Interpretation and the Administrative State

J. Joel Alicea (00:00:01):

Good afternoon. Welcome to the American Enterprise Institute. I’m Joel Alicea. I’m a non-resident fellow here at AEI, and also an associate professor of law at the Catholic University of America’s Columbus School of Law, where I serve as the co-director of the project on Constitutional Originalism and the Catholic Intellectual Tradition or CIT, which is a co-sponsor of this event along with AEI. Because we’re a co-sponsor, let me say a little bit about CIT before we get started. CIT explores the relationship between the Catholic Intellectual Tradition and American constitutionalism. We do that through events like this one, fellowship programs for students and young lawyers courses offered at CUA and on many other activities. You can find out more about us at, sign up for our email list. So if you can know about events like this one in the future. Now this is one of several events that CIT has co-sponsored with AEI and we are, as always, very grateful to AEI for hosting us and for being such wonderful partners for these events. I want to especially thank you, Volavin, for serving as our host and Sophie Ruzzieri for helping organize the event.


Our topic today is interpretation and the administrative state. Let me just set the table for our conversation so that we’re all on the same page before diving in. The premise of our discussion is that how a judge approaches questions of legal interpretation might have ramifications for the scope of authority wielded by administrative agencies of the executive branch collectively known as the administrative state. And one example of those dynamics and the focus of our conversation today is the administrative law doctrine known as Chevron deference. We’re going to talk about several other administrative law doctrines but that’s just going to be probably the focus of our initial conversation about interpretation and administrative state. Now, Chevron Doctrine takes its name from the 19 84th Supreme Court decision, Chevron versus NRDC, which has come to be understood as having established a two-step process that courts are supposed to employ when reviewing an agency’s interpretation of a statute. At step one, the court should use all the traditional tools of statutory interpretation and ask whether after exhausting those tools, the statute remains ambiguous on the precise issue the agency was confronting.


If the statute is unambiguous, that’s supposed to be the end of the analysis, but the court should at that point just say what the law is. But if the statute is ambiguous, as courts often find statutes to be, a court’s supposed to proceed to step two of Chevron where the court defers to the agency’s interpretation of the statute so long as that interpretation is reasonable. Let me just quickly caution here. What I just said, I think, is a pretty conventional recitation of parts one and two of Chevron, but even the way I just recited that might be contestable and so I just want to flag that as I move through. In practice, Chevron often functions as a rule of broad deference to administrative agency interpretations of statutes, why latitude is given to agencies to act under Chevron. Now in a couple of cases called Loper-Bright and Relentless the Supreme Court, this term is due to decide whether Chevron should be overruled, which would ostensibly at least do away with deference to agency statutory interpretations and cut back on the scope of agency authority.


The court heard oral argument in those cases in January so this seems like a good time to assess the relationship between theories of interpretation and judicial deference to administrative agencies along with some other administrative law topics. And we have two of the best people possible to do that on our panel here today. I’ll briefly introduce them and then turn it over to each of them to give about 15 minutes of opening remarks. I’ll then facilitate a conversation between them and we’ll open it up to Q&A from the audience for the last 15 minutes of our event. And I have a note here that I believe means that if you want to submit questions for the Q&A portion of the event, you should use the hashtag AEIAdminLaw on X, formerly known as Twitter or email So let me just quickly introduce our speakers, our distinguished speakers, and I’m going to be pretty brief in my introduction so we can get to the substance of the event because honestly, if I read their full bios, we’d be here for a while. They’re very accomplished speakers.


I’ll start with immediately to my right, Jeff Pojanowski, who’s the Biolchini Family Professor of Law at Notre Dame Law School, specializing in administrative law and jurisprudence, clerked for Justice Kennedy on the US Supreme Court, for then Judge John Roberts on the DC Circuit, Princeton undergrad, Harvard Law School. And I would especially commend to your attention his article, Neoclassical Administrative Law in the Harvard Law Review, which I think does a deep dive into a lot of the issues that we’re going to be discussing today. I think AEI put that on the event page a link to that so I highly recommend that article to give you a better sense of some of these questions. On the other side of the stage, Kevin Stack is the Lee S. and Charles A. Speir Chair in Law at Vanderbilt University Law School. He also specializes in administrative law and in theoretical foundations of law. You can see that just there, the nature of their scholarly expertise lends itself to this kind of discussion. Professor Stack graduated from Yale Law School, Oxford, getting an MLitt from Oxford on a Fulbright Scholarship and from Brown University for his undergraduate degree.


He clerked for Judge Kimba Wood on the US District Court for the Southern District of New York and Judge Wallace Tashima on the US Court of Appeals for the Ninth Circuit. And so with that… Oh, and I should say, as to Professor Stack, I would highly recommend several pieces from him, but one that is short and that I really enjoy is Chevron is a Phoenix, which is excellent. I think that was in Vanderbilt Law Review. And then another one, Purposivism in Agency’s Interpretation… Is that basically the title? What’s the title of it? I want to make sure they get it.

Kevin Stack (00:06:37):

Purposivism in Agency’s statutory Iterpretation.

J. Joel Alicea (00:06:41):

Yes. I think that was it. So I recommend that one as well. Both of those articles are terrific and give you a lot of insight into, again, the nature of the debate that we’re about to have and the where Professor Stack is coming from in terms of his theoretical commitments and doctrinal commitments in terms of administrative law. So with that, I’ll hand it over to Professor Pojanowski.

Jeffrey Pojanowski (00:07:02):

Thank you, Joel. Thank you so much for having me here and thanks for everyone for listening. Catholic University Law School is doing some really exciting important stuff, especially with respect to the CIT and so it’s great to be here and I’ve never been in an AEI event before. This is very exciting. I’m waiting for my invitation to a Georgetown cocktail party now that I’m here. I feel like I’m in. And it’s great to be here having this discussion with Professor Stack. You’d be hard-pressed to find someone doing more interesting work on the intersection of interpretive theory in admin law or a more generous, kind scholar in the field. So it’s great to be here. So I guess we’ll talk a bit today about Chevron deference to administrative agency’s interpretations of law, and I guess in the future of the doctrine. As Joel indicated in his introduction, there’s widespread speculation that the Supreme Court is going to overrule Chevron. In fact, close court watchers noticed in what may or may not have been a Freudian slip. Justice Jackson referred to the doctrine in the past tense at an oral argument last week.


But now we may not want to get ahead of ourselves. People have been speaking about Chevron’s demise for the past five or six years. Everyone thought its cousin of a doctrine, Auer would be overturned, but Auer survived in wounded form so we don’t want to get ahead of ourselves. Nevertheless, Chevron does seem to be under siege, at least at the Supreme Court. The Supreme Court rarely gets to step two. It resolves most questions at step one, making Chevron alive at the court, but kind of a zombie. The situation is different at the lower courts, the Court of Appeals in the district courts, Chevron still lives on where courts are applying it, but the court’s example allows some judges, at least in the Courts of Appeals in the district court to ignore it or to minimize it by applying a very strong step one. Judge Ray Kethledge on the United States Court of Appeals for the Sixth Circuit recently wrote, I think in the Vanderbilt Law Review online, that he’s never reached step two.


So there’s an interesting question about why is Chevron under siege, why might it be overruled or if it’s going to be not overruled, why is it likely to be narrowed or why have courts fallen out of love with it. Now, one potential story we could tell is a crassly political story. You would say, look, Chevron became a thing in the Reagan administration when conservative judges were annoyed at the liberal DC Circuit getting in the way of the deregulatory of Reagan administration’s work. And they fell out of love with it when the Obama administration was using the administrative state to enact progressive policies. Now, I’m skeptical enough about human nature to never rule that out, but I think that that story’s probably a little too simple. First off, probably the Justice Scalia’s greatest defense of Chevron in the city of Arlington opinion took place five years into the Obama administration. And the real opposition to Chevron really started picking up steam under the Trump administration. So if you were a real hardheaded political opportunist, the criticism wouldn’t be happening when conservatives were in fact in power.


And there’s also an extent to which on one level, the doctrine is politically neutral. It helps conservative agencies when they’re in power. It helps liberal agencies when you’re in power. Any person who’s overruling it has to realize that their ox is going to be gored at some point. And in fact, probably the optimal strategy, if you’re politically really calculating, is to not overrule it, never apply it when your party’s in power and always apply it when the other party’s in power. We wouldn’t be talking about overruling it. So I think at least the raw politics argument may not be the whole story. So what else might be going on? Now, one of the occupational hazards of being a law professor is a willingness to believe that the power of ideas and intellectual trends can affect the real world. And another one is believing that your scholarship is accurately forecasted or explaining what’s going on.


I’m going to embrace both dangers and suggest that at least one vector in what’s going on with the courts falling out of love with Chevron is a form of interpretive formalism that’s become more prominent in the courts working itself pure. As professor Tom Merrill wrote a while ago, I think presciently, that maybe the seeds of Chevron’s demise were sowed with the rise of textualism and that although textualist defenders of Chevron, like early stage Scalia and Justice Thomas and Judge Silberman, just may not have appreciated the implications of their interpretive commitments. That’s the argument I’ll make. I’ll start by identifying how a form of moderate legal realism most plausibly underwrites Chevron’s jurist prudential or interpretive presuppositions. A touchstone here would be the great legal realist, Max Radin’s 1930 Harvard Law Review article called Statutory Interpretation, a nice field preempting title. Very good. So Radin argued, look, there are circumstances in which the text of a statute plainly decides a case. There are easy cases, he says, words are not portmanteaus, you can’t stuff anything you want into them. There are easy cases.


But he also said, look, there’s lots and lots of hard cases and most of the cases that get litigated aren’t the easy ones. So what do we do? What do interpreters have to do in those cases? And he said like contemporary textualists, he was a skeptic of legislative intent. We can’t look to intent. Like a good legal realist, he thought canons, especially linguistic canons or substantive canons were indeterminate and at best masked policy purposes. So he says, in the end, what you really need to do, or one thing you look at is you appeal to the statute’s purpose. And he says, purpose of interpretation, while it’s different than just doing whatever you want on a blank slate, nevertheless involve significant policy choices. And he says there’s two kinds of policy choices that purpose of interpretation requires. The first one is ascribing a purpose to a statute. For any statute, you can identify a number of plausible purposes and you can describe those purposes at various levels of generality. Statutes don’t come with purposes clearly stapled to them. So you have to pick among purposes. And he says that is inevitably a policy choice.


What’s the interpretation that makes the statute the best it can be? That’s policy choice one. Policy choice two is once you’ve selected a statute or purpose, which interpretation best advances that purpose? So in the relentless cases, the question is, whatever fees are necessary and appropriate for advancing the purposes of this maritime act? Would requiring the fishermen to pay for the fees advance the purposes of the act or not? That’s going to be an empirical policy judgment. It might, it might not, but you need to make some prediction about what the results of your interpretation are going to be. Now, if that’s what you think goes on with purpose of statute or interpretation and you’re a purposivist, Chevron makes a lot of sense. Now, not all cases are hard. There are easy cases where the law clearly speaks to what’s required, and that’s step one where the court resolves it. But then there’s hard cases, right?


And Chevron’s insight is, although the legal realists believe judges had to pick the purposes or judges had to figure out whether applying the statute would advance the purpose, Chevron’s insight is there’s someone else around to do that, and that’s an administrative agency. And if picking the statute’s purpose is a political choice, agencies are more politically accountable than courts. And if figuring out whether applying a statute would advance a given purpose, agencies are more technically adept at identifying what the likely consequences of the interpretation are going to be. So this moderate working form of legal realism can help shed light on the worries voiced by Justices Kagan, Sotomayor and Jackson at the oral argument at Relentless and Loper-Brite about how getting rid of Chevron is essentially going to make judges make policy choices on hard cases. Because in the end, filling in gaps in the statutes is inextricably intertwined with policy choices. Now, for a certain kind of interpretive formalists who are increasingly prominent in the federal courts, that structure and argument doesn’t make as much sense.


Textualists increasingly seem to believe that the original public meaning of text, canons, structure give much more guidance than the moderate form of legal realism would say happens with respect to statutory interpretation. They would conclude that resolving hard cases of statutory interpretation may have political implications but it doesn’t necessarily mean that’s a political choice. A 60/40- call on what a statute means is a lawyerly craft judgment about the likely inferences about legal meaning as opposed to whether you think that’s a better meaning or not. So although there may be judgment, it’s not the kind of judgment that you have with respect to policy choices. And that explains, for example, an oral argument of how justice Kavanaugh and Justice Gorsuch would press the Solicitor general saying, look, if we’ve done all the work that the footnote in Chevron tells us to do with all the traditional tools and we think one answer is better than the other, why defer? And that also explains why the Solicitor General had a hard time identifying or explaining or theorizing ambiguity.


She didn’t seem to want to give the legal realist answer to these formalists. So she had a hard time explaining what kind of ambiguity requires deference. Now, this kind of textualist formalist faith in the relative autonomy of law and craft doesn’t mean you believe law never runs out and it doesn’t believe that anything involving a statute has to be resolved by a court. As the petitioners in Relentless and Loper-Bright argued, and some of the justices, the Chevron skeptical Justices in the court indicated, sometimes you have statutes that say reasonableness or in the public interest. And there, the formalist might say that’s actually kind of an explicit delegation of policy-making discretion that only on the surface appears to be a question of statutory interpretation. There’s no legal tools to be brought to bear about what’s just and reasonable or what’s in the public interest or what’s reasonable unless it’s some legal term of art or something like that. And in those circumstances, what you’re doing is not Chevron or interpretation, but you’re doing a standard form arbitrary and capricious review.


You ask, did the agency consider all reasoned arguments adequately explain itself and give a rational explanation for its choices? And that explains an oral argument why the petitioners were trying to cordon off those cases in an arbitrary and capricious review and why Justices Gorsuch and Kavanaugh were keen to take it there. Now, this is controversial, right? The big question, the hard question for this formalist approach is, how do you sort questions that are lawyers’ questions into policy judgment questions? Because again, the statuary questions don’t come with a label affixed to them, right? And Justices Kagan, Sotomayor and Jackson are skeptical about our ability to do that. And I think any honest formalist has to admit there’s going to be hard questions where you’re not sure whether this is a lawyer’s question or whether this is in fact a policy judgment that’s been delegated to the courts. But I don’t think that recognition necessarily entails that you have to have Chevron deference. It’s possible that these cases occur on a continuum. There are clear instances of policy delegations, there are clear instances of lawyers questions, there’s going to be a twilight in the middle.


And the fact that there’s a twilight in the middle doesn’t necessarily mean you have to construct an entire doctrine of deference around those edge cases. You could start off saying, look, we’re going to handle lawyers’ questions and if we do the hard work and realize it’s actually a delegation of policymaking discretion, we kick into arbitrary and capricious review. So I think if you’re going to defend Chevron deference, I think you have to attack the formalism head on and be a straight-up purposivist or a dynamic interpreter or a more hardened realist, I think. So what would interpretation look like after Chevron if they were to overrule it? It’s possible that a part of it depends on what the court would decide to adopt, but it’s very likely that you would have courts sorting cases into two buckets. You’d have the lawyers’ questions bucket, you’d have the policy question bucket. Some form of nondeferential review would apply to the lawyers’ questions, whether it’s the pre-APA approach of giving some epistemic deference for longstanding or contemporaneous interpretations, or whether it’s something like Skidmore deference where you give respectful consideration to the agency’s interpretation.


Something along those lines will probably happen. And I think that’s probably sensible, in the same way that a federal trial court in Arkansas might look to the Second Circuit on a novel questions of securities law or a court may look to an amicus brief of an expert for advice while realizing the advocate has some skin in the game. It makes sense to listen to people who know things and especially in really hard questions. In instances where they feel like this is not a lawyer’s question, this is a reasonableness, just and reasonable, arbitrary and capricious review will likely apply. Now, part of my argument that I mentioned in the article Joel talked about is maybe arbitrary and capricious review should actually be a little more forgiving than the court currently has it. This would be the flip side of the interpretive formalism. If courts aren’t there for policy judgments, they should be less rigorously policing that the choices that the agencies make in the gaps. But that’s a question for another day. So in some sense, what it might look like is kind of what the court does now.


The court basically, the Supreme Court at least, resolves every case at step one when it thinks this is an interpretive question, and then it does arbitrary and capricious review when it doesn’t think there’s interpretive questions on point. And basically they would be telling the Courts of Appeals and the district courts to essentially do the same thing. So a last word on the politics of judicial review and administrative law. I talked earlier about how the crass political story about conservatives don’t like deference because it helps Democrats and vice versa may be a little bit oversimple, there may be differences in interpretive theories doing work here. And if the differences between formalists and realists, it’s not obvious there’s a strong political valence between whether you’re a formalist or a legal realist, right? Early textualists use the tools of legal realism to critique intentionalism and purposivism. And Ronald Dworkin, who’s probably the least legal realist person you can find, is not exactly a man of the right. So that doesn’t neatly cash out into political camps, although there is a kind of sociological story that most formalists today tend to be on the right, so to speak.


But that doesn’t mean I think overruling Chevron would be without any kind of political valence. If you think of the effects of repealing Chevron or overruling Chevron, you would essentially be shifting interpretive authority to declare what the law is to the courts. And there are political implications for that. Now, think of it this way. There’s a certain kind of conservatism that like stability, incrementalism and decentralization. And although while overruling Chevron may be destabilizing, a world without Chevron would take away tools from a centralized dynamic engine of policy change and move interpretive authority to scattered Courts of Appeals and district courts who are decentralized and whose institutional methods and procedures are much more incremental and wide sweeping. So for that certain kind of conservatism, the shift of power from agencies to courts may have some political implications. Now, progressives are not fans of that type of political conservatism and some conservatives who are worried about the concentration of private power and want to use the government to counterbalance that may not be fans of it either.


But I think that might be some of the political implications of overruling Chevron. So thanks very much, Joel.

J. Joel Alicea (00:24:16):

Great. Thank you, Jeff. Professor Stack.

Kevin M. Stack (00:24:19):

Great. First, Joel, thank you so much for the invitation. It’s really such an honor to be here and thanks for coming. I think this is a great time to engage these kinds of topics. I just want to first begin by emphasizing Joel’s praise for Jeff’s article, Neoclassical Administrative of Law. It’s one of those rare kinds of pieces of scholarship that its categories inhabit the way you think, you can’t really get out of them. And I just really couldn’t commend it enough. I think I probably have some disagreements with it at some parts, but I really think it’s extremely exceptional. So here’s maybe one way of thinking about actually, just hearing what Jeff says, is I’ve long joined those who’ve thought that Chevron only has one step, right? Chevron only has one step. So we can think of that that as just step one. But let’s just say Chevron only has one step, which is this permitted by the statute, is what the agency’s done permitted by the statute?


Let me just quickly defend why we should think about it that way, which is, it doesn’t make sense the way Chevron was set up to say, is the statute clear? Which is an interpretive question. And then if you say no to that, then ask again at a second step, is what the agency’s done permitted by the statute from an interpretive perspective? Why have these two different interpretive moments? Let’s just have one interpretive moment which basically ask, is this permitted by the statute? Is this permitted by the statute? And that’s not original to me, but I think that that is, in many respects, the best account of what the court has been doing or had been doing for a while as to Chevron. So what I’d like to talk about today a little bit is how could we ground a version of deference that’s slightly independent of the categories in which Chevron was working? I think that’s really important to our structure of government. So it’s going to move aside from the clarity and ambiguity framing which I think is somewhat distracting, and think about deference from another perspective.


So I want to begin in a really ground-up way by, let’s just ask how we think about the nature of the agency’s duties. How do we think about the nature of the agency’s duties under statute? And this is going to sound pretty simple and pretty general, but just hang with me for a little bit because I think that you’ll see some implications that [inaudible 00:26:35] so I think the agency’s duty is a duty to implement the statute, that is to further the purposes of the statute consistent with its terms. So it has to carry forward the purposes of its statute consistent with what’s permitted by the statute’s terms. And that seems pretty general. I think almost, we will see, incontestable, I don’t know, but pretty general. So the agency, just to say, just to get that in mind, so to do that, it has to engage in two interpretive moments. It has to ask, what are the purposes of the statute and what’s permitted? And these are both fundamental and necessary. It has to develop a construction of the purposes of the statute, otherwise its actions can’t be rational and they can’t be non-arbitrary.


We can’t think about whether what the agency has done is means end rational unless we have an understanding of what the agency was trying to do. You have to have that in there. So the agency officials have to have an idea of what their aim is. And of course they can’t do anything that’s not permitted by the statute, right? That’s uncontested in any account, any account of Chevron. So that’s pretty simple, right? And let’s see though, what I think follows from that. I think what follows from that, if that’s right, it doesn’t make sense for courts to ask, has the agency interpreted the statute in the way I think is best? Because that seems to me an odd formulation. On the one hand, it would seem to allow that the court is asking a wholly different question than whether the agency has complied with its duties. The duties of the agency are to have a valid construction of the purpose of the statute and ask whether what it’s done is permitted by the statute.


And if the court were asking a different question than that, it would seem like almost systemic form of sandbagging, which is like, nice that you’ve done your job, now we do something else. And that can’t be our legal system, I don’t think. So that maybe is a functional premise but I think maybe once a formalist could follow on there. So then how do we resolve this? What’s the court to ask? The court’s to ask, has the agency validly understand the purpose of the statute and is what the agency’s done consistent with what the statute permits? And that, I think should be the basic framework of thinking about a review of agency action. And I think that that is actually importantly deferential. It’s not deferential in the same way that Chevron is but I want to point out a couple of features of it first before I talk about how this, I think, could still be the law, even if Chevron is overruled. So first what I’m saying, I want to make sure it’s clear, and this gets to some possibly points of theoretical disagreement.


What I’m saying, the position I’m defending, is inconsistent with the idea that every time a court engages interpretation, it has to, it must find the “best interpretation” or the “most natural interpretation.” I’m taking the fair inside of that coin and I’m going to say the courts engage in interpretation for all different sorts of reasons. And we can ask courts to engage in interpretation to say, is this a permissible interpretation? I think sometimes we ask the courts to do that. And sometimes we ask the courts to say, is this the best interpretation of the statute? And they’re forced to do that. But the question of what the interpretation is for is not determined by the idea of interpretation itself, nor do I think it’s determined by the judicial role itself, nor do I think it’s determined by the Constitution itself. In fact, I think congress can direct courts to engage in interpretation for different purposes. So it could ask courts to engage in interpretation to ask, say, is this permitted by the statute? Which I think is the background idea in statutory interpretation.


And that I think relies on an idea of the necessary proper clause. So second, relatedly, I think the view I’m defending is consistent with the APA, right? It basically is an account of what the APA’s requirement, the Administrative Procedure Act’s requirement to resolve relevant questions of law, interpret constitutional and statutory provisions, and decide and overrule agencies if they’re outside of the scope of their statutory jurisdiction. I think what I’m saying is consistent with that. I’d be interested to see if others have different views. So third, I think the position I’m defending can be compatible with both textualist judicial interpretation and more purpose-inflected judicial interpretation. Okay, so let me just do this both ways. So a textualist could ask, using textualist methods, is what the agency’s done permitted by the statute? A purposivist could ask, is what the agency’s done permitted by the statute? I think in both cases, each interpreter has to be aware of tendencies and check them. So let me begin with the purposivist.


The purposivist has to check its tendency to say, when the agency has done something that’s consistent with the overall purposes of the act, that necessarily means it’s permitted, okay? A purposivist can’t help themselves to too much broad purpose at this question of what’s permitted. So just to give an example for the administrative lawyers around. So there was a case Babbitt v. Sweet Home where the court under Justice Stevens upheld the agency’s decision that habitat modification was permitted by the Injured Species Act and did that in part based on the idea that the Injured Species Act was meant to protect endangered species. There is some complicated textual limitations that were built into that. And maybe Justice Stevens helped himself with too much of the purpose to endorse the idea that what the agency did was permitted by the act. So position I’m defending, a purposivist has to check how much purpose can come in in determining what’s permitted by the scope of the statute. And so to a textualist, has to check their tendency to close off meanings through the application of textual tools.


So I don’t know, maybe we could say the most recent case, Biden v. Nebraska, which had to do with the student loans, was there too much of a tendency to look at the text as an opportunity to close off meanings as opposed to say what’s permitted by the text? So let me get to now my final point. So anyway, I think that this approach could be compatible with either. Final point, so what does this mean for the future of deference for agencies? So I think if, as it seems likely, the court is going to overrule Chevron at least in name, then it may come resort to some form of de novo or Skidmore review. I think that even if de novo in name or Skidmore review replaces Chevron, the court is still going to be thrust upon the structure of deference that I’m talking about. It’s still fundamentally going to have to ask the questions I’m talking about.


So let’s think about it. Suppose the court says we review agency action de novo. So they’re going to review agency action de novo so they’re going to have to isolate what are the interpretive acts the agency has done. Interpretive acts the agency has done is they’ve decided what the purpose of the statute is and they’ve decided whether what they’ve done is permitted by the statute. And I don’t think the court’s going to be able to, shouldn’t be able to engage in interpretation beyond those interpretive moments. And that involves then a lot of deference for agencies, I’ll finish up here, which is if… And the same could be said under Skidmore, and we can talk about that, which is it’s basically saying here’s a zone in which what the agency decided was permitted by the statute in a broad terms. And that’s significant deference to agency. And so frequently the court will then uphold agency action even though the agency didn’t interpret the statute in the way that the court thought would be the best interpretation by its own lights.


So that’s kind of a little different account. We’ll have to see how this marries with hard questions. I think there’s probably some hard questions about how this marries with hard questions, but there’s a little overview of some things I’ve been thinking about. So really the fundamental anchoring questions are purpose and admissibility.

J. Joel Alicea (00:35:29):

Excellent. Thank you, Kevin. Before I ask my own questions, I just want to see, Jeff, if you had anything you immediately want to react to or…

Jeffrey Pojanowski (00:35:36):

We’ll talk about it in the Q&A. Yeah.

J. Joel Alicea (00:35:37):

Okay. Both of you, I thought very helpfully identified various theoretical tensions that are present in this debate over deference, realism versus formalism, textualism and purposivism, or maybe I should flip those. And there’s another strain of argument about what role does the Constitution play in thinking about deference doctrines. So a lot of the arguments made against Chevron are based on the Constitution itself and whether the Constitution prohibits Chevron. But in these debates, all of these different pairings tend to get jumbled together and you hear arguments from different sides of those pairings that aren’t clearly identified as one or the other. And I was hoping that the two of you could disentangle these for us.


So to what extent is, and both of you did this a little bit in your opening remarks, but to what extent should we think of Chevron deference as necessarily traveling with, purposivism is an argument you were making, or rather realism is an argument you were making, Jeff, and does that also necessarily come with purposivism? Does it necessarily come with a view of what role the Constitution has in informing these debates or can some of these things be disaggregated from each other and you could be, as I think you were suggesting, Kevin, maybe a textualist but also have a pro-deference view on this. Could you all just disentangle these for us a little bit? I’ll start with you, Jeff.

Jeffrey Pojanowski (00:37:14):

Sure. It’s a great question and I’ll explain why they seem to run together, at least they tend to. But I’ll talk a little bit about why they may be sometimes separate. So two easy cases for why they run together. So imagine if you’re a functionalist in terms of constitutional interpretation. We’re going to make sure the government works well enough, the branches don’t dominate each other too much. We can work reasonably well together. We don’t care about what they thought in 1789 about the administrative state or something like that. So there’s no constitutional barrier to something like Chevron deference and you’re going to be keen on Chevron deference because you think that’s a reasonably good way of implementing statutes. So you could be kind of functionalist or kind of purposivist or a realist the whole way down. And then there’s the formalism all the way down. You can say, look, I am an originalist formalist about the Constitution. I’ve done the homework and my interpretation of the Constitution is such and I’ve got this strict understanding of separation of powers and Chevron is problematic because it essentially gives the judicial power to the administrative agency.


So I have a particular understanding of the original Constitution which is hard to square at the administrative state, which leads me to separation of powers, concerns about deference. And therefore Chevron is not only against the APA or a bad idea, it’s unconstitutional. So you can see the arguments have these often run together and there’s a logic to that. But I don’t think they’re necessarily logically entailed. So even if you are a constitutional formalist or an originalist or something like that, you got to do the work and show that Chevron in fact violates the Constitution. And there are originalists who think it doesn’t. John Manning, for example, thinks the constitutional structure suggests that we should have Chevron or we say the constitution is silent on it and we need to go to secondary sources or something like that. Or even if you’re an originalist who thinks maybe as a general matter, and I’d probably fall into this faint-hearted category, even if it’s the case as original matter, there’s problems with the administrative state or something like that, that ship has sailed.


And precedent is a thing. And the judicial power includes precedent and constitutionally blowing up the administrative state from the get-go is not the thing we’re going to do. And we may want to avoid making constitutional arguments about Chevron if there’s arguments on the ground about the APA or the nature of the judicial function that don’t require us to constitutionalize that. So I don’t think they necessarily run together. And one last point. What’s neat about Kevin’s argument is, look, let’s say the Constitution doesn’t say anything one way or the other about Chevron or it allows it and let’s say the APA or the best reading of statute says defer, there’s don’t be more Catholic than the Pope aspect of it. Courts are happy to defer on habeas questions with respect to Congress has told them to. If the original law of the APA says deference or if the best reading of particular statutes say only look for permissible, I think you have to do some pretty heavy lifting to say that’s unconstitutional or something like that. But the question is, do they in fact say that?

Kevin M. Stack (00:40:49):

Yeah, I think that’s helpful. I wonder, one of these questions implicated here is the scope of delegation, which is to what extent can you have different views about the scope of delegation and to what extent does that travel independently from Chevron? And here I think is a really important point, which is that to think about Chevron was really only about implied delegation of interpretive authority. It said sometimes we’ll infer or implicate a delegation of interpretive authority. And I think this sets up a really important point, which is that Congress often says the agency shall specify the meaning of a term. And I don’t view that structure, the agency shall specify the meaning of say, even stationary source [inaudible 00:41:37] said that. I don’t view the character of deference that would follow from that to be implicated in the Chevron debate at all, although I think it sometimes gets swept into it. Congress can say the agency shall specify the meaning of a term and presumably that would be an indication of you would give quite a bit of deference to how the agency specified the meaning of the term.


And that it should have persisted alongside Chevron and it should continue to persist alongside Chevron. I worry the court might write an opinion that actually overrides that. Now, if you had a different view about delegation and you thought it actually was not constitutionally proper for Congress to say the agency shall specify the meaning of an important statutory term, then you might have a different view on that. So I think it’s worth pointing out in some sense what’s the domain of Chevron is really about, it has been set up as ambiguity and implied delegation. But where Congress has expressly delegated authority, that I don’t think there’s any way in which the debate about Chevron should impinge.

Jeffrey Pojanowski (00:42:47):

Following on that, one way of thinking about the stakes of overruling Chevron is you have two overlapping circles. You can think of legal questions and then policy discretion or something like that. And the questions that, just and reasonable or in the public interest or the agency shall specify the meaning of the term, I think everyone agrees, or at least most people agree, who don’t have the non-delegation problem, that’s arbitrary and capricious review. And I think everyone also agrees that if the statute clearly speaks to something, that’s for the courts because it wouldn’t be permitted for the agency to do it. I think the big difference is what do you do with the overlap between interpretive questions that don’t have clear answers. And I think the Chevron critic wants to say lump those into the first circle with respect to treat those as kind of pure legal questions. And whereas although I think the Chevron defender would not say this is the same thing as arbitrary and capricious review, it would say we are going to defer if it’s a reasonable interpretation of the purpose or something like that.


So I think it’s about fighting over that overlap about which… Does that sound fair or…

Kevin M. Stack (00:44:01):

That sounds fair, which is in a question whether there’s a hard case, and I want to say the court would approach it with a generosity which would say, is this permitted by the text as opposed to what do we think is the best interpretation of the text. And then if they say it’s permitted, then we still have to engage in, a relatively robust or not, arbitrariness review.

J. Joel Alicea (00:44:22):

And Jeff, very quickly on what you just said there. Is the reason that critics of Chevron would want to lump that kind of middle category into the. Let’s make this an interpretation question. Is that motivated by a kind of formalism that distinguishes between the types of reasoning that goes into an interpretive question versus the kinds of reasoning that goes into a policy discretionary question? What’s motivating them to shift it over?

Jeffrey Pojanowski (00:44:54):

I think that would be the best justification for shifting it over.

J. Joel Alicea (00:44:56):


Jeffrey Pojanowski (00:44:56):

I think this is kind of Diceyan as opposed to Therian. We’d say, look, lawyers’ questions, or hard lawyers’ questions are lawyers’ questions, but if it’s not a lawyer’s question, let’s stay out. And so to the extent if we’re debating over whether a source is one smoke stack or the whole plant. And it’s not just and reasonable or adequate margin of safety or something like that. Say, look, that may be a hard question of statutory interpretation, but it’s a question of statutory interpretation as opposed to just and reasonable. So I think there’s this institutional role concern, I think.

J. Joel Alicea (00:45:33):

Yes, please.

Kevin M. Stack (00:45:34):

[inaudible 00:45:35] Just very, one small thing. And one thing I worry about in this regard, which is I can understand lawyers’ questions, which are in the domain of my permissibility column, is just because it can be characterized in interpretive terms can’t be enough. I can characterize almost any issue. I think a court can characterize almost anything in interpretive terms. So there has to be another disposition which helps to check it, I think, which is, it can be characterized in interpretive terms but what we’re asking is, is this within a zone, as opposed to what we’re asking is what’s the best reading of that, say if it was a tort statue or something.

J. Joel Alicea (00:46:11):

Yeah. You both have touched on in your remarks what happens if Chevron is overruled. And you both touched on it, but I want to just squarely present the question because it goes to your piece, Kevin, about Chevron is a Phoenix, where you argue, as you were arguing here, that something like deference will endure even if Chevron is overruled. Can you just elaborate on that point? And then I’d like to hear your response to that, Jeff, after Kevin.

Kevin M. Stack (00:46:41):

Yeah, so I guess I have maybe three thoughts. So the first one I’ve already mentioned, which is that deference, when there’s an explicit delegation to the agency to specify terms, that should endure. Second is the conceptual argument I was making in my open remarks, which is the structure of review, if we’re taking seriously the agency as good faith actors, is that we’re asking for them to act within a zone. And in fact, that’s the structure of our government. We want them to rely on some interpretive issues, but also policy and political preference all come into what we think are appropriate sources for the agency to rely on when it’s filling out what a statute means and what to act on. And I think that that’s part of the agency’s duty and we have to… I think it’s almost a formal argument for why deference will persist. And then there’s a functional point, which is what I put in the Chevron is a Phoenix paper as well, or in part what is in Chevron is a Phoenix paper, is that courts are just not equipped.


Courts are just not equipped to handle the level of technical detail and they’re just inevitably going to default to some kind of deference because these are not questions for them. These are the questions they’ll realize are so technical or they have a broad prevent reasonable risk kind of quality that they’re just going to find a way of deferring, whether they call it deference or not.

Jeffrey Pojanowski (00:48:08):

And I’ll answer in a line with Kevin’s three points. So I agree with him on the first. So if it says the agency shall define this term or it should be just and reasonable or in the public interest, absent someone bring up non-delegation issues, which I think we’ll talk about later, I think courts are going to defer, although there may be some temptation to turn those into lawyers’ questions. And I think there’s some reason to be cautious about that. On the second point, I think I’m a little bit more skeptical. I think if the courts were to buy your framework, then they would and they would do it. It’s like a formal argument to saying, Hey, regulatory legislation comes with a heart and sacks methodology attached to the rack. And if you buy that, you will do that. But I think a lot of judges won’t buy that. They would say, look, statutes are there. Yes, statutes have purposes but their purposes are done through particular means and the agency’s supposed to do it through their particular formal means, et cetera, et cetera, et cetera.


So I think a textualist, I guess you have to convince a textualist that these laws come out of Congress with these kind of interpretive instructions attached to them and that seems trickier. You have to convince them this is something like EDPA where you have to defer. And that might be a good argument but I think you’d have to convince them of that. On the third point, I think there’s a lot to what Kevin says. It’s very easy for the Supreme Court to never defer because they have what, 45 50, 60 cases a year now, and they’ve got an armies of law clerks and they can run everything down. But if you’re a district court who has to deal with agency challenges or if you’re the Court of Appeals who’s dealing with a whole bunch of those, especially if you’re not the DC Circuit who only has nine cases a month or something like that, you’re going to be really busy. And there may be a temptation to be like, eh, this seems good enough. Let’s just move on. This is a really complicated statute. This is really boring. I’m tired of these cases.


Let’s go and we’ll cite Skidmore instead. There’s going to be some temptation to that, but I think it nevertheless might change the dynamics. First off, it might change dynamics at the agency level. If you know you’re not asking, what can I get away with in Chevron as opposed to what are courts likely to say what this is going to be? That may change your menu of options and change the way you go about rulemaking. The decision of making environment may change. And I think just the default setting of courts may change as well. It’s not available to you to say, Chevron only has one step and I have to ask whether something’s reasonable. I have to ask the first question, do I think this is right? And so I think framing the decision making environment in that sense, I think, will nevertheless change. So you wonder if you’re going to have… So yeah, I’ll stop there. Yeah.

Kevin M. Stack (00:51:19):


J. Joel Alicea (00:51:20):

Oh, please.

Kevin M. Stack (00:51:22):

Sorry, Joel. That’s super interesting. It feels like maybe the point of disagreement is really on what I was calling my point to, which is, does a textualist judge have to think that statutes come with heart and sacks bundled in? Which in some sense, I am bundling away. And I guess I feel like no, a textualist judge has to just realize there are different kinds of statutes. There are different kinds of statutes. Some statutes are difficult and lots of statutes pose difficult questions and they require me, there’s no one else in the room to interpret it and we’re left, even if it’s ambiguous, to figure out what the best interpretation is. And some statutes ask you to do a different task. Maybe that’s the character if you’re [inaudible 00:52:11] they ask you to do a different task. They say, has another actor acted within their lane? That’s a structural premise. A structural premise. I think one… But you’re right, I would have to maybe defend to the textualist that that makes sense of structure.


And I think ultimately that what this reveals is the defense of Chevron that I’m suggesting, or of deference at least I’m suggesting, is really bottomed on a kind of departmentalism. Departmentalism says we the court should be respectful of the interpretive methods engaged by other institutions, Congress and the executive branch, and we give a margin of appreciation of both. And so ultimately I think that that’s right. And so then the question would be, then we’re next to whether a textualist can also be a departmentalist, and I think some are.

Jeffrey Pojanowski (00:53:01):

Oh, yeah, for sure.

Kevin M. Stack (00:53:02):


Jeffrey Pojanowski (00:53:03):

And this is really helpful, actually. I’m understanding your argument even better now. I think another way of characterizing the difference would be… Kevin sees, look, regulatory statutes are just slightly different beasts than other types of statutes which come with a particular type of interpretive method. I think the Chevron critic that I’m describing would say, look, I’m not going to treat statutes wholesale. There are parts of statutes that are deferential zones and parts that are not and we’re going to have disaggregate the statutes into lawyers’ questions versus non-lawyer questions as opposed to taking them wholesale. I think.

Kevin M. Stack (00:53:35):


Jeffrey Pojanowski (00:53:35):


J. Joel Alicea (00:53:36):

Just a reminder, if you have any questions you can tweet them at, or I guess not tweet anymore. You can use X with hashtag AEIAdminLaw or email them to But I want to pick up on a question or an issue that you teed up actually, Kevin, which was the interaction between the Chevron debate and the delegation debate. So in my view at least, there are really three major administrative law doctrines that interact to form the modern administrative state. There’s broad delegation of lawmaking authority from Congress to the agencies, there is for some agencies, so-called “independent agencies”, insulation from, one could say political accountability, one could say political interference, through removal protection. And then there is broad deference to the agencies through Chevron deference. And if you combine those things, agencies has to have lawmaking power, insulated from political interference/accountability and have a lot of deference from the judiciary. That’s a pretty powerful actor and that is the modern administrative state.


But let’s leave aside for now the removal question because I think you’re right that, in what you were saying, the delegation has much more relevance to this issue about Chevron and deference. So if the court overrules Chevron, as you said, it’s not clear yet what the court will do, as Jeff said as well, but if it did, what implications, if any, does that have on the delegation side what the court would have to do or would be confronted with on non-delegation questions? Does it depend on the way in which they write the opinion overruling Chevron? I’ll start with you, Kevin, and then Jeff.

Kevin M. Stack (00:55:29):

Okay. Yeah, I was super interested to hear what Jeff has to say.

Jeffrey Pojanowski (00:55:32):


Kevin M. Stack (00:55:34):

I think it’s going to depend. I think atmospherically… Here’s one just atmospheric cultural take, which is that if they overrule Chevron, then that would maybe take some of the motivation and a little bit of the gas out of the energy behind a more strict delegation doctrine. We’ll say there’ll be a sense that agencies are more checked by the courts, and so that’ll be a substitute mode of accountability that the non-delegation doctrine would’ve served. So I think that that’s a pragmatic dynamic. I don’t know if it’s a… I don’t think it’s entailed. I think that you could have a different view on deference and delegation. I think you could have a view that courts shouldn’t be deferential and that they have a broader narrow delegation. I don’t think it’s entailed. So that’s a first thought. Super interested to hear what Jeff has to say.

Jeffrey Pojanowski (00:56:27):

Yeah, so you could imagine a certain kind of argument, especially if you’re inclined to rule on the constitutional grounds on Chevron, you might be also willing to say, okay, this is the first step. The next step is the non-delegation doctrine. If you believe the original Constitution believes in strict separation of powers such that deference is giving the agency judicial power, you could say, look, it is just bad to give agencies legislative power. So Chevron is the first one and we’re going to keep moving. I didn’t get much sense of that at oral argument. I listened to a good chunk of the oral argument but I also did a quick word search, and there’s no discussion in the non-delegation doctrine, including from Justice Gorsuch, who seems to be one of the biggest revivalists. Maybe he’s just lurking in the tall grass and waiting to pounce next, but I don’t think… Here’s why I’d be inclined to at least, a person who’s skeptical of Chevron for the reasons I would be, would also not be enthusiastic about the non-delegation doctrine.


One of the worries about the non-delegation doctrine is the manageability of it. How much delegation is too much? And you see, Justice Scalia would worry about this, a lot of old school textualists worried about this. And if for the same reason, you’re inclined to say, look, I want to stick to lawyers’ questions. I don’t want to manage these kind of hard policy choices that don’t seem legally tractable, I’m not sure how excited you are to jump into the non-delegation problem and decide I’m going to go around drawing lines about how much delegation is too much. So I think the same reasons that might make you as a judge want to hold onto and be jealous with separation of powers reasons, hold on to “lawyers’ questions,” if you feel the non-delegation doctrine is not amenable to lawyerly craft, I don’t think you’re looking to pick that fight. And I think as Kevin suggested, maybe the pressure, this is kind of let the air out of the tire with respect to a little bit worries about the Leviathan or something like that.

J. Joel Alicea (00:58:34):

Yeah, please.

Kevin Stack (00:58:36):

Just picking up on the point Jeff made, the classic concern voiced by Scalia that the court has not felt often that it could second guess the permissible degree of discretion to agencies. And I haven’t seen articulations even from Gorsuch, Justice Gorsuch of the non-delegation doctrine that avoid that problem. So he has a couple of formulations, including coming back to the idea of contingency delegations. But contingency delegations themselves, there’s a lot of room as to what the contingency is. So contingency delegation basically says, if you find X, then you have the following power. And what’s in the finding, what’s in the contingency was historically quite broad. If you find the terms of trade were no longer reasonable or didn’t equalize trade between two countries, which is a complex judgment, then you have the following powers. So I have yet to see an articulation, even from a formal perspective, that seems to avoid this or [inaudible 00:59:33] and around this question of degree as to delegation.

J. Joel Alicea (00:59:38):

Yeah, I don’t think that his Gundy, separate opinion, Gorsuch’s separate opinion really tries to draw a clear line in that way.

Jeffrey Pojanowski (00:59:46):

It seemed more of like a call to scholars like, Hey, give us some ideas here, folks.

J. Joel Alicea (00:59:54):

But this is actually related to another question, which is the major questions doctrine. And to what extent is the major questions doctrine really just an adjunct to Chevron, a way to rein it in, such that if the court overruled Chevron, there would just be no need for the major questions doctrine or goes away. Or just in general, maybe putting it in more broadly, what are the implications potentially for the major questions doctrine, if the court does overrule Chevron?

Jeffrey Pojanowski (01:00:28):

I think it depends… The major questions as two kinds of things. One would be made obsolete by Chevron. You don’t get Chevron deference on major questions. And if we give it to Chevron deference, all that goes away. But then the other is outside of Chevron. And I think it depends how you frame what the major questions doctrine is, if it’s a substantive canon requiring a clear statement. So imagine if you were to overrule Chevron because you’re a textualist formalist. Textualist formalists are traditionally pretty uncomfortable with substantive canons. So a person who’s going to want to overrule Chevron on those grounds may not be excited about having this strong normative canon here on the table, at least at the pains of inconsistency. Now if you frame the major questions doctrine as a tool of ordinary interpretation, as like a descriptive canon, and that’s what you see from Justice Barrett in her concurrence in the Nebraska case and saying, look, this isn’t a clear statement requirement upholding constitutional values. This is just kind of a presumption about how language usually works.


We usually don’t hide elephants in mouse holes and you want to give me a good reason to do that or give me strong evidence that we do that. And I think that strikes me as something it’d be independent of whether you have Chevron or not. We say, look, we’re doing ordinary interpretation. And that would just be another question, another descriptive canon in the toolkit.

Kevin M. Stack (01:02:01):

Yeah, I think that’s interesting. I think this is pretty much in parallel, but maybe there’s a little difference, which is that if the decision in Loper is a broad overruling of Chevron that invokes an anti-delegation atmosphere, then I think that’s going to bolster the idea of major questions. If on the other hand they do a narrow sort of, Chevron was formulated awkwardly, this whole one step, two step thing was not so great and we want to inject a little more de novo review more clearly, then major questions I think will just sit out there separately. We could also talk, if you want, about to what extent the major questions doctrine is defensible outside of thinking of it as a substantive canon. But I don’t…

J. Joel Alicea (01:02:58):

No, I’m fine with that. If you want to do that, yeah.

Kevin M. Stack (01:03:00):

Yeah, to me it seems like it… So Justice Barrett is the one who’s on the forefront of this and she argues that no, this is just a mode of context. Mode of context. We bring context to bear on statutory interpretation. There’s a nice, I like to quote John Manning on this, even textualists look at context. But here’s my question. Here’s my question on that one, would be, Manning also makes the distinction, he says, and I thought this was really perceptive. He says the difference between a textualist and a purposivist is the kind of context that matters for them. And for textualists, what matters is semantic context. And for a purposivist, what matters is policy context. And if we could contest that but if that’s right, that it seems hard for me to defend the major questions doctrine as something other than looking to the broader policy context of the enactment, which then makes me think it’s what major questions doctrine is doing is bringing a purpose-based conception into interpretation. Now, we may or may not like that, but it feels like then it’s harder to defend as a textualist.

Jeffrey Pojanowski (01:04:20):

Mm-hmm. This is a really hard and interesting question, and I think part of it may be the blurriness between the policy context and the semantic context line. But let me give a stab at defending it. So we understand statutes in light of the context> We know why people passed the Affordable Care Act. It deals with healthcare and stuff like that and all those kinds of things, or we know the Telecommunications Act is roughly about, or something like that. You’d imagine some… And you can get a sense about how much big game this particular statute was seeking to hunt or how incremental it was or whether this is an ancillary provision of a statute or a core provision of the statute. And you might be able to make judgments saying, look, this seems to be… Any reasonable reader who understands this context, understands that this is kind of a narrow gauge type of statute, whether I like it or not, or something like that. Or this is a broad gauge, big time reform statute that wants to comprehensively occupy a field or something like that.


Or you could say, look, this seems like a ancillary provision that this seems to be being like an et cetera as opposed to a central part that completely defines what’s going on. And I’m not sure if that’s semantic context or policy context. I think you can draw an inference about whether someone’s trying to do something big or not, as opposed to doing something small or not, as opposed to whether what they’re trying to do is good or not or something like that. And again, maybe I’m a little squishy on this because I think people like Sam Bray are onto something with respect to the mischief rule or something like that. So I think I may be willing to let policy context bleed in a little bit more than others. But that’s my take on it.

J. Joel Alicea (01:06:22):

I suppose one way in which major questions could be affected by the overruling of Chevron, if Jeff is right that the agency’s approach to regulation changes, if you’re right, that the overruling Chevron makes the agencies think, okay, I need to hew more closely to what I think the courts will actually think is the best interpretation as opposed to what is permissible, then that could lead to less ambitious regulatory policy that could then implicate the major questions doctrine less often. I would think that’s at least possible.

Jeffrey Pojanowski (01:07:06):

Yeah, there’s a chicken and egg thing though. If they know the courts are likely to not find… If courts are willing to find elephants in mouse holes all the time, then they’d be willing to identify them. But if the court’s skeptical of that, if they have some soft form of the major question, that seems less permissible, I think.

J. Joel Alicea (01:07:26):

Yeah. Last question that I wanted to pose to the two of you before I open up all that… And there is one other thing that I might ask you about. But the last one that I had kind of thought through beforehand was this point about de novo review, which you mentioned, Kevin, that at the oral argument there was a lot of discussion among the justices about to what extent is Chevron inconsistent with the idea of de novo review of statutes and that if we overrule Chevron, that means that we’re now doing de novo review of statutes, statutory interpretation questions. But there there’s this argument that has been made before by Henry Monaghan and others. I think Adrian Vermeule raised the point both before and after oral argument as well, that that’s really just not helpful in thinking about the Chevron question is to frame in terms of de novo review because it could very well be that the statutes themselves require deference to the agency under a de novo standard, so this isn’t really helping us think clearly to frame the issue of Chevron as de novo versus deference.


What are your thoughts about that, Kevin?

Kevin M. Stack (01:08:36):

I guess, yeah, I think it’s an interesting question. In a way, it’s de novo about what. De novo about what. So I think that if you think that de novo inherently means that the judge has to decide any interpretive question has to be decided what’s the best or most natural reading of the law, then that’s quite a broad view. But I think you could think de novo review, you could say de novo review of what’s permitted seems like at least conceptually coherent thing to say. It doesn’t mean there’s not going to be hard questions about that. So I think that there’s a way in which they could come up with de novo review that doesn’t mean that only agency action that will be affirmed is agency action which fits within the court’s best interpretation of how the statute should be construed wholesale.

Jeffrey Pojanowski (01:09:26):

Yeah, the structure to the argument would be something like this. So the court should follow the law and if the best reading of the law says to defer, deferring follows the law. And saying following the law is de novo begs the question. And I guess the response to that would be like, okay, but prove. Making that argument doesn’t prove that deference is in fact is the default. You’re going to have to make an argument about why it’s reasonable to assume review of questions of law builds in that deference. EDPA says it clearly. APA is more contestable or making that presumption with respect to all regulatory statutes is contestable as well, especially if many of the people who make that argument also agree that, look, this implied delegation is a fiction, but it’s a good one. And then we’re just arguing about fictions based on turn of what’s the best way to running judicial review systems.

J. Joel Alicea (01:10:33):

But couldn’t the response by someone like Monaghan or Vermeule to what you just said, Jeff, be to say, why is the burden on us to prove that there is such a default rule of delegation? Why isn’t the burden on you to prove that there isn’t? Haven’t you implicitly adopted a default that you haven’t proven which is that there isn’t this form of delegation, that either way there’s some default going on that has to be justified? The major questions doctrine could be seen as putting in place that kind of default. So what’s the basis for it one way or the other?

Jeffrey Pojanowski (01:11:11):

Yeah, and this, I think, goes down to, I guess the disagreement between me and Kevin or the formalist in Kevin to be like, look, judicial role is, unless you’re told otherwise, to give your best answer of what you think the law is. And if you’re going to be deferential, we’re going to need some reason for that. So I guess it would be kind of like a Marbury type presumption and you’d say, give me some reason to suggest why I shouldn’t do anything besides what I think the best answer is, I guess.

J. Joel Alicea (01:11:41):

And this is getting to this, as you were saying, Jeff, this is, I think getting to this disagreement with you, Kevin, about you seeing agency statutes as really just a different breed where the default might be different there…

Kevin M. Stack (01:11:54):

That’s right.

J. Joel Alicea (01:11:54):

… than another statutory interpretation domains.

Kevin M. Stack (01:11:56):

That’s right. Because Congress has made a choice to best lawmaking authority in another agency. And I guess the thought is that’s for a reason. It’s for a reason that they’ve chosen to capitalize on the expertise and political accountability of this other body. So to me it seems like it wouldn’t make sense that we say, go do that thing that Congress has asked you to do but watch out for our sandbag, it’s coming. And so it’s implied in, we’ve given you that authority, so we should ask, are you in your lane? Those are lawyers’ questions to be sure. Are you in your lane? But basically the position I’m suggesting really is that some level of deference is implied by delegation. But it’s implied by delegation not in the Chevron way exactly.

J. Joel Alicea (01:12:49):

But I assume, or maybe this is a bad assumption, I just wonder, maybe I should just ask you, Kevin, that that could be overcome, that default could be overcome in a particular context. So I think you, Jeff, in your article, argue, or I don’t recall if it was your own argument or you were serving scholarship on this, that the historical context of the APA’s enactment could be read to suggest no deference, at least under 706.

Jeffrey Pojanowski (01:13:18):

Or reduced [inaudible 01:13:19]

J. Joel Alicea (01:13:18):

Or minimal deference. Yeah. So if that were true, if that history were right, do you think then, Kevin, that could overcome this default that you perceive in agency statutes?

Kevin M. Stack (01:13:33):

Yeah, I think it’s, to me… Yeah, Congress I think could overrule the default I’m suggesting. Whether it’s done so in the APA is a contestable question, especially given that Congress was familiar with agencies and understands what they’re asking of agencies. But yeah, I think this is a default that could be overcome and it would be a little bit, there may almost be arguments where Congress is trying to do, which wouldn’t be surprising, two different things at once. It wants to vest agencies with authority. Does it all the time or has it all the time. At the same time, it also gets excited about the role of courts. So it may not be integrating these thoughts that well. But yeah, I would concede that Congress could overrule this default and it may say, as to some issue, we want the court to determine the best interpretation of stationary source. And maybe there are reasons because we’re worried about capture or something else. There could be reasons for that specific default.

J. Joel Alicea (01:14:44):

Yeah. And the question then would just be whether it has done so. Right?

Kevin M. Stack (01:14:48):


Jeffrey Pojanowski (01:14:50):

To speak against interest for the deference skeptic, so let’s say people like Aditya Bamzai or Mike Rappaport are correct that the best interpretation of the APA is not de novo, plain de novo review, but a non-deferential, if there’s a contemporaneous interpretation or a longstanding interpretation that gets some epistemic deference, which is much weaker than Chevron. Let’s say they’re right about that and that’s the law of interpretation that the APA creates. I’m not sure that automatically defeats your argument because it’s the same kind of paradox about rules against statutory interpretation. So Larry Alexander’s got this great article called Mother May I? The Imposability of Rules of Statutory Interpretation. So if I were to say, Hey, listen to nothing what I say, unless I say the words, “Mother may I,” beforehand. And if I walk in and I’m on fire and I say, water, water, and I don’t say, “Mother may I,” you just stand there with a bucket of water, you’re not being a faithful agent because at this point, I don’t care about, “Mother may I.” I want you to throw water on me.


It’s possible that Congress with the APA says, we presumptively don’t want deference. But let’s say in the 1970s, it’s enacting all these statutes in the best interpretation of the statute is actually whatever we said in the APA, right now, we actually want agencies to do this kind of thing. So at best, the APA would be a default that could be, and it’s very possible that with later statutes, they weren’t intending something as deferential as the APA. So I’m not even sure the APA can necessarily be a silver bullet against you.

J. Joel Alicea (01:16:20):

Although that would still at least clarify what the default was.

Jeffrey Pojanowski (01:16:23):


J. Joel Alicea (01:16:24):

And that default would be not the Chevron default.

Jeffrey Pojanowski (01:16:27):

That’s true. Absolutely true.

J. Joel Alicea (01:16:28):

Okay. Let’s open it up to questions from the audience if there are any questions.


Adam White, our first question from our admin scholar here at AEI.

Adam White (01:16:39):

I learned from everything from Professor Stack and Pojanowski, but your discussion about different kinds of statutes reminds me of the quip, maybe Frankfurt or [inaudible 01:16:47], that sometimes Congress just expresses a mood. And you kept returning to that class of statutes that’s the hardest for all of this. The public interest, just and reasonable, public convenience and necessity. And those statutes were largely written a century ago for multi-member independent commissions. And I wonder, to what extent can the discussions around deference and delegation treat those statutes differently not just because they’re being administered by multi-member commissions rather than one headed agency, but because those really were intended to be quasi-judicial, quasi-legislative commissions in the sense that they were supposed to go about their work with a different mindset and maybe with different tools than the speedy energetic executive agencies?


I know for the last 50 years or a hundred years, a lot of people have said that’s a phony distinction, agencies are agencies, but they really weren’t at the time that those statutes were being drafted. So I guess the short question at the end of a long statement is just, is there a way that the deference doctrines should approach those statutes in a way to really shape the mindset and approach and toolkit of the agency and not just what comes out at the end of their process?

Jeffrey Pojanowski (01:18:13):

That’s a great question.

Kevin M. Stack (01:18:15):

That’s a great question.

Jeffrey Pojanowski (01:18:16):

That’s a great question and I haven’t thought about it this way. One, my first thought is I’ve been reasonably skeptical of hard-look in arbitrary and capricious review. The internal Diceyan in me wants to separate functions as like, I don’t know why the DC Circuit should check FERC’s homework so closely or decide really get on the EPA’s case if they don’t want to do cost benefit analysis if the statute’s silent on it. But if these statutes were written with a certain deliberation in mind and we need some sort of compensating doctrine, some sort of administrative law of the second best, that might be some justification for a more robust reason to explanation requirement. Not the hard look where you say, I don’t know. I would’ve done something different in terms of policy but yeah, make them answer things more. Make them respond to questions better. Make them show their homework. And again, it leads to the temptation of saying, I’m going to grade them too hard and I’m going to be deregulatory or I’m going to be more regulatory.


But that can maybe, that would change maybe the way I think about hard look review, at least.

Kevin M. Stack (01:19:41):

I think it’s a really interesting question. In a way, my particularism about statutes to say, regulatory statutes are different. You could also translate your question as, let’s go one step level down and are different agencies different? And different agencies are different. And so I don’t know whether… Maybe we could ask whether we really need a totally trans-substantive doctrines of deference. We should understand there’s a law, there’s FDA law, there’s FTC law in some level. I think that some of that hemming in just goes hand-in-hand with the fact that there are a lot of judicial constructions out there, what the statute meant. And some of those with early statutes do hem in certain ways. So I don’t know whether that should make us more or less willing to overrule old precedents that hem in, say, old agencies, but those seem like that’s one thing I’d want to be thinking about as to how they’re different.

J. Joel Alicea (01:20:36):

Other question here. Yes.

Speaker 7 (01:20:42):

My question is related to the one you just answered but it’s a little bit different. For those of us who are not expert in this era of the law and recognize its importance, one of the more useful things that was done at the beginning of this discussion was a taxonomy about the politics of this. In other words, with deference, the decisions are being made by commissions that are political and changed with administrations. Without it, they’re being made by the courts, which are in most cases, less responsive. But I think the question I have, which it builds on the past one, of them is factual. Again, for those of us who don’t follow this, and that is how much of regulatory and administrative laws actually made by commissions where there’s a three to two vote?


And if you don’t like the rules that come out, then elect a new president and then it’ll be three to two in your favor so that it is in a sense, responsive versus how much of this area is made by a civil, or for that matter, military employees, civil servants or military servants who are not elected by anybody and not even nominated and confirmed by the Senate but just completely unresponsive? And do the arguments around this debate distinguish between those agencies that are purely administrative and those agencies that are, shall I say, political or made up, in many cases, actually, of politicians. Thank you.

Jeffrey Pojanowski (01:22:30):

That’s a great question. So the two biggest distinctions that we make in administrative law are between the commissions, which are usually five commissioners, three to two in terms of politics, and who serves set fixed terms and are protected from removal. And then cabinet agencies where the head of the agency is removable by the president. So there’s a sense which cabinet agencies are more directly politically accountable. If Biden doesn’t like what the head of the EPA is doing, you can go. It’s harder for a president to control the Federal Communications Commission or the Federal Energy Regulatory Commission. There tends to be three to two, it staggers. So there’s a sense in which they are more insulated, insulated from politics. So the commissions are actually more insulated from politics than the cabinet agencies. And civil servants in those agencies can make lots and lots and lots and lots of decisions, but there’s usually some route of appeal up to either the agency head or to the commission. The main things where you don’t have as much direct review, I don’t know much about appeal of things within the DOD and military stuff, but usually most…


My sense is most agency decisions, whether they’re multi-member commissions or cabinet heads, some unelected civil servant or bureaucrat will make a decision but there’s always going to be appeal to some responsible political officer up the chain. But an interesting point is the deference doctrines don’t seem, and I could be wrong in this, but don’t seem to make a distinction between independent agencies and cabinet officials for deference. And maybe it’s not obvious that they shouldn’t. If one’s more politically responsive than the other and if one of the justifications for Chevron is political responsiveness, maybe the case of it for multi-member commissioners is weaker than it is with respect to cabinet heads. I don’t know.

Kevin M. Stack (01:24:36):

Just to add onto that, I agree with everything that Jeff said. I don’t know of any agency itself, anybody other than an agency itself that can engage in rulemaking at least. So that’s always going to be, whether it’s a multi-member commission or the head of the EPA or the EPA itself that’s engaging in rulemaking, and there’s different levels of responsiveness there. Yeah, And ask the question about who should get more deference. I think the law treats them uniformly. There was a opinion by Justice Roberts in the DC Circuit many years ago with the question of, how much deference should the president get when the president’s acting under statute? And you could argue, this should be… You could go both ways. You could say, this should be heightened deference because it’s obviously the height of political accountability. Or you could say, no, in its respect, the president’s like a normal agency and we should apply our norm. This is an administrative action so we should apply normal doctrines of administrative law there.

J. Joel Alicea (01:25:33):

Might have time for one more question, if there is one, or we could just wrap up if anyone has a final question.


Okay. Before we wrap up, I do want to just say it because I botched this at the beginning, and it has proven very important to this conversation, that Kevin’s piece is Purposivism in the Executive Branch: How Agencies Interpret Statutes. That was Northwestern University Law Review. I just want to make sure to get that out because it’s a very worthwhile piece and it’s one that does, I think, has related very to our conversation today, especially this whole question about whether agency statutes really are different in kind from other statutes with respect to interpretation. So please join me in thanking our two guests.

Please enable JavaScript in your browser to complete this form.
How did you hear about this event?