There, no view about the legal standard for judging affirmative action programs in general commanded five votes. Because it was "in the middle," Justice Powell's opinion was generally treated by lower courts as controlling for precedential purposes, although it didn't get the full force of precedent until it was adopted by a majority opinion 25 years later in Grutter v. Bollinger. But long before that, the outcome of the Bakke case was at least clear to the parties, because there were five votes for the same outcome: Bakke won.
The case I am hypothesizing would be different. No single outcome would command a majority. In 1834, Chief Justice John Marshall said that the Court wouldn't deliver a judgment in a constitutional case unless a majority of the members of the Court voted for the result, but as noted on p. 6 of the 10th edition of Supreme Court Practice (formerly Stern & Gressman but now Shapiro et al), the modern practice permits a majority of the Justices participating (so long as there is a quorum, which these days is six), to rule. However, with eight Justices participating, a majority still means five. So it looks like there would still be no opinion of the Court in Fisher II if the voting pattern hypothesized above comes to pass. What then? I spent about half an hour researching this question and some more time talking this over with proceduralists, only to come away uncertain. It is possible that there is a definitive answer out there in some case, but if so, I haven't found it. Nonetheless, I'm about 80% confident that the right answer would be to treat Justice Kennedy's vote as analogous to Justice Powell's vote in Bakke.
Even so, Fisher II would be more complicated than Bakke. In Bakke, there were only two outcomes proposed by the various Justices, even though there were more than two rationales. Thus, as noted above, although the precedential value of Bakke was up for debate, the outcome in the particular case was clear: Bakke won and the university lost. In Fisher, the votes (by hypothesis) divide among three outcomes, none of which has a majority: reverse (4); affirm (3); remand (1).
A reasonably close nine-Justice analogue is the 2006 case of Rapanos v. United States. The Court considered whether four Michigan wetlands fell within the regulatory authority of the Army Corps of Engineers pursuant to the Clean Water Act. Four Justices (in an opinion by Justice Scalia) effectively said no. As a technical matter, they merely vacated the appeals court's ruling so that their standard could be applied, but it was quite clear that under that standard, the absence of a steady flow of water meant that the wetlands fell outside of federal jurisdiction as the Scalia Four would have read the Act. Another four Justices (in a dissent by Justice Stevens) said yes, and would have affirmed the appeals court ruling to that effect. And then one Justice (Justice Kennedy, naturally) said maybe. Justice Kennedy thought that the Scalia test was too stringent and that the Stevens test was too permissive. He voted to remand for the application of an intermediate standard.
In light of the fact that Justice Kennedy's Rapanos vote was to remand under his unique standard, which was about halfway between the Scalia and Stevens poles, why did Scalia rather than Stevens announce the judgment of the Court? The answer, I think, is that there is at least a common core of the Kennedy and Scalia results: Both undo what the appeals court on its own did, whereas the Stevens group would simply affirm. The tricky bit is what the appeals court does with the case once it comes back. What it in fact did was to punt, by remanding to the district court and the Army Corps. After that, the case disappeared (from WestLaw, anyway, which is where I looked), but it is pretty clear that if it had come back up, the appeals court at that point would have been wise to apply the Kennedy standard. If application of the Kennedy standard resulted in finding the wetlands regulable, then five Justices (Kennedy plus the Stevens four) would affirm; if application of the Kennedy standard resulted in finding the wetlands non-regulable, then a different five Justices (Kennedy plus the Scalia four) would affirm.
Should the Rapanos approach apply in Fisher? In Fisher, as in Rapanos, no position commands a majority, but one could argue that in the case I've hypothesized, there would be five votes for a common core of undoing what the appeals court did. Once the case ends up back in the Fifth Circuit, that court could then apply the Kennedy standard from Fisher II (assuming, as seems plausible, that Justice Kennedy announces a unique one).
In the event that the Court splits as I've imagined, I think the foregoing is the most likely result, but let me suggest two alternatives. One such alternative would be that Fisher simply wins. Here the idea would be that Justice Kennedy's vote to remand is sufficiently different from either of the other alternatives that it's effectively not counted. It would be almost as though he were recused. Of the Justices participating and voting on the record as presented, the result would be 4-3 for Fisher. There is, I think, some logic to this approach, but I don't think it's consistent with the Court's practice of requiring at least a majority of the Justices participating in the case to get a judgment--and while Justice Kennedy's vote would be somewhat analogous to him recusing himself, it is in other respects quite different.
So consider another alternative. What if the three more liberal participating Justices decide to join Justice Kennedy in voting for a remand? Then the Court would be divided between four votes to reverse and grant summary judgment for Fisher versus four votes to vacate and order the record to be reopened (or for a trial) in the district court. Then what? Neither disposition is obviously "narrower" than the other (the much-criticized standard for determining the holding of a divided Court under the much-criticized and puzzled-over standard of Marks v. United States). And in any event, the Marks standard is used in figuring out what rule of law applies when there is an intra-majority disagreement over the rationale for a majority result, but here there is no majority result.
This leaves the arresting possibility that if the liberals strategically abandon their (presumably) preferred option of affirming the Fifth Circuit and join Justice Kennedy in voting to vacate and remand for a new trial, that would leave the Court evenly divided and so the judgment would be affirmed by an equally divided Court. That would be very weird, since (in this scenario), the Court would be equally divided between two outcomes, neither of which is affirming the appeals court, but possibly appropriate depending on how one conceptualizes the rule that an evenly divided Court leaves the lower court judgment intact.
If the idea is "the tie goes to the runner" (I know, that's debated in baseball), then application of the rule here makes little sense. It's not too close to call as between affirm and reverse. However, maybe the affirm-by-an-equally-divided Court rule is based on a different idea--simply that if the Supreme Court doesn't act, the status quo ante is undisturbed. If so, then it doesn't matter what the outcomes the various Justices favor are. If a majority of them can't agree on one, the appeals court decision stands. If so, then even if the three liberal Justices don't change their vote to join Kennedy for a remand, the Fifth Circuit opinion would be affirmed, albeit by an unequally divided Court (4-3-1 in this scenario).
Is that really a possibility? Maybe, but if so, then one can readily imagine the four conservative Justices switching their vote strategically as well, to get a remand to the district court on Justice Kennedy's terms rather than leaving the Fifth Circuit victory for UT in place.
The bottom line is not surprising: The Court will do what Justice Kennedy wants it to do. But the way it gets there could be intriguing.