Yes, Gorsuch Matters

In a post on his blog, Eric Posner asks whether Gorsuch matters for administrative law.  The argument goes that while Gorsuch, a critic of Chevron deference, would be replacing Scalia, the Court's leading defender of Chevron deference, and joining Justice Thomas as perhaps the only other member of the Court that would like to overrule Chevron outright, data shows that Scalia and Thomas actually voted almost exactly the same way in administrative-law cases, perhaps because Scalia, while touting deference in theory, avoided deferring to "liberal" regulations by declining to find statutory ambiguity where less conservatively motivated Justices would.  It follows, Posner reasons, that if Gorsuch turns out to be "a clone of Thomas" in administrative-law cases, he'll be "a clone of Scalia" in administrative-law cases too.  I think this argument is completely wrong for a few reasons, and that Gorsuch's confirmation will very possibly be the undoing of deference doctrine as we know it.

Gorsuch Could Change Deference Doctrine in Lower Courts

First, assuming the argument works as far as it goes, i.e., that Gorsuch's confirmation will not result in materially different bottom-line results in administrative-law cases at the Supreme Court from what obtained before Scalia's death, it's almost entirely besides the point because bottom-line results in the handful of administrative-law cases the Supreme Court hears is not what "matters" for these purposes, or at least not what matters most.  What "matters," it seems to me, is whether, as a general matter, the federal system as a whole will become one in which courts decide (or at least more frequently decide) the meaning of ambiguous statutes that agencies administer, or continue to be one where agencies usually decide, subject to reasonableness review, the meaning of ambiguous statutes that Congress asked them to administer.  

It may be that what the Court has to say doctrinally on that score has never had overwhelming influence on how the Court actually decided the small set of difficult and often highly politically charged administrative-law cases that come to it; that was very possibly true, at least, of Scalia.  But all evidence suggests that what the Court has to say about deference doctrine has great influence on whether the lower courts defer to agencies in practice.  Lower courts, a recent comprehensive study shows, are much more likely to rule for and defer to agencies when applying the Court's Chevron deference rule than a softer deference standard like Skidmore, and vastly more likely to rule for agencies when applying Chevron than when they apply no deference standard at all.  Were the Court to hold that Chevron deference is not applicable in some large class of cases where it currently is, or to overrule Chevron altogether, results in lower courts would be quite different.  On the great many statutory questions on which agencies regulate that Congress has not clearly addressed and the Supreme Court does not grant certiorari to decide, the law would meaningfully more often be court-made law than agency-made law.  That seems like a big deal.  (I quite appreciate that "court-made law" is a tendentious formulation and that Judge Gorsuch and others would claim that overruling Chevron would replace agency-made law with the law Congress actually made, as divined, imperfectly but fairly accurately, by courts.  For reasons I'll address in a posting next week, I find this a deeply implausible view of the world.)

Now, it is happily or unhappily the case that if confirmed, a Justice Gorsuch would only be the second member of the Court to even express interest in overruling Chevron.  But by my count, he will likely be the fifth or sixth Justice open to a narrowing of Chevron that would be tantamount to overruling it in all but name.  In City of Arlington v. FCC four years ago, just five Justices, including Justice Thomas, joined an opinion of the Court authored by Scalia stating the once-uncontroversial view that if an agency with general rulemaking or adjudicatory power under a statute announces a reasonable interpretation of an ambiguity in that statute in a notice-and-comment rulemaking or formal adjudication, it is entitled to Chevron deference, period--regardless of whether the question it is deciding can be characterized as whether or not it has the power to regulate on some substantive question in the first place. The other four members of the Court had other ideas.

Justice Breyer, consistently with opinions and writings of his going back decades, could only concur in the judgment, opining that an agency is entitled to Chevron deference only when ten or so factors lead him to believe that Congress would have wanted the agency to receive deference on the particular question at issue.  In dissent, the Chief Justice, writing for himself, Alito, and Kennedy, complained that an agency can't receive deference on whether or not it is entitled to deference, and that to say an agency is entitled to deference on whether or not it is entitled to regulate on some substantive issue is to say precisely that, given that if it can regulate on the issue, it will be entitled to deference.  That sounds reasonable enough, but as the argument reveals itself, it turns out that Roberts isn't only arguing that agencies cannot receive deference on whether some subject matter is within their regulatory "jurisdiction," but also that the question of whether an agency has been delegated "interpretive authority" on any given ambiguity within its organic statute must be decided "specific statutory ambiguity" by specific statutory ambiguity, by "simply apply[ing] the normal rules of statutory construction."  How the normal rules of statutory construction (as opposed to some abnormal set of antediluvian hyper-purposivist rules) are to shed light on whether a statute that by its terms gives an agency general rulemaking authority to carry out all its provisions "really" gave the agency interpretive authority on any one of those provisions is not explained and unfortunately never illustrated on the facts of the case, as Roberts would, oddly, have remanded to the court below to let it decide in the first instance whether Congress truly "delegated interpretive authority" over the provision of the Telecommunications Act at issue to the FCC.

Jibes aside, the underappreciated upshot of City of Arlington is that even before Justice Thomas began to opine that Chevron was unconstitutional, four Justices who still are on the Court today were of the view that Chevron deference did not apply to all ambiguities in an agency's organic statute, so long as the agency addressed them in a sufficiently formal way, but rather only applied where a court determined that Congress really meant for the agency to have "interpretive authority" on a given ambiguity.  While that approach would lead to deference some of the time--realistically speaking, in cases presenting issues that courts didn't care about enough, or disagree with the agency enough about, or want to exert the effort, to decide themselves--it has nothing to do with Chevron as the lower courts understand it.  Rather, without coming out and saying so, it would amount to a reincarnation of the confused, permissive pre-Chevron deference doctrine that Chevron replaced.

It's hard to say whether, failing overruling Chevron in so many words, Justice Thomas would have any interest in cutting back on deference by joining these four Justices in licensing lower courts to engage in delegation mind-reading.  I am inclined to believe that Justice Thomas is too principled, too discomfited by contentless standards, far too disinterested in guessing at what was in Congress's heads or attributing intentions to Congress that it lacks, and, at least in the past, too smart an expositor of a truly rule-like Chevron doctrine to go in for it.  It is, of course, possible that Justice Thomas may simply vote to overrule Chevron in every Chevron case, in the same way that Justices in the 80s who lacked a majority to invalidate the death penalty would nevertheless vote their precedentially foreclosed views in death cases.  Were Justice Thomas willing to vote in that way, which I think is possible and not without precedent in his career, Gorsuch wouldn't need to join the Court for the Breyer/Roberts reading of Chevron to become law; they could make it the law by plurality, with Thomas concurring on broader grounds.  Query, however, whether all five of the Justices it would take to effectively restore pre-Chevron deference doctrine in this fashion would be inclined to so fundamentally change the law by way of a Marks holding, or whether such a holding would be stable or bind the Court.

Gorsuch, it seems to me, is the answer to the Breyer/Roberts coalition's problems; he would almost certainly give them the needed fifth vote to narrow Chevron.  Why do I assume Gorsuch wouldn't be just as principled a coherentist about deference as Thomas?  Because, in the past, he hasn't been.  In his famous concurring opinion in Gutierrez-Brizuela, Gorsuch boasted that in his opinion for the panel and another opinion before that, he had "sought to tame some of Brand X's more exuberant consequences"--yes, that's a lower-court judge talking about "taming" a Supreme Court precedent that's only 11 years old--by holding that in the future when his court gave Chevron deference to an agency interpretation and overruled contrary circuit precedent under Brand X, it would never apply its decision retroactively to the parties before it, thereby turning future Brand X cases in his circuit into bizarre exercises in advisory opinion-giving.  Gorsuch would defend this advisory approach to Brand X by likening it to qualified-immunity cases that announce a new constitutional rule without sanctioning the defendant that violated it.  In qualified-immunity cases, though, plaintiffs attempt to show that their rights were clearly violated under existing law and that the rule of constitutional law they claim the defendant violated should apply in their cases; in Gorsuch's version of Brand X, the parties to a case implicating Brand X would always know that nothing they argued could possibly affect the court's ultimate judgment.  Whatever the court said about its precedent going forward, it would always apply its old rule to the parties before it.  (It's true that some parties would be repeat players and care about prospective application, but, to give just one common example, individual immigrants would not be and thus would have no reason to argue about whether an agency interpretation that couldn't possibly affect them was owed deference.)  I'll say more about Gorsuch and deference in subsequent posts, but for now suffice it to say that he has no problem with making an incoherent mess of deference doctrines he doesn't like so long as doing so cuts back on deference.

Gorsuch Would Change Outcomes in Big-Ticket Supreme Court Cases

On the merits of Posner's actual argument--namely, that whatever Gorsuch might do to doctrine, he wouldn't change the results of the Court's administrative-law cases too much--Posner's inference from data that Scalia and Thomas voted roughly the same way in administrative-law cases goes astray in a couple respects.  The first is that Justice Thomas only became anti-Chevron Thomas on the last day of the 2014-15 term (though he hinted at his direction a few months prior).  Prior to that, while Justice Thomas did not, like Scalia, eccentrically urge Chevron deference to opinion letters, amicus briefs, and other forms of guidance that no one else on the Court deferred to, he was about as staunch an advocate of mainstream Chevron doctrine as there was.  The most noteworthy example, of course, was his 2005 opinion in Brand X holding that the lower courts, if not the Supreme Court, are obliged to overrule their precedent on ambiguous statutes when an agency subsequently offers a contrary reasonable interpretation; a more quiet example is his joining Scalia's opinion in City of Arlington.  The data set Posner borrows is from 2005-15, during which time Thomas and Scalia not only agreed on politics and statutory interpretation, but deference as well.  So to point to Thomas's 2005-15 votes as indicia of where opposition to Chevron would lead Gorsuch, or as evidence that theoretical disagreement between Thomas and Scalia on Chevron made no difference in terms of the real degree of deference they gave agencies, doesn't work, not only because it assumes all sorts of things that may or may not be the case, but because Thomas's 2005-15 votes were not the votes of a Chevron opponent.

Second, Posner's data does show that, in 31% of the agency cases involving "liberal" agency decisions that were hard enough and important enough for the Court to grant cert, Scalia voted for the agency, while Thomas voted to uphold liberal agency decisions 26% of the time.  (Similarly, it shows that Kagan, Ginsburg, Breyer and Sotomayor have voted to uphold conservative agency decisions 33, 34, 36 and 40% of the time respectively.)  Posner says this is evidence of Scalia refusing to defer to liberal agencies in practice, whatever he said about Chevron in theory.  31% isn't great, but, with the possible exception of the criminal defense bar, I would be surprised to learn of some liberal interest group other than liberal regulators that Scalia voted for 31% of the time, or that Thomas voted for 26% of the time.  And I would be fairly surprised if absent a meaningful deference doctrine liberal regulators could even claim a 31% win rate before Scalia or a 26% win rate before Thomas, or whether conservative regulators could claim win rates of 33 to 40% before Sotomayor, Ginsburg, Kagan and Breyer.  Narrowing Chevron to oblivion could make all members of the Court a lot more predictable and ensure that whichever party was in the majority on the Court at any given time won the bulk of agency cases.