Qualms About a (Henry) Friendly Court

By Michael C. Dorf

Chief Justice Roberts was a law clerk for the great Judge Henry Friendly and is fond of invoking his former boss for a standard proposition of judicial restraint: If it's not necessary to decide an issue to resolve a case, then it's necessary not to decide that issue. It is not always clear that Roberts means this. Although I have no doubt that as a matter of tone the Chief Justice prefers a Court that is restrained and perceived as apolitical, he has certainly authored and joined opinions that could be said to reach out to decide controversial questions. But if we are to take Roberts at his word, he should be delighted with the Court's current 8-justice configuration, which appears to be leading the Court to avoid deciding important issues.

Obviously, when the Court splits 4-4 and affirms by an equal division it decides nothing as a matter of precedent. But sometimes even unanimous action has an indecisive character, as in Monday's non-decision decision in Zubik v. Burwell. As I noted last month when the Court sought supplemental briefs from the parties, the justices were probably divided on key issues and trying to find a way to get the parties to settle. Monday's order sending the cases back down confirms that, but as Marty Lederman explains here, full settlement is unlikely because there remain outstanding differences between the government and at least some of the plaintiffs. Thus, I agree with the assessment of Nelson Tebbe, Micah Schwartzman, and Richard Schragger, who say that "[w]e will see Zubik in the Supreme Court again, most likely."

In the meantime, I want to question both the logic and the wisdom of Judge Friendly's dictum. Often it is not necessary to decide any issue to resolve a case, but it doesn't follow--indeed it can't follow--that it is therefore necessary to decide nothing.

In many cases there are various potential routes to resolution. Consider a schematic example. In order to win on appeal, petitioner must prevail on either issue X or issue Y. Suppose that after a preliminary vote, the court agrees that petitioner should win on both issue X and issue Y. Is it necessary to decide X? No, because a decision based on Y would suffice for the petitioner to win. Is it necessary to decide Y? No again, because a decision based on X would suffice. And yet it can't therefore be necessary not to decide X or Y, because deciding neither X nor Y means not deciding the case.

The foregoing schema is more common than you might realize. Although the formal presentation of a case might not starkly present alternative equally sufficient grounds for a particular outcome as in the schematic example, usually there are such alternatives presented at the level of sub-issues. It may be the rare case in which it is necessary to resolve all of sub-issues that a court in fact resolves en route to its bottom line.

How then should we understand Judge Friendly's aphorism? It was clearly meant as a kind of avoidance canon, but note that it is distinct from constitutional avoidance. Judge Friendly did not say that if it's not necessary to decide a constitutional issue . . . . He said that if it's not necessary to decide an issue . . . .  To the extent that this is logically possible, I have always understood this as stating a principle about avoiding difficult or politically charged issues.

In the Zubik context, these imperatives arguably point in different directions. Whether the contraception mandate with the opt-out offered by the government amounts to a RFRA substantial burden on some or all of the Zubik plaintiffs undoubtedly is politically charged. But it might not be very difficult. At least it seems less difficult than the myriad subsidiary issues that the lower courts will need to resolve in figuring out what sorts of accommodations the government can provide consistent with other laws.

Of course, the Court in Zubik did not offer Judge Friendly's maxim as the basis for its action. Rather, it looks like the remand was the result of an inability to agree on the outcome. Or does it? My colleague Steve Shiffrin suggests that the effort to induce settlement in Zubik might not be a product of a 4-4 split.

I'm inclined to agree normatively but not causally. That is, as I noted in my Verdict column comparing the initial request for supplemental briefing to convening a settlement conference, it is laudable for the Court to play this settlement-facilitation role when it can. But given how rarely the Court in fact plays this role, I find it hard to imagine that the justices would have done so were there five clear votes for any outcome. That seems especially true now, given that the Court's remand rested on papering over at least some differences between the parties.

But even if I'm right and Prof. Shiffrin is wrong about what went on behind the scenes in Zubik, there does seem to be a broader spirit of decision avoidance at the Court, even in cases that don't divide 4-4. Consider Monday's ruling in Spokeo, Inc. v. Robins. That case produced a 6-justice majority and yet the Court went out of its way to decide as little as possible. After finding that the Ninth Circuit erred by recognizing standing in a plaintiff who had alleged a "particularized" injury without also making a determination that the injury was "concrete," the Court could have readily made its own determination whether the injury alleged was sufficiently concrete to satisfy Article III. Instead, the Court remanded back to the Ninth Circuit for it to apply the concreteness test.

To be sure, that kind of remand is not necessarily noteworthy. The Court often gives the lower courts the first crack at applying some new test, especially where a factual record is needed. But here the question is whether the allegations in the complaint are sufficient to establish standing, so no factual findings will be necessary. Moreover, the SCOTUS did not fashion a new test. The requirement that an injury be "concrete and particularized" (my emphasis) has been quoted in dozens of prior cases. Nor did the SCOTUS opinion in Spokeo provide some new definition of "concrete."

Indeed, as I argue in my latest Verdict column, the Spokeo Court can be faulted for unhelpfully defining "concrete" as "de facto" and "actually exist[ing]." As I explain, were it up to me, I would dispense with judicial review of "concreteness" where Congress has granted standing to sue for remedies for particularized injuries. But the Court doesn't go down that route. Instead, it pretty much just repeats prior formulations of the standing rule.

Having taken the case and neither clarified the law in general nor applied it to resolve the particular case, the Spokeo Court unwittingly offers a pretty good argument against the Friendly maxim. Even if it's not strictly necessary to decide some issue, often it's a darn good idea.