Does the Administrative Procedure Act Provide for Universal Vacatur?

Event Video

Listen & Download

Section 706 of the Administrative Procedure Act instructs courts to "set aside" agency action that is unlawful. These two words—"set aside"—have sparked much debate among lawyers, jurists, and scholars. In particular, administrative law enthusiasts disagree about whether the "set aside" language means that courts must enter universal injunctions against unlawful regulations. Some suggest that "set aside" contemplates wholesale invalidation of regulations. Others take the position that the scope of the "set aside" remedy is more limited. In United States v. Texas, which is before the Supreme Court this term, the Court is set to decide whether the "set aside" language requires universal vacatur of regulations. In advance of the Court's ruling, this teleforum panel will discuss this difficult question.



  • Hon. Beth A. Williams, Member, United States Privacy & Civil Liberties Oversight Board, Professorial Lecturer in Law at the George Washington University Law School, and former Assistant Attorney General for the Office of Legal Policy, United States Department of Justice
  • Prof. John Harrison, James Madison Distinguished Professor of Law and Thomas F. Bergin Teaching Professor at the University of Virginia School of Law
  • Prof. Ron Levin, William R. Orthwein Distinguished Professor of Law at the Washington University in St. Louis School of Law
  • [Moderator] Judge Steven J. Menashi, U.S. Court of Appeals for the Second Circuit






As always, the Federalist Society takes no position on particular legal or public policy issues; all expressions of opinion are those of the speaker.

Event Transcript



Chayila Kleist:  Hello, and welcome to The Federalist Society's webinar call. Today, February 17, 2023, we discuss the question, does the Administrative Procedure Act provide for universal vacatur? My name is Chayila Kleist and I'm an assistant director of practice groups here at The Federalist Society. As always, please note that all expressions of opinion are those of the experts on today's call as The Federalist Society takes no position on particular legal or public policy issues.


In the interest of time, we'll keep our introductions brief but if you'd like to know more about any of our participants, you can access their impressive full bios at Today, we are honored to have with us as our moderator, the Honorable Judge Steven Menashi who has served on the U.S. Court of Appeals for the Second Circuit since 2019. Prior to being appointed to the Second Circuit, he served as special assistant and associate counsel to the president in the white house and as acting general counsel in the U.S. Department of Education. He was also an assistant professor of law at Scalia Law School at George Mason University and is a partner at Kirkland & Ellis in New York City. I'll leave it to Judge Menashi to introduce our panel.


One last note, throughout the webinar, if you have any questions, please submit them via the question-and-answer feature so we have access to them when we get to that portion of today's webinar. With that, thank you all for being with us today. Judge Menashi, the floor is yours.


Judge Steven J. Menashi:  Thanks very much, and thanks to The Federalist Society for organizing this panel today. Today, our topic is the availability of universal vacatur under the Administrative Procedure Act. That is whether courts may invalidate entire regulations for everyone in the country consistent with the APA rather than being limited to providing relief to the parties before the court. Given the rise of policy making through regulation and some recent discussions about forum shopping among judicial districts, the stakes are potentially very high. And it's an issue the Supreme Court might address this term because in a pending case, United States v. Texas, which is about federal immigration policy, the question of universal vacatur under the APA is implicated.


So to help us puzzle through these issues, we will hear from three distinguished panelists. I will introduce them in the order in which they'll speak, and then they'll each give us opening remarks, and then we'll open it up for discussion among the panelists and questions from the audience. So first, we'll hear from Beth Williams. Beth is a member of United States Privacy & Civil Liberties Oversight Board and a Professorial Lecturer in Law at the George Washington University Law School. She served as Assistant Attorney General for the Office of Legal Policy at the Department of Justice from 2017 to 2020.


Then we'll hear from John Harrison. John is the James Madison Distinguished Professor of Law and the Thomas F. Bergin Teaching Professor at the University of Virginia School of Law. In 2020, he published an essay in the Yale Journal on Regulation Bulletin called “Section 706 of the Administrative Procedure Act Does Not Call for Universal Injunctions or Other Universal Remedies,” an essay that's become influential in this area. Then we'll hear from Ron Levin. Ron is the William R. Orthwein Distinguished Professor of Law at the Washington University in St. Louis School of Law. His essay, “Vacatur, Nationwide Injunctions, and the Evolving APA” is forthcoming in the Notre Dame Law Review. So with that, Beth Williams.


Hon. Beth A. Williams:  Good afternoon and thank you very much to Judge Menashi and to The Federalist Society for inviting me. It's a pleasure to be on this panel with such distinguished scholars on this topic. Just to give the standard disclaimer, in case my agency's general counsel is watching, I'm not here in my capacity as a member of the Privacy and Civil Liberties Oversight Board and any views expressed here do not necessarily reflect the views of the U.S. government, the board, or any other board members. So onto the fun part.


The oral argument in United States v. Texas in November was a very interesting one. As background, the state plaintiffs brought a challenge to the Department of Homeland Security's guidelines for the enforcement of civil immigration law. Essentially, the Biden administration has been trying to prioritize certain groups of unauthorized immigrants for arrest and detention. The states on the other hand are arguing that the guidelines are at odds with the brightlines or categories that Congress laid out. And the merits of the case are interesting, but for our purposes today, the focus is on the remedies discussion and that's where some of the fireworks were.


It's not often the chief says, "Wow," to an argument made by the solicitor general and there was some good-natured ribbing among the justices between those who belonged to the D.C. Circuit cartel, as Justice Kagan named it and the other justices. The cartel, as it were, would be comprised of the chief, Justice Kavanaugh, and Justice Jackson all of whom served on the D.C. Circuit before being elevated to the Court. But of course, there's one more member of the Court who also served on the D.C. Circuit, Justice Thomas. And it's not clear whether Justice Thomas shares the views of his fellow D.C. Circuit alumni on this issue.


So what made the chief say, "Wow"? Well, the solicitor general advocated a very different interpretation of the meaning of "set aside" in the Administrative Procedure Act than the D.C. Circuit or really most anyone had taken before. Section 706 of the APA provides the reviewing court shall hold unlawful and set aside agency action, findings, and conclusions found to be in violation of six listed standards of review among which of course is if the court finds the action to be arbitrary and capricious.


For decades, the D.C. Circuit has taken "set aside" to be the equivalent of an authorization for the courts to stand up for the administrative state yelling, "Stop." And for many people who have issues with the size and insulated power of the administrative state, that's a good thing. In many ways, the APA and the changes of presidential administrations have provided the only meaningful checks on the actions of regulators. But originalism. In recent years, there's been some fascinating academic work, most notably by Professor Harrison and by Professor Sam Bray at Notre Dame that "set aside" never meant what everyone thinks it meant. Did Congress, when it passed the Administrative Procedure Act in 1946 with a very different administrative state, mean to give a district court judge the power to completely strike down or vacate regulations as to everyone? Or did it mean to allow courts to set aside or disregard an unauthorized regulation with respect to adjudicating the case for the parties before it?


And of course, regardless of what Congress intended, what did people understand "set aside" to mean at the time? There's good historical research to suggest people did not understand the APA to be majorly displacing well understood norms of equitable practice. And if that is what Congress was doing, wouldn't they have put it in the remedies section -- section 703 -- rather than the scope of review section -- section 706? And wouldn't they maybe have used stronger language than "set aside?"


So this is a fascinating case of statutory interpretation and in some ways sets up a textualist originalist split on the Court. At oral argument, you have Justice Kavanaugh making a textualist argument that "set aside" means "set aside." That's always been understood to mean. He said that the rule is no longer in place. Contrast that with Justice Gorsuch who focused more on the structure of the statute and also seemed more receptive to the government's and Professor Harrison's originalist arguments.


So predictions for this term. Where does that leave us? I could of course be wrong, but my prediction is that the Court is going to punt on the question this term. There are a couple of reasons for that. First, this is a major issue of statutory interpretation, and the briefing was not extensive. It's just a few pages on remedies at the end of each brief. Now you would think that if the Court were going to tackle a question this big now they would ask for additional briefing. Second, as many have noted, there hasn't been percolation in the lower courts on this issue, so the Court doesn't have the benefit of circuits weighing in.


Third, the government made only the statutory argument and not the constitutional argument before the Court. The statutory argument is that universal vacatur is not actually authorized by the APA. The constitutional argument would be that even if universal vacatur were authorized by the APA, such a statute would impermissibly extend the Court's power beyond the case or controversy limitation of Article III. In the somewhat related context of nationwide injunctions in Trump v. Hawaii in 2018, Justice Thomas raised this constitutional question in his concurrence when he said that even if Congress someday enacted a statute that clearly and expressly authorized nationwide injunctions, courts would need to consider whether that statute complies with the limits that Article III places on the authority of federal courts. Again, that constitutional question was not briefed in this case. And finally, if the Court gets to the remedies section, it can resolve the question through the alternative arguments section 1252 and not get to the question under the APA.


So implications. What if the Court sided with the government and concluded that "set aside" and the statute does not provide for universal vacatur. What then? Well, the irony of all this is that it wouldn't actually change much in D.C. As Professor Bray explains, the reason for that is the power of precedent. So if Party A sues the federal government in D.C. and wins and it also wins on appeal to the D.C. circuit, then that ruling will have precedential effect for every person who comes later who sues the federal government in D.C. The Court doesn't need to vacate the rule. It just needs to issue a judgment. By the power of precedent, everyone who comes next will win and everyone can sue the federal government in D.C. So after a loss in the D.C. Circuit that's not reversed by the Supreme Court, the government as a practical matter is not going to keep going with a rule held to be unlawful in a place where everyone can sue it.


Reading "set aside" not to mean universal vacatur matters more in every other circuit. And for where the government can't be sued by everyone in the country, the precedential effect of one decision is necessarily not as broad. But still the government may then make the decision not to enforce its regulation within that district or circuit or may make the decision that nationwide uniformity is so important to the particular regulatory scheme that it will rescind its rule. But the relative importance of uniformity or the workability of dis-uniformity sits with the elected branch, not the judiciary. All of that is to say that reading the APA not to authorize universal vacatur still maintains the APA as a significant check by the judiciary on unlawful regulatory action. Thanks.


Judge Steven J. Menashi:  Thank you very much, Beth. John.


Prof. John Harrison:  Thank you. First, briefly my answer to the question is, no. The first thing I'm going to say is because I'm a law professor, I'm going to quibble with the question and say that universal vacatur is redundant. The interesting thing about vacatur is that the lower courts have come to see it as a remedy distinct from other remedies and in particular, distinct from injunctions against enforcement, distinct from declaratory judgments and vacatur as a distinct remedy intrinsically is universal because it operates on the legal status of what the agency has done and therefore operates as to everyone.


An injunction against enforcement by contrast, can be universal, can tell the agency not to enforce against anyone, but doesn't have to be universal. That is governed by principles of federal equity, I think. Whereas the interesting thing about vacatur, if it's available, is that it's always universal and a large part of the argument here is the idea that by calling for that kind of remedy, not just injunctions and declarations, the APA pretermits the whole argument about whether universal remedies are appropriate by creating universal remedy and calling for it. So it's important to keep straight the distinction between remedies that might be universal, notably injunctions, and the idea of vacatur which has to be universal because of the way it's conceived.


Second, my substantive claim is that the argument that the APA itself resolves this question in favor of vacatur is incorrect. Rather, the APA doesn't contain any of the law of remedies. It points to other sources of the law of remedies. It does so most notably in 5 U.S.C. §703 which describes the system of judicial review, divides the world into special statutory review and other forms of proceedings for review. Special statutory review is conducted pursuant to specific statutes that usually provide for review directly. In the court of appeals there are a lot of statutes like that. There were already several when the APA was adopted.


Some of them used the word "set aside." Some of them say "set aside." My claim is they mean whatever they mean. The reference to "set aside" in section 706 doesn't add anything to that nor does 706 change anything in the other forms of proceeding, notably the suits for injunctions, declaratory judgments, damages, habeas corpus, all of which are listed in section 703. They all have their own law of remedies and the place to look for whether universal remedy is available is in that body of remedies.


Some of the proceedings with their associated remedies that are referred to in section 703 clearly are remedies that are specific to a part. Habeas is on the list. The remedy in habeas is for the petitioner to be released from confinement -- debt from confinement. Damages are on the list. The remedy in damages is an award of damages to the specific party. Declaratory judgments are on the list. Declaratory judgments declare the rights of parties, but injunctions are on the list. Maybe injunctions can properly be universal, maybe not but that's a question about the law of injunctions. It's not a question about the APA.


That then raises the question -- that's basically a structural observation about the way APA judicial review is set up. It's divided into these categories and each one has its own law of remedies. That raises the question, what about section 706? There's no indication that section 706 was designed to change all of those other components of the law of judicial review. Section 706 doesn't reach into special statutory review provisions and say, "Stop doing whatever that statute was doing and now do what section 706 means by 'set aside'." Section 706 doesn't change what goes on in habeas, it doesn't change what goes on in the declaratory judgment proceeding. Declaratory judgment proceedings are governed by the declaratory judgment act.


So what does "set aside" in section 706 mean? I think it means one of two things. I think the best reading is that it means what "set aside" sometimes means in other contexts -- put aside, disregard, don't decide in accordance with, and then go on and do whatever the court should do. That matches section 706 which is about scope of review and tells the courts what issues they are supposed to decide, what issues they are to leave to the agencies. And section 706 read that way tells the courts "Here is what you decide for yourself and if you decide that the agency is on the wrong side of the law, then don't decide in accordance with what the agency has said."


Another reading -- I think the second best reading -- because maybe -- I don't think so -- maybe section 706 refers to remedies in some sense when it says "set aside." The other possibility is that's a generic placemark for whatever is the appropriate remedy. So once again, section 706 points back to special statutory review proceedings, points back to injunctions, points back to the declaratory judgment act, for example, but that's where the answers are found. They're not found in section 706. So that's why I say the attempt to look to the APA for the answers to these questions is unavailing. The answers are found elsewhere. There are bodies of law that answer this question but it's not the APA.


Judge Steven J. Menashi:  Okay. Thank you very much, John. Ron.


Prof. Ron Levin:  Okay. Well, let me start off with where I agree with John. I agree that universal vacatur is a redundant term because vacatur means, by definition, applying across the board. However, his answer to the question posed in the title is incorrect. The correct answer is yes and I will try to explain why that is so. It's true that when the APA was adopted, most judicial review of the agency action occurred in regard to adjudication and substantive rulemaking was relatively uncommon until the 1970s. So the idea of using the APA to invalidate an entire rule may not have been specifically contemplated by the drafters, but I think Congress had intended to provide the courts with sufficiently broad remedial authority to keep up with emerging challenges. And in this instance, the "set aside" language of 706 is very broad anyway. So there's no real tension between the literal words of the act and the manner in which they have used it over the decades.


In my forthcoming article in the Notre Dame Law Review which has been mentioned and also an earlier piece I wrote in the Chicago Kent Law Review, I reviewed a host of situations in which the APA has been construed creatively to meet emerging challenges even if these holdings departed somewhat from the original expectations of the drafters. In other words, originalism is not a very common phenomenon in APA jurisprudence. So in particular, as rule making has emerged as an important force, courts have recognized the need for controls over the rulemaking process.


So standing has been broadened beyond what the APA text says, we have the hard look doctrine which was certainly not commonplace in 1946. The practice of requiring that the factual premises of a rule have to be based on the administrative record, the idea that agencies have to reply to material comments they received during the rule making, those things are not in the text. They were not in the original expectation, but they've come to be accepted as part of our culture and administrative practice. Now I see vacatur as being in a very similar light.


In fact, I go on to argue in my paper that a court's ability to nullify a rule in its entirety is a practical necessity. Why is that? Well, think about applying an injunction that applies only to one plaintiff. In an extensively regulated industry governed by a host of complex rules, if a single company -- one pharmaceutical manufacturer, one airline, one pipeline company obtains judicial review and is told "All right. You don't have to comply with the rule but nobody else does," the result is -- "but everybody else has to continue to comply," the result is chaos. If the rule is going to be revised, it has to be revised to apply to all the similarly situated companies in this type of situation.


And furthermore, the reviewing court is not entitled to specify just how the rule should be revised because that would invade the agency's responsibilities to execute the law. So we need to be talking about a remand, not piecemeal relief for an individual plaintiff. And the rule is either going to be remanded or not remanded. So the courts have very sensibly interpreted section 706 to allow vacatur. And it's a basic premise of administrative law practice. In cases that are familiar to us all as foundational cases in administrative law Chevron, State Farm, Brown and Williamson tobacco case, American Trucking, Regents, the recent DREAMers case, all of these presuppose that if the challenger wins, the rule would be set aside in the sense of nullified. So it's not just the D.C. Circuit, this is part of what administrative law practitioners have accepted for a long time. It would be deeply revisionist to depart from it.


Now I want to add that I think courts have and they should have discretion to decide whether and when to issue a vacatur or a nationwide injunction. Admittedly, the language of section 706 seems to present a problem in that regard because it says, "The court shall set aside agency action as one of its review criteria." But the answer I've given in my work is that the word "shall" should not be read literally. The cases don't read it literally. 706 is seen as a declaratory provision that sets forth broad principles but leaves it to the courts to flesh out what the language means and it's consistent with that to treat it as authorizing vacatur but not commanding the courts use it in every situation. And I've got some ideas about how courts can use that discretion to tamp down the ill effects of nationwide injunctions, but we can talk about those in the discussion period. But for now, I'm just going to stand on my claim that vacatur is at least lawful under the APA.


Judge Steven J. Menashi:  All right. Thank you very much, Ron, for an interesting discussion. I want to remind everybody in the audience that if you have any questions for our panelists, please use the question-and-answer function and write the question in there and I'll read it for the group. But let me ask -- I had this question just to John who seemed to say it's possible to understand 706 as providing some remedies in some sense. So 706 does also say, in addition to the "set aside" language, that a reviewing court should compel agency action unlawfully withheld. Does that not suggest that 706 is about remedies? Is that not a remedy that you would get in an APA challenge?


Prof. John Harrison:  Good question. First, that appears in section 7061. The "hold unlawful and set aside" language is in section 7062. 7061 is about agency inaction. 7062 is about agency action. I think that structural difference gives a reason to think that section 7061 does have something to say although it's completely generic. It just means give whatever affirmative relief is appropriate. 7061 has a reason to address the question of remedies on a very high level of generality because it's reiterating a point that might otherwise not be clear, which is that should a court give any affirmative relief at all? For example, when an agency has failed to regulate, should a court say, "Yes, the agency must regulate." That's a sufficiently questionable matter that I can imagine why Congress would want to underline that and say, "Yes, give whatever is the appropriate remedy."


I think there's a difference between section 7061 and section 7062 and even if there isn't, that reference to remedies in section 7061 is completely generic. It just means give whatever remedy is appropriate and indeed, section 703 names an appropriate affirmative remedy which is a mandatory injunction. That's language that appears in section 703 and not section 7061. And so if they are parallel in that respect -- I don't think they are -- then what I said is I think the second-best reading of set aside in section 7062 is appropriate which means it just means "and give whatever remedy should be given under federal equity under the special statutory review provision in habeas, whatever it is." So one way or another section 7062 is -- actually 7061 also -- neither one of them is adding anything to the law of remedies. For both of them, the actual law of remedies is found somewhere else.


Prof. Ron Levin:  Could I ask John a question following up on that?


Judge Steven J. Menashi:  Sure. Go ahead.


Prof. Ron Levin:  Under your second-best reading, why can't you treat that as meaning that if vacatur is the appropriate remedy, you can give it as part of whatever remedy is appropriate? It sounds as though that would be consistent with the reading that I've been proposing.


Prof. John Harrison:  Two points about that. Thank you, Ron. That's an important point. First, something I've been trying to emphasize and I'm going to keep trying to emphasize and it's important for the way this debate has developed because in my view, section 706 is not giving any new instructions concerning remedies and if [inaudible 25:17] a generic reference, that points us primarily to federal equity. And before the debate started about section 706, there was a preexisting debate, largely connected in its origin to Trump against Hawaii, about whether for example, universal injunctions were known at the time of the framing, whether they are an appropriate development of equity from the time of the framing. Those I think are the crucial questions. That's what courts should be -- and lawyers should be talking about is, is this an innovation in federal equity? If so, is it a proper innovation in federal equity? The discussion of section 706 I think is a distraction.


The other thing I'll say is that in a number of the forms of proceeding referred to in section 703, something like vacatur is simply irrelevant. A petitioner for habeas doesn't want anything to be vacated. The petitioner for habeas wants an order saying, "Let me out of jail." The damages plaintiff wants money damages. Somebody who wants a declaratory judgment, wants a declaratory judgment about that person's rights. So the idea of vacatur is relevant for equitable remedies and the question is, is that a proper use of federal equity?


Judge Steven J. Menashi:  Okay. I have a question from the audience which is about the language of the -- or what "set aside" might mean. And so the question cites Black's Law Dictionary from 1944 and it says, "To set aside a judgment, decree, award, or any proceedings is to cancel, annul, or revoke them at the insistence of a party unjustly or irregularly affected by them." And the questioner says, "It seems to me this is an argument against Professor Harrison's point that there is no evidence "set aside" meant to act upon the agency action itself. Sure, the definition suggests Abbott Labs overextended 706 reliefs but that isn't an intentional attack on "set aside," rather frustration with Abbott Labs. And so I guess maybe I'd ask two questions about that. So one is, is the problem the expansion of pre enforcement review from Abbott Labs?


And second, if "set aside" meant to annul or cancel a judgment, if at the time of the APA everybody expected agencies mostly to regulate through adjudication, is it just an emphasis on rule making now that asks them -- that is the cause of the problem, but if the statute is asking the Court to sit in judgment over our rule making the same way it would over an adjudication, doesn't that suggest there's a power to annul the underlying action which would be rulemaking. I hope that wasn't too complicated, but does anyone want to -- it seems like it's directed to John, first, and then we'll get reactions from everybody else.


Prof. John Harrison:  Let me start about that. And the first point I want to make about that is, yes, that is one of the things that "set aside" could mean at the time of the APA. It's one of the things that "set aside" can mean today and it's one of the things that "set aside" may mean in the various special statutory review provisions that use those words "set aside." So yes, that's possible. The question though about section 7062 is, what does "set aside" mean in a provision that governs the whole gamut of proceedings for judicial review and some of which that kind of remedy that might be thought to act on the agency's action is not relevant, like damages in habeas and declaratory judgments. So for structural reasons of the APA, that understanding of "set aside" can't be appropriate in section 7062 even if it's appropriate elsewhere, including in some special statutory review provisions.


I think it's also true that part of the story here about how that language got into section 706, originally 10E, of the APA is that when the APA was adopted, although there were rules, the focus was much more on adjudication. And it is possible to imagine a remedy in an adjudication that operates both on what the agency has done and specifically on the relations between the parties because when an agency does an adjudication, it binds that particular party and so dissolving the -- what's normally called an order -- that is to say, depriving it of its legal force might also not have effects beyond those of the parties. So it's, as you're suggesting, it's very important to keep in mind the difference between adjudications and regulations and the predominance of adjudications at the time of the APA.


I think it's also true that Abbott Laboratories has changed the focus of review of regulations through the pre-enforcement context and in the pre-enforcement context, a remedy like -- something like vacatur seems more appropriate and so I think the courts and the lawyers have tended to forget -- post Abbott Labs -- have tended to forget a lot of the other forms of proceeding and have forgotten the fact that section 706 has to be understood so that it can abide by all of them.


Judge Steven J. Menashi:  Any other panelists want to weigh in on these questions?


Prof. Ron Levin:  Sure. I can. Although maybe Beth wants to speak to this. If not, I will -- okay. Well, so in the first place, surely people have understood "set aside" to mean nullify in a number of contexts that are really not controversial. In the adjudication context which has been pointed out was the usual context in the initial decades of the APA, people understood that if the adjudicative order is found to be unlawful, the Court would nullify it. I think John has indicated in his writing, and I think maybe today, that in the case of special statutory review provisions for rules, "set aside" was understood to mean you can kill the rule and it's gone.


So what do we do then about cases that are reviewed under the APA but not under a special repeal provisions? I think that the adaptation to Abbott Labs has been a sensible one and given that Abbott Labs has been on the books for a long time and really is not controversial on its own terms, you would expect that people would apply the APA in a manner that responds to it. And if you have a pre-enforcement review proceeding where you're just asking the Court to get rid of the rule itself, it's pretty hard not to think that sometimes vacatur is the only sensible remedy.


Think about Chevron. The question is, can the EPA define stationary source to mean an entire facility rather than a particular piece of equipment? Are you going to say there's one answer for one company that sues and another answer for the company that doesn't sue? That would make no sense. They've got to treat it as a whole. And so it's been a necessary adaptation, I think. I also want to speak to the section 703 point. Section 703 says that the Court may resort to any applicable form of legal action including the particular ones that are listed there. Declaratory relief, injunctive relief, and habeas. It sure doesn't sound as though they're ruling others out. In the legislative history including what's glossed as "such as." So why can't it be the case that a remedy that wasn't specifically contemplated is now read into the APA? As I said before, originalism is just not a common feature of APA interpretation and I wouldn't expect that it should be in this instance.


Judge Steven J. Menashi:  Yeah. So Beth, what about that point? So Ron says that it was a necessary adaptation in the way the cases have evolved, that our practices have evolved in administrative law. I think based on the chief justice's questioning at oral argument, it would be disruptive to say that that's not what the APA means. So why are the benefits of the dis uniformity or the case-by-case adjudication -- why does that outweigh the disruptive effects of reconsidering the precedents?


Hon. Beth A. Williams:  Yeah. That's an excellent question. And I think Ron made a really good point at the outset that there's lots of different elements of current administrative law practice under the APA that are atextual and that have developed through common law and through judicial orders and maybe this is not that different. And maybe this power or this reading of "set aside" has grown up in order to provide a natural check post Abbott Labs and to match that type of pre-enforcement review. But practically, I think it's a question not only of what will be the implications if it is read this way but who's in the best position to decide what the implications should be?


So if a court adjudicates a claim against a single plaintiff and assume it's not in the D.C. Circuit and the court doesn't strike down the rule, is the court better situated to decide that the entire regulatory scheme is therefore unworkable and unenforceable or are the elected branches -- the Executive Branch in this case -- in a better position to decide how they will respond to this adverse court ruling and what the scope will be of how they're going to respond to the court ruling, whether they're going to basically not follow the regulation, to rescind the regulation, or to keep fighting in other courts. If they get an adverse ruling in the Fifth Circuit, maybe they'll continue to litigate in other circuits and to raise it up to the Supreme Court. But I think someone considering the implications would say that the elected branches rather than the courts are in the best position to decide that question of workability.


Judge Steven J. Menashi:  All right. Thanks. I have another APA textual question for John. And the question is, 705 authorizes courts to issue all necessary and appropriate process to postpone the effective date of an agency action and 55113 defines agency action to include rules. So if 705 authorizes courts to postpone the effective date of an agency rule, couldn't 706 refer to vacatur of an agency rule as well?


Prof. John Harrison:  Once again, I think the possible seeming parallel is not a real parallel. Yeah. Section 705 is about preliminary relief and hence is saying something about remedies. One interesting thing about section 705 is that it has a slightly convoluted language for all necessary forms of process which quickly became understood to mean an injunction, that is to say something governed by the law of injunctions. So even section 705, which does point to remedies, then points to the law of injunctions whereas section 706 is largely about another issue and it's about a very important issue. It is titled scope of review. It was titled scope of review when the APA was adopted. It still is today and that question that section 706 is about that section 705 is not about is of fundamental importance and again suggests that 705 and 706 are not that much parallel because 705 is on the topic that is much more about remedies than section 706 is.


Judge Steven J. Menashi:  Okay. Thanks. I have another question. The questioner says that several of the panelists have premised the arguments on the idea that vacatur is inherently universal and notes Jonathan Mitchell in his “Writ-of-Erasure Fallacy” article suggests the possibility of geographic severability. And the question is, can vacatur be subject to geographic severability in the same way that an injunction can and if not, why not? And that's related to another question we had which is, could it be that a federal court might vacate a rule but only within its geographic boundaries and maybe that would allow for some containment of the practice and allowance for different views? So I guess maybe I'll start by asking Ron because you said you had some proposals for how this should be handled and it does seem like you are -- that you recognize it shouldn't be done in every case. And so what about geographic limitations or other limitations? And I'll ask the other panelists if they have proposals along those lines as well.


Prof. Ron Levin:  Right. So I think I would start off by just saying on a semantic level, we would more likely refer to that type of relief as granting a limited injunction not a nationwide injunction. And there's no tension with the concept of injunctions there. Vacatur, I think, is more commonly described as "across the board" but that's only a semantic point. On the substance, yes, I think that there should be situations where a court should say, "It is feasible in this situation to limit the scope of our injunction to our circuit so the other circuits would have a chance to weigh in. We don't want to preempt them by giving nationwide relief." And where that is workable, I think they should do it because nationwide injunctions do have perverse effects in a number of ways. So that is a question of granting no broader an injunction than is needed which is a premise of equity.


Judge Steven J. Menashi:  And can I just ask, why is that a better approach, limiting it to the geographic boundaries of the circuit than limiting it to the parties that are before the court?


Prof. Ron Levin:  Well, there could be situations where that would do it, but then you will often have situations where the rule does not operate coherently because all of the players in the game are interacting with each other, competing with each other, or the agency is trying to manage a program. And if it's done totally retail, one plaintiff, you lose that effect.


Now what I do think you can do, and this I think is implicit in your position, is that if the agency adopts a rule and then applies it to a particular respondent, then the court can't say, "Okay. We're going to set aside the application in order to apply the rule just to you." And then be used to contain and fight on other adjudicative cases. That's what non acquiescence is and there's nothing wrong with that because it does give other circuits the chance to weigh in.


Judge Steven J. Menashi:  Yeah. Beth, I think you wanted to weigh in.


Hon. Beth A. Williams:  Yeah. I was just going to say, I think just to put a point on it, I think what we're saying is that vacatur is necessarily universal but injunctive relief is not necessarily universal. And so if a court is vacating something, they're vacating the rule as to everyone but they wouldn't have to do that. So if they're deciding the case just as to the parties before them, the operation of precedent would control anyway. And that would have natural geographic limitations. So even if you were deciding the case with an injunction just as to the party before you, that case went up to the Fifth Circuit and the Fifth Circuit agreed, the injunction by precedent would only apply in that circuit. Now it's different for the D.C. Circuit as I was talking about because anybody can sue the government there.


Judge Steven J. Menashi:  John, you want to say something?


Prof. John Harrison:  I want to say that Beth's point about precedent is extremely important and one practical application it has is that if this is a good way to go and it may be as a matter of federal equity, it's something the courts of appeal should do and not something the district courts should do because it's -- the courts have circuit precedent and the district courts don't do that. And as for nationwide uniformity and the practical imperatives that Ron has been talking about, let me point out that there are a lot of important areas like the interpretation of federal criminal statutes where there is not nationwide uniformity where it can be years where an important statute has a different interpretation in one court of appeals compared to another. That's unfortunate, but it brings with it the benefits of percolation and those are real. So yes, There is a practical case for nationwide uniformity but it's not a lay down, I don't think, because in other areas, the legal system manages with that in order to achieve other good things.


Judge Steven J. Menashi:  Yeah. So let's ask about the question of percolation and nonacquiescence because I have a question from the audience about inter circuit nonacquiescence. And so the question is, what bearing does this debate have on inter circuit nonacquiescence? The general understanding is that if say, the Second Circuit finds a regulation invalid, the agency might still enforce it elsewhere and give other courts of appeals the chance to consider and perhaps uphold the validity of the regulation and that process has systemic benefits including percolation. If one accepts that, does that have implications for whether courts can or should vacate a rule altogether? And Ron in his last answer suggested that inter circuit nonacquiescence is limited to cases where you can vacate an order applying a rule to somebody but is that a more limited conception of inter circuit nonacquiescence than maybe you would advocate Beth or John? I don't know. Why don't we say Beth?


Hon. Beth A. Williams:  Yeah. I mean, that obviously can play an important role, but I would just say that I don't know if you would need that if you read "set aside" in the way that Professor Harrison is explaining. I know the solicitor general, and the United States government is explaining. It's up to -- you can imagine in my previous role I was the chief regulatory officer of the Department of Justice -- you can imagine that the United States government -- there would be plenty of situations where they would choose a regulatory scheme or that they would feel that it was important to enforce it even if they didn't think so uniformly. And so I would think that you would want to allow as much percolation as possible and not have one district court judge be able to shut down everything immediately.


Prof. Ron Levin:  So you said as much percolation as possible. We just disagree on what's possible. Yes. So I think the traditional role of non-acquiescence is in context of adjudicative orders. So social security cases, immigration cases and tax cases and the like. Those are situations where the court isn't wiping out the rule, they're saying that the order that applied the rule is invalid and other circuits can take a different view because the rule is still on the books. On the level of regulations, if it's adequate to give declaratory relief, that wouldn't shut off other circuits. Also I have proposed in my paper that there should be a presumption that a nationwide injunction should be stayed pending appeal not instituted immediately. And the stay would give other circuits time to weigh in. So those are practical accommodations, but at the end of the day, there are times where the only meaningful relief is to say that we find this rule to be invalid and it has to either stand or fall for the entire industry. So our view is, we hold that it's unlawful. We set it aside subject to the checks that go with that like appellate review, stays, and the rest.


Judge Steven J. Menashi:  Could you explain that a little bit? So when is it that it might be enough to declare the rights of the parties as averse to a circumstance where you must declare the rule invalid?


Prof. Ron Levin:  Well, if they ask for a declaratory ruling. When you say there's no -- that the equities do not favor an injunction right now given the usual traditional formula about balance of the equities and the like. These are possibilities but I'm not trying to say that vacatur or nationwide injunctions don't have a place on the menu.


Hon. Beth A. Williams:  I would just say that I think another thing we can probably agree on is that relief can be fairly broad even if it's limited to the parties in the case because the principle is you give as much relief as necessary to give complete relief to the parties before you. So sometimes that is broader than applying to the particular plaintiff. Sometimes if the plaintiff is a state, in fact that relief is pretty broad and is still consistent with traditional principles of equity. I think in vacatur I could imagine -- I mean, the problem with vacatur is just it's so universal that it's hard to tailor it in any meaningful way.


Judge Steven J. Menashi:  Yeah. John.


Prof. John Harrison:  Yeah. First, I think what we're now talking about is, what's good federal equity practice? And I think that's the right question. I do think that for regulations as well as adjudications, there's a place for non-uniformity, for inter circuit nonacquiescence because of the benefits of percolation and the point I'll make again is that yes, there are costs to having what amounts to variation in the content of regulatory law from one circuit to another but the same thing often happens now with federal criminal law. Important federal criminal statutes can be interpreted significantly different in different courts of appeals. That's unfortunate but it does bring with it the benefits of percolation.


Judge Steven J. Menashi:  So I have another question from the audience. Can offensive non mutual, collateral estoppel solve the problem in district courts? That is if a district court finds a rule invalid and enjoins its enforcement against the plaintiff, can future plaintiffs come along to invoke that judgment in the same district court? Any takers?


Prof. Ron Levin:  Isn't the U.S. government protected from that by Mendoza?


Hon. Beth A. Williams:  I think it doesn't usually apply to the U.S. government.


Prof. Ron Levin:  So on some other panel that would be a great answer, but I think on this one, it doesn't work.


Prof. John Harrison:  That too would be a change here in the federal law of collateral estoppel. Right now it doesn't work that way against the government. I think that would be a major change, the kind of thing -- to pick up on [inaudible 49:43] -- the kind of change that should be done by Congress rather than by the courts on their own book.


Judge Steven J. Menashi:  Are we losing the -- so does the practice of the universal -- I take it that the opponents of universal vacatur would also say that we're losing the benefits of the prohibition on non-mutual offensive collateral estoppel against the government by having universal remedies, right?


Prof. John Harrison:  Yes. It's an end run around it.


Judge Steven J. Menashi:  Right. So if we recognize the Mendoza principle, is that an argument against finding universal remedies. How do you reconcile the Mendoza principle with your position, I guess, Ron?


Prof. Ron Levin:  Well, the legalistic difference is the difference between an adjudicative setting and a rule making one, but the broad principle is, yes, there's a tension between them because if a court vacates a rule and institutes it nationwide, then that's going to mean no other court can deal with it and that's a cost in terms of the potential benefits of percolation. So it should be rationed and limited to situations where you really need it. But I'm suggesting there are situations where you really need it because it would just have a great deal of confusion and inequity if you instituted it to fewer than all the relevant parties. I mean, if you have State Farm seat belt rule, if you really wanted an injunction that says, "Okay. So the policy holders of State Farm get to keep the Carter administration rule but everybody else is governed by the Reagan administration's rule." You just can't have a regulatory regime like that. So there are times where it's got to be all or nothing and if that means the first court triggers the dispute up to the Supreme Court, that is a cost.


Judge Steven J. Menashi:  Well, can I just ask about that also. So you say that you can't have a regime like that, but the agency would still be allowed to stop enforcing the rule across the board if they thought a patchwork rule was unworkable, right? So I think this is a point Beth made earlier. Isn't the question, just who should decide to what extent the regulation should be applied in a disuniform or a uniform manner? Is it the court or is it the agency?


Prof. Ron Levin:  Well, I do think you should give the agency flexibility in a number of ways. I mean, somewhat extent of my opening talk was that it is primarily the job of the agency to fix the rule on remand if the court has found an error in it. And so the court can't dictate what that's going to be. They have to leave it to the agency to do it. Also the practice of remand without vacatur, which I've mentioned, I have defended as being in part leaving it up to the agency to decide how to fix the mistake that the court has found on remand and that kind of flexibility is desirable to the extent you can do it. So I'm not indifferent to that at all, but sometimes if the court says, "Here's the law," that would foreclose the agency from deciding what the law can be otherwise. They can limit their enforcement, that's true. But they can't construe it differently from the way the court construes it.


Judge Steven J. Menashi:  I'd like to go back to an earlier point that came up. So in her opening remarks, Beth made the point that the government was making a statutory argument but not a constitutional one. That is whether Article III would allow Congress to provide for universal vacatur. So I'm curious about John's position on this. So is your position that the APA doesn't authorize universal vacatur but if Congress did authorize it, that would be a permissible use of Congress's authority or is there a constitutional limitation on this kind of relief?


Prof. John Harrison:  I'm not a hundred percent sure about what I think on that, but I do think this. I do think that there are probably reasons for Congress, if it decided to authorize courts to do this, there are reasons Congress might have to do that that probably would make it permissible under Article III although I'm not sure in large part because there's a longstanding equity principle that a broader injunction is permissible when it's necessary to give complete relief to the actual party. And that may be what's underlying the assumption in section 705 that stays of regulations generally operate as to everyone although I'm not sure that assumption has been thought through as well as it should've been. And so I can see the courts saying, yes, Congress has made a judgment here that this kind of broad relief is necessary so that actual parties will get what the parties are entitled to. So I suspect that's probably permissible. I haven't completely thought that through, so I don't really have developed --


Prof. Ron Levin:  Can I add one other thing which is that under special review statutes, Congress has for decades authorized a court to set aside a rule in its entirety if they find it to be illegal. And this was true well before the APA was adopted. There were cases involving the interstate commerce commission and the federal communications commission in which the legislation provided for it and the court said we would do it if the challenger wins. And these cases were called to Congress's attention when the APA was under consideration. With a hundred years of track record, I think it's a little tardy to suggest that it's beyond Congress's power to authorize a rule to be set aside in its entirety.


Hon. Beth A. Williams:  I think structurally though what you're seeing here from a larger perspective is as the regulatory powers grew, the regulatory check grew, and the check grew in the judiciary and not in Congress. And so you have potential legislation like the REINS Act which would provide checks on the regulatory state that are beyond the parties to a case -- that are to entire rules and not in specific cases. I think the constitutional question is interesting. The constitutional question was raised in the attorney general guidelines that were issued in 2018, and as far as I know, those guidelines are still in effect. That's one of the few things the Biden Justice Department has not rescinded, and I think those were probably guiding some of the arguments that were made along with Professor Harrison's academic work in the SG's arguments.


Judge Steven J. Menashi:  So we have a question about this constitutional argument, which is the following, Congress can assign certain cases to one court. The federal circuit has some cases that no other circuit court hears. Unless that's unconstitutional, how can it be that Congress lacks the authority to pass a universal vacatur statute? Would that be the same principle -- if Congress could restrict it to one court, why couldn't they authorize one court to give an answer for the country?


Hon. Beth A. Williams:  I think the argument is case or controversy in Article III. So even if Congress passed a statute, the argument would go, they're not allowed to pass a statute that conflicts with Article III, which limits the judiciary to deciding cases or controversies. I think that's what Justice Thomas was referring to in his concurrence.


Prof. John Harrison:  Let me say on that I have what might be a minority position. I think Congress can legislate concerning the rules of precedent in the federal courts, and so I think that Congress can say that if say one federal court of appeals makes a decision, that it's precedential for all the other federal courts of appeals. I think Congress can do that and many people I think would say Congress can't legislate about that. And so yes. That would be a way for Congress to do something a lot like vacatur by establishing say the first court of appeals or just one the federal circuit or the D.C. Circuit or any court to establish a nationwide precedent that, like changing the rules of collateral estoppel, would be a substantial change and one I think is for Congress and not for the courts themselves.


Judge Steven J. Menashi:  Okay. Well, we're in our last minute of this discussion so I guess I'll just ask, does anybody have any final thoughts that they want to add? If not, I'll turn it over to our host.


Chayila Kleist:  Thank you so much. Sorry to cut the discussion. I know it's been really good. On behalf of The Federalist Society and myself, I wanted to thank our experts for the benefit of their valuable time and expertise today. I want to thank our audience for joining and participating. We welcome listener feedback at [email protected] and as always, keep an eye on our website and your emails for announcements about other upcoming virtual events. With that, thank you all for being with us today. We're adjourned.