Predictions are hard, especially about the future. It’s much safer to hide behind a broad trend: There was a time, just a couple of years ago, when it seemed like Auer and Seminole Rock were not long for this world. Auer’s author, Justice Scalia, turned his back on the doctrine. The tide of scholarly opinion—led by former Scalia clerk John Manning—seemed to rise comfortably, almost casually, in opposition to Auer. Fueled by a few encouraging concurrences, cases percolated through the federal courts that seemed to provide a ripe opportunity for review.

But no longer. The Court reaffirmed Auer last year. And the scholarly waters—if the previous contributions in this symposium and a few longer papers are a reasonable guide—seem a little more brackish than they once did. Casual contempt is now matched with casual acclaim. (Spake Sunstein and Vermeule: “the [Auer] principle is entirely correct.”) Can these datapoints be shoehorned into a prediction? Sure, why not: The Supreme Court isn’t going to overturn Auer and Seminole Rock anytime soon.

The brute facts mentioned above are a big reason. The Court just reaffirmed Auer; the doctrine lost one of its staunchest critics. The votes to overturn it don’t seem to be there. And the existence of a symposium like this one is surely another brute fact in Auer’s favor. If the doctrine really were just conspicuous lunacy, it seems unlikely that smart people would spend so much time disagreeing over it.

And there are two additional reasons why Auer will persist. The first has been flagged by many previous contributions to this symposium: The big criticism of Auer— the incentive for vague regulations—can feel a little underwhelming. Even if you buy the criticism in theory—and I do—there is always the intractable-feeling question of how to weigh it against the other values of administrative law. Administrative law serves many masters. Vague regulations can be bad—but so can fewer regulations, slower regulations, and en-masse-suddenly-rewritten regulations. It’s anyone’s guess how these various levers will shift in a post-Auer world; it seems to me that modesty dictates a bit of uncertainty about the pluses and minuses of it all. And even if you buy the criticism in practice—that is, even you’re confident that the cost of vagueness outweighs these other effects—you may run up against an equally intractable-seeming question of how to prove the point. Proving Auer’s vices decisively requires more than counting cases; it requires finding an acceptable measure of vagueness and a plausible source of exogenous legal variation. No mean feat.

The second reason: Even if we think Auer is broken or imperfect, there are plenty of ways to fix it short of scrapping the whole doctrine. Many possible fixes have been flagged by other contributors. But it seems to me that a perfectly reasonable way to cabin Auer is to operate at the so-called step zero: We can limit the types of agency interpretations to which it applies. The basic idea is simple: If we’re most concerned about opportunistic or feckless agencies surprising regulated entities with last-minute interpretations, we can keep Auer from rubber-stamping the worst of the bunch.

Indeed, the Auer /Seminole Rock test — “the administrative interpretation . . . becomes of controlling weight unless it is plainly erroneous or inconsistent with the regulation”— has already been circumscribed in a number of key ways, all of which one can think of as amounting to an Auer step zero. I catalog these step-zero (or sometimes step-one) cases in a semi-recent essay. First, Auer deference is warranted only when the regulation in question is really ambiguous. Second, the agency’s interpretation must be a considered judgment; it can’t be simple after-the-fact convenience. Third, the agency’s interpretation cannot impose too much unexpected liability. Fourth, in at least one circuit, the agency’s interpretation cannot be interpreting a pre-existing interpretation. And fifth, in several circuits at least, the relevant interpretation cannot conflict with the agency’s previous interpretations.

These cases matter because they suggest that Auer can fix itself—should we believe it’s broken, of course. This step-zero circumscription “tailor[s] deference to variety,” as Souter once said of Mead. A little careful tailoring might serve regulated entities much better than blowing up Auer entirely.

Conor Clarke is a graduate of Yale Law School and a current law clerk for Judge Alex Kozinski.

––

This post is part of an online symposium entitled Reflections on Seminole Rock: The Past, Present, and Future of Deference to Agency Regulatory Interpretations. You can read the entire series here.