doctrine or keep it.

chief justice roberts: we'll hear argument first this morning in case 22-1219, relentless versus the department of commerce. mr. martinez. oral argument of roman martinez on behalf of the petitioners martinez: mr. chief justice, and may it please the court: for too long, chevron has distorted the judicial process and undermined statutory interpretation. it should be overruled for three reasons. first, chevron violates the constitution. article iii empowers judges to say what the law is. it requires them to interpret federal statutes using their best and independent judgment. chevron undermines that duty. it reallocates interpretive authority from courts to agencies, and it forces courts to adopt inferior agency constructions that are issued for political or policy reasons. in doing so, chevron blocks judges from serving as faithful agents of congress. it mandates judicial bias and encourages agency overreach. and by removing key checks on executive power, it threatens individual liberty. chevron also violates the apa. the most straightforward reading of section 706 requires de novo review of legal questions. congress put constitutional and statutory interpretation on equal footing, and it required independent legal judgment as to both. as justice scalia wrote, the apa's text contemplates that courts, not agencies, will authoritatively resolve ambiguities in statutes. and, third, this court's only justification for chevron is the implied delegation theory, but that theory is a fiction. there's no reason to think that congress intends every ambiguity in every agency statute to give agencies an ongoing power to interpret and reinterpret federal law in ways that

override its best meaning. in this case, the agency misinterpreted the msa to force struggling fishermen to pay up to 20 percent of their annual profits to federal agents. the government says that even if all nine of you agree with us that the agency's construction is worse than ours, you should nonetheless defer to that construction and uphold their program under chevron. that's not consistent with the rule of law. if we have the best view of the statute, we should win this case. i welcome the court's questions. justice thomas: mr. martinez, how much deference is in tension with the judicial role? martinez: i think it's very much in tension, your honor. justice thomas: no. how much would it require? i mean, your argument is that chevron deference is problematic.

but how do we determine how much deference is too much deference? martinez: i think you've certainly crossed the line when you have a rule that says that we're going to allocate interpretive authority from -- from article iii courts to an agency. and so, when -- when you've got deference that amounts to that, which is what chevron deference is, then i think you've -- you've crossed the line because what you've really done is -- justice thomas: i think what i'm trying -- what i'm asking is, how do we know where the line is? we show deference. you -- there's skidmore deference. martinez: sure. justice thomas: we are deferential in fact finding, et cetera. so i'm just trying to determine whether you're saying that we -- if it's not de novo review -- martinez: right. justice thomas: -- without any presumptions or deference, then it's problematic. martinez: i -- i think deference becomes problematic when it requires a judge to say that the law means x when really the judge thinks the law means y. i think skidmore deference is not problematic because it doesn't require that. skidmore deference essentially says -- and we would be very comfortable with skidmore -- that because the agency has a -- has an important role to play in the process, often the agency has helped draft the statute,

the agency has knowledge of the policy context surrounding the statute and its implementation. of course, courts should pay special ultimately has to bring its expertise to bear in a way that's persuasive. and if the -- the court isn't persuasive, if the court thinks that the law means x even though the agency thinks the law means y, then the court needs to go with the best interpretation of the statute, just like it does in every other -- chief justice roberts: well -- martinez: -- area of statutory or constitutional -- chief justice roberts: -- let's -- martinez: -- interpretation. chief justice roberts: -- let's suppose the statute says the department of transportation will set length limits for trucks that are reasonable. martinez: right. chief justice roberts: is that a legal question for the court, or is that a policy question for the agency? martinez: i think that -- chief justice roberts: it's a -- the -- the legal authority says they've got to be reasonable. that's a term that courts apply in many situations. martinez: i -- i think that a court looking at that statute would try to determine the best meaning of the statute, and the

best meaning of the statute there would be that -- that the use of the term real -- "reasonable" confers upon the agency discretion to choose among certain policy options. now that doesn't mean that the agency can just do whatever it wants because there are limits, and the court has to police that limits. michigan versus epa is a good example. congress used a broad term like "appropriate" and the question was -- which is similar to "reasonable," in giving the agency a -- a range of discretion. but, at the same time, when the agency said, well, we don't have to consider costs in figuring out whether something is appropriate, the court said no, that, as a legal matter, the best interpretation of the word "appropriate" in the context of this statute requires the agency to consider costs. chief justice roberts: well, what if the statute says that the agency can regulate truck -- truck length for vehicles that travel in interstate commerce and there's a question whether or not interstate commerce -- the -- the delegation for interstate commerce is satisfied when particular -- martinez: right. chief justice roberts: -- circumstances are present. martinez: i -- i think that that would be a case if you're -- if the court were called upon to interpret what -- if the dispute was about whether -- what

interstate commerce means, i think that would be a classic legal question that would be a legal question for the court. and i think it actually highlights -- because interstate commerce is probably there because of the constitutional limitations, it highlights the fact that, really, the same rules should apply to interpreting constitutional -- chief justice roberts: well, i mean -- martinez: -- provisions as statutes. chief justice roberts: -- you could imagine -- you could imagine situations where the interstate commerce determination is peculiarly fact-bound, you know, trucks transferring loads and -- at transfer points on the border. is that in interstate commerce for each one or not? and isn't the policy judgment of the agency pertinent in that situation? martinez: i think, certainly, the policy judgment of the agency is -- is pertinent in determining sort of the facts because the agency might be on the ground and understand the factual scenario. but i think there's a -- an important legal component to that question, that in any other context, like, for example, if you were interpreting the

constitution, i think the court would -- would quite reasonably think it's its own job to interpret the constitutional requirement of intstate commerce and would -- would say -- would give it its best meaning. and i think -- justice kagan: well, let me give you -- martinez: -- the same approach -- justice kagan: -- a few more examples along the same lines, mr. martinez. is a new product designed to promote healthy cholesterol levels a dietary supplement or a drug? martinez: sorry. can you give that one more time? justice kagan: a new product designed to promote healthy cholesterol levels, is it a dietary supplement -- that's a statutory term -- martinez: ok. justice kagan: -- or a drug? martinez: i -- i think it would depend on -- on the -- the original understanding of the text of that statute in -- read in context. justice kagan: you -- you want the -- martinez: and i think that's a -- a legal question for a court. justice kagan: -- you think that the court should determine whether this new product is a dietary supplement or a drug without giving deference to the agency, where it is not clear from the text of the statute or from using any traditional methods of statutory interpretation whether, in fact, the new product is a dietary supplement or a drug? martinez: i -- justice kagan: you want the courts to decide that? martinez: justice kagan, i think

with respect to that question or any other of the -- a legal question, i think what the court would do, there -- there are going to be hard questions, but i think the court would bring all the traditional tools of construction to bear -- justice kagan: they do that -- martinez: -- and would -- justice kagan: -- under chevron. they -- you know, we have made clear all the traditional tools, if you can find an answer, that is the answer. so the court is very rarely in the situation in which you're talking where it thinks the law means x and instead it says y. if it thinks it means x, under chevron, as we've understood it and made clear and reigned it in a little bit over these last few years, it's supposed to say x. but sometimes law runs out. sometimes there's a gap. sometimes there's a genuine ambiguity. and i -- i don't know. in that case, i would rather have people at hhs telling me whether this new product was a dietary supplement or a drug.

martinez: so, your honor, i think a couple things. first of all, i don't think chevron is a doctrine that only applies to tie-breaker 50/50 scenarios. it's never been understood that way. you know, justice scalia in his famous article in 1989 -- justice kagan: it's not a tie-breaker. there are just some times where you look at a statute and the most honest reading is that there's -- there's -- there's a gap there -- martinez: but -- justice kagan: -- because of the limits of language, because of the limits of our ability to predict the future. and so who fills that gap? martinez: but i -- i guess what i would sort of push back on is i don't think there's a gap if the court looks at the statute and thinks, hey, this is a really hard case, it's a really close statute. fifty-two percent likely, i think -- you know, i have 52 percent confidence that x is right -- justice kagan: i'll give you -- martinez: -- 48 percent likely -- justice kagan: -- i'll give you

another one, mr. martinez. does the term "power production" -- i'm just -- these are real cases. martinez: right. justice kagan: these are -- these are prototypical chevron cases. martinez: but -- justice kagan: does the term "power production capacity" refer to ac power that is sent out to the electric grid or dc power that's produced by a solar panel? martinez: i think same answer as the first hypothetical. but let me try to -- let me try to sort of give you a different framework for thinking about this problem. let's imagine that that statute came to a court before an agency had even acted in the first place. what would a court do? would a court look at the statute -- a statutory term like that that's a hard -- presents a hard interpretive question and say: well, this is hard, it's sort of 52/48, it's kind of close. i think the law has run out and i'm just not going to be able to decide this. i think the court would go with the best interpretation. justice kagan: the -- the -- the -- the court might -- the court in that case would have to make a choice. but, you see, here, the court can say, you know, the best option is to listen carefully and to defer if it's reasonable and if it's consistent with everything that we know that congress has said, to defer to people who actually know things about these things -- martinez: but -- justice kagan: -- to -- you know, to people who understand the way particular questions fit within a broader statutory and regulatory scheme, to people who

have understanding of the policies and of the facts that led to this. i'll give you a third example. martinez: can i respond? justice kagan: and this will be my last one, mr. martinez, and it's going to be fairest one because it's going to be one you know about, which is chevron. as a stationary source in the clean air act, does it refer to whole plants or to each pollution-emitting device within the plant? martinez: we think that the decision in chevron was -- reflected the best interpretation with much respect to justice gorsuch's mother's epa. we think that that was the best interpretation. but -- but can i just go back and i think what you described earlier about listening to the agency and taking into account all those things, our -- our rule would allow that. that's skidmore. i think the only difference between our rule and -- and the skid- -- what -- the skidmore sort of approach and the chevron approach is that after listening to the agency's explanation of all the things that you said, if the court isn't persuaded by the agency that the agency's interpretation is correct, chevron would say you still have to go with the agency. and that's just like a dramatic thing. justice sotomayor: buthy not? meaning i -- i think all of the

play in disagreement is around the word "ambiguity." i know that there have been some earlier cases that suggested if there were two plausible meanings, you went with the agency meaning. i think we've gone far beyond that. it has to be two reasonable meanings. assuming -- you -- you make an assumption that there is a best answer. i don't know how you can say there's a best answer when justices of this court routinely disagree and we routinely disagree at 5/4. is the best answer simply a majority answer? i don't think so. martinez: but, your honor, if -- justice sotomayor: i happen, when i dissent, think the others got it wrong. [laughter] justice sotomayor: and they often do. [laughter] justice sotomayor: but putting that aside -- but putting that aside, in those situations, there are two plausible -- not nearly plausible. there are two best answers. and the question is who makes

the choice or helps you make the choice. and if the court can -- can disagree reasonably and comes to that tie-breaker point, and it could be 51/49, it could be 52/53, if it's that close, why shouldn't the person with all of the qualities you spoke about, the entity with all of the qualities, expertise, experience, on-the-ground execution, knowledge of consequences, why shouldn't deference be given to that entity? martinez: justice sotomayor, i think your explanation of ambiguity just now just proves the problem with chevron because i think what you said is that whenever there's a case, a statutory se in which the members of the court disagree with one another, that -- that's essentially saying the statute is ambiguous because reasonable people can disagree. justice kagan: that's what nobody believes -- martinez: well -- justice kagan: -- about chevron, mr. martinez. as we've described it, if you -- you work hard to figure out a

statutory problem. you don't say, oh, it's difficult. oh, there are two interpretations. oh, you know, not everybody agrees with this in three seconds flat. you don't say that. you do everything you do, look at the text, look at legislative history if you believe in legislative history. look at context. look at every tool you can, and still there are places where we don't know whether this drug is a -- is a -- is a -- whether this product is a drug or a dietary supplement, and it's best to defer to people who do know, who have had long experience on the ground, who have seen a thousand of these kinds of situations. and, you know, judges should know what they don't know. martinez: i -- i agree with that, justice kagan.

but, with -- with all due respect, i -- i think i understood justice sotomayor to be saying that whenever judge -- justices of this court disagree about the be meaning of the statute, because, obviously, everyone on the court is reasonable, that shows that there's an ambiguity. if that's the test, which i think was the implication of the question, then that can't be wrong. that's much broader than -- justice sotomayor: that wasn't -- martinez: -- step one. justice sotomayor: -- my implication. my implication was that using all the statutory tools, you can still come up, using them in good faith, using them, you can still come up with no answer -- martinez: well, i think -- justice sotomayor: -- with no clear answer. martinez: -- i -- i think you can can come up with no clear answer because some -- justice sotomayor: or no best answer. martinez: -- because some statutes are hard. but i think you can come up with a best answer, and -- and the reason i think that is because -- justice sotomayor: best only because a majority agrees? justice jackson: but -- martinez: no, no, because -- because, if you had the same statute with the same interpretive question posed to you without the agency having acted, i don't think you would say there's no answer here. i think you would choose the best answer. justice gorsuch: i mean, mr. martinez -- justice jackson: but, mr. martinez -- justice gorsuch: -- i guess i'm struggng to understand what -- what -- what's at stake here given the questions because, as i understand justice kagan's hypotheticals, which are -- are hard ones, that one option would be to say it's ambiguous and, therefore, the agency always wins.

that -- that's what i understood chevron to mean at least coming in here today. another would be to listen carefully to both sides and provide special weight under skidmore to a coequal branch of government's views about the law, which one would think we would do anyway, and that they would have -- have -- be considered great weight in arriving at the best answer and that that's what a court would do if -- if there were no interpretive principles advanced by the executive branch, if there hadn't been some sort of rule or adjudication. is that -- is that correct? martinez: i -- i think that's correct. and i think the difference between the skidmore approach that you just laid out and the chevron approach is just, at the end of the day, once you've considered all the expertise and all the information the agency has to bear -- justice gorsuch: who decides? martinez: who decides? who -- is the judge persuaded or not persuaded? justice gorsuch: is the judge persuaded at the end of the day, with proper deference given to a coequal branch of government, or does the judge abdicate that

responsibility and say automatically whatever the agency says wins? martinez: right, even -- even if the judge is not persuaded. justice jackson: but, mr. martinez -- justice gorsuch:nd then -- and then -- justice jackson: -- doesn't that -- justice gorsuch: -- and then if i might just -- just finish up, what -- what's the effective difference of that? it seems to me that in the first case, when -- when a judge says here's the law, it's settled, we're done, right? it can be appealed, but at the end of the day, if the supreme court of the united states upholds that interpretation, we're finished. whereas, under the chevron approach, are we finished? martinez: no. justice gorsuch: what happens? martinez: i think the agency can overrule what the court said. the agency can overrule what itself said. i think that's a very strange thing, that in every other area of statutory interpretation, we understand the law to have one fixed meaning and the goal is to try to figure out that fixed meaning, but chevron by design creates this world in which the agency is -- is -- because there's this zone of discretion, the -- the agency and ambiguity, the agency can kind of flip-flop and then force courts to flip-flop with them.

justice gorsuch: and i'm struck on that score by the brand x case, which involved broadband, in which this court said, okay, agency, you automatically win with respect to one interpretation of the bush administration, i believe it was, and then, of course, the next administration came back and proposed an opposite rule. martinez: right. justice gorsuch: and then the next administration came back and flipped it back closer to the first. and as i understand it, the present administration is thinking about going back to where -- martinez: that's -- that -- that's exactly right. justice gorsuch: -- where we started. martinez: that's exactly right, justice gorsuch, and i think it -- it plays up the real problem. chevron really is a reliance-destroying doctrine. imagine if you're a person or a regulated entity and you're trying to figure out what the law is. you should be able to rely on the best interpretation of the law and not have to, you know, check the -- the c.f.r. every couple years to see if the law has somehow changed, even though congress hasn't acted. justice gorsuch: and that's the delta between skidmore and chevron? martinez: i think -- i think that's right. i mean, skidmore, i think, would allow for -- for courts to give meaningful weight and consideration to -- to

persuasive opinions by agencies. the only thing skidmore doesn't do is require a court to give up its -- its interpretive -- ultimate interpretive say and defer to an interpretation that is not persuasive. justice gorsuch: thank you. justice jackson: mr. martinez, what -- what i'm stuck on is what seems to be an assumption in your argument that every question posed with respect to interpreting -- interpreting a statute is a legal one. i see chevron as doing the very important work of helping courts stay away from policymaking, and so i -- i'd like for you to sort of think of it through that lens and help me understand why, if we do away with chevron's framework, we won't have a problem of courts actually making a policy decision.

so justice kagan gave you a number of examples, and i think the reason why those examples are hard or why they're ambiguous or whatever is because, at bottom, they're not asking legal questions; they're asking policy questions. how is it that, you know, stationary source is to be defined? that's not really a legal question. i mean, there could be several reasonable ways of interpreting that. and at the end of the day, i think the way i've been thinking about chevron is congress has given that policy choice to the agency. and my concern is that if we take away something like chevron, the court will then suddenly become a policymaker by majority rule or not, making policy determinations. so how can we avoid that? martinez: so we agree, obviously, that -- that courts should not be in the business of policymaking. and i think the whole enterprise of statutory interpretation, when properly understood, is -- is designed to take courts out of policymaking because what the court is trying to do is -- is act as a faithful agent of what congress has done and find the best -- justice jackson: but isn't that --

martinez: -- interpretation. justice jackson: -- isn't that what chevron does? i mean, isn't chevron, step one, even in this very case, asking the question, one, has congress made that policy determination? so, for example, here, the question is whether or not monitors on the boats have to be paid for by the owner of the boat. i see that as a policy question. congress could have said yes or no. there's nothing about law really inherently in the question of should the monitors on the boats be paid for by the owners or the government. so step one is has congress in the statute answered that question. when we say no, everybody agrees that's not in the statute, then we say the agency can make that determination so long as they do so in a reasonable way. so, for example, here, the question is whether or not monitors on the boats have to be paid for by the owner of the boat. i see that as a policy question. congress could have said yes or no. there's nothing about law really

inherently in the question of should the monitors on the boats be paid for by the owners or the government. so step one is has congress in the statute answered that question. when we say no, everybody agrees that's not in the statute, then we say the agency can make that determination so long as they do so in a reasonable way. and the -- and courts sort of police the boundaries of reasonableness, but whether or not the monitors are paid for is not really a legal question. martinez: i think the question of whether or not the law allows the agency to -- to force the monitors to be paid for by private industry is absolutely a legal question. i agree with you that when congress -- justice jackson: but isn't that the same question as to whether or not -- isn't that just another way of saying, can this policy determination be made by the agency? martinez: no, i don't think so. i think the difference is when the -- when the -- when the policymaker, whether it's congress or the agency, is sitting there and trying to figure out, like, what the best policy is, would the world be a

better place if industry has to pay for these monitors or not, that's absolutely a policy question. justice jackson: ok. so that's the question -- martinez: but -- but -- justice jackson: -- right? martinez: no, because, when it comes to a court, the court is not figuring out what the best thing for the world is. the court is figuring out, well, what did congress actually want here. it's -- justice jackson: but i guess i'm afraid that the court really is figuring out what the best thing in the world is if we -- martinez: but -- but -- justice jackson: -- look at it through your lens, right, because, if the answer to the question is, you know, should -- should they pay for it or not, the agency has a view, and unless we're deferring to that view, i don't see why we aren't overriding the -- the agency's policy prerogative. martinez: the -- the question that the court should be answering is not should agency -- should industry pay for the monitors. the question that the court should be answering is, did congress require or allow agent -- industry to be forced to pay for the monitors? and that's a very different question. that's the different between law and policymaking. and i think the whole assumption and the whole understanding of statutory interpretation under this court's cases is there's a difference between law and policymaking. judges are there not to exercise force or will. they're there to exercise judgment.

they're -- they're serving as neutral umpires. they're not players on the field. justice jackson: all right. so how does that -- justice barrett: mr. martinez -- justice jackson: -- play out under your interpretation -- so, here, what -- what is the question we're supposed to be answering? martinez: the question you're supposed to be answering is, did -- does this statute require -- has congress required -- either required the -- the monitors to be paid for by industry, or has it given the agency the authority to make that decision? and i don't think -- i think that is a legal -- both of those versions of that question are legal questions, and the answer is no. justice barrett: mr. martinez, can i ask you a question about the line between law and policy? and i want to ask you in the context of one of justice kagan's examples, the dietary supplement or drug. where is the line between something that would be then subject to arbitrary and capricious review and something that's a question of law? because i'm just wondering whether we could say that the definition of dietary supplement or drug might be something that's a question of statutory interpretation in the context of the statute, but which category any one thing fell in might be a question of policy for the agency. martinez: right. i -- justice barrett: is that

possible? martinez: -- i -- i think that's right. i think that would be more of a -- of a, you know, application of law to fact or a factual question. but i think the core question of, like, you know, what is the meaning of dietary supplement, and i forget what the other alternative was, those are legal questions. justice barrett: but whether the particular cholesterol-reducing drug fell -- martinez: right. justice barrett: -- in one category or the other, i mean, you know -- martinez: that -- that would be a -- justice barrett: -- presumably, that depends on how does this function? what is the mechanism by which it decreases cholesterol? martinez: i -- i think that's right. but i think it's -- i do think it is important to make -- retain the sort of legal component of that question and -- and make sure that the courts have authority over that legal component. justice barrett: i want to ask you something about your article iii argument too. you know, justice thomas asked you what the line is. and, you know, courts all of the time make judgments about whether things are reasonable. but i -- i don't understand you to be disagreeing that things like whether something -- that an agency could be tasked with deciding what was the most feasible, most useful, most reasonable. well, courts could do that too. so is that a delegation of judicial power that would offend article iii in your view, to give those kinds of --

martinez: no, i think -- justice barrett: -- decisions to an agency? martinez: -- i think the way to think about those kinds of -- of statutory provisions would be that the best interpretation of the statute, given the nature of the word "reasonable" in context, is to confer a range of discretion on the agency. and so i think a court in that case -- if -- if the agency is operating within the range of discretion, that's arbitrary and capricious review. if the agency is sort of operating at the edges, you have to figure out where the guardrails are. that's the legal question. so, if the -- if the statute says, you know, the agency can pick red, blue, or green, then the choice among those three options is for the agency. but, if you have a legal question like, oh, does pink count as red, that's a legal question. chief justice roberts: thank you, mr. martinez. how much of an actual question on the ground is this? i saw some study that said we haven't relied on chevron for 14 years. and judge kethledge has written -- he's been a judge for 10 years. he's never invoked chevron step two.

you know, judges are used to deciding things, and when they get around to doing it, they tend to think what they've come up with is not only the best answer, but it's the only answer. [laughter] and -- and i just wonder how often this comes up? martinez: i think it comes up a lot, your honor. and this court hasn't relied on chevron since 2016, but the lower courts still have to apply it. and i think these two cases, the -- the two that you're going to hear this morning, sort of show what happens when -- when courts are applying this doctrine because they're -- they're essentially getting to a point where they don't really have to figure out the best answer and they can just -- you know, instead of asking what does the statute mean, they can ask a different threshold question, which is, is this statute ambiguous enough that -- that we should just, you know, let the agency do the work for us? chief justice roberts: thank you. justice thomas? justice alito? justice alito: mr. martinez, would you agree that one of the reasons why chevron was originally so popular was concern that judges were allowing their policy views, consciously or unconsciously, to -- to -- to influence their

interpretation of the statutes in question? martinez: yes. justice alito: why was that fear unfounded? why do you think now that the fear was unfounded? martinez: well, i think three things. first of all, i think the fear has -- it's reasonable to think the fear has diminished over time, regardless of what it was then, in large part due to the very salutary developments in the way that this court and the lower courts generally now think about statutory construction. in the old days, there was a lot of reliance on legislative history and on sort of more free-form analysis that i think made it easier for policy considerations to infect the judicial decision-making process. but this court has now made clear that, you know, really, we should be text-focused, we should be focused on faithful agency to congress. so i think that is one difference. i think another difference is courts now have become more

appreciative of the fact that we're not just talking about, you know, judicial -- rules of, like, judicially made common law about how to interpret statutes. we have the apa here. justice scalia was a big defender of chevron in its original incarnation but, over time, came to realize that the apa had text that actually bore on this question. and i think, when you're enforcing that text, you come to the same place as our article iii argument, which is that courts have to exercise independent judgment. justice alito: do you think that the canons of interpretation that we have now and all of the other tools that we have in our statutory interpretation toolkit are like the enigma machine and so we have these statutes and they're sort of written in code and we run them through the enigma machine and, abra cadabra, we have the best interpretation? do you really think that's how it works? martinez: i -- i think that what this court does with respect to the normal canons of construction is it's used the -- it's -- it's generated those canons as rough rules of thumb to help guide the interpretive process because, if the court

believes that the canons best approxime the best original meaning of the statute, especially -- and then there's some canons that -- that sort of are not purely textual canons but that sort of are informed by constitutional -- foundational constitutional values. i think chevron's very different from that because, with chevron, you're doing something -- you're not trying to find the best interpretation anymore. you're, in fact, agreeing that you have to impose the not-best interpretation because you have to defer. and so, unlike all the other canons, chevron is the only one that says to courts, you can stop doing your normal interpretive function and we're going to allocate that interpretive function outside of article iii. justice alito: thank you. chief justice roberts: justice sotomayor? justice sotomayor: i counted over, i think -- not i -- the solicitor general or someone has given us a list of 77 cases in which the court has used the chevron approach and interpreted what the law was. your overruling chevron puts a question to all those 77 cases.

martinez: no, your honor, i think -- justice sotomayor: no, your out is it's stare decisis now? martinez: right. so -- justice sotomayor: until the agency does something else? and then people can come back because it's not stare decisis anymore? martinez: so i think, with respect to the effects of -- of applying normal rules of construction here instead of chevron, i'd say two things. first of all, the 70 holdings or whatever, the bottom-line holdings in those cases would get stare decisis, so they would not be undermined. so there's no convulsive change of the law with respect to that. justice sotomayor: i don't understand how that happens. once you have a new approach, i'm not sure. martinez: i -- justice sotomayor: but let me move on to the second part of my question, which is the cases that come to the court are usually the hard cases. so you say in the last 14 years we've barely referenced chevron. and do you know what the breakup is? how often have we consistently upheld the agency in those cases? martinez: in -- in the cases since 2016?

justice sotomayor: yes. martinez: i -- i don't know the track record on it, your honor. justice sotomayor: i know, it's interesting. martinez: but i will say, i mean, there -- there's some prominent -- justice sotomayor: but -- but putting that aside where we disagree, do you suggest that our disagreement was based on ignoring of chevron or us doing exactly what you say we should be doing, which is to say this is outside the bounds of reasonableness or around the guardrails because you're going outside of plausible -- martinez: i -- justice sotomayor: -- of reasonable interpretation? martinez: -- i think the court in cases like the american hospital case or the digital realty case, which i think are two really good recent examples, the court unanimously overturns the lower court decision because it does exactly the right thing. it does all the canons at step one and it -- and it essentially says, like, the statute is clear. but i think what those 9-0 decisions show is how confusing and unworkable chevron is because the lower courts, you know, purported to do or didn't really do what they were supposed to do and they came to the opposite conclusion, not

necessarily because they thought that -- that your interpretation wasn't the best but rather because it thought that the statute was ambiguous enough that it required deference. and so it's like a threshold -- justice sotomayor: counsel, that judgment is inherent in every question. i mean, that -- that kind of problem is just a part not just of judging but of decision-making, period, of life. and so it's not clear to me that the fact that there may be some ambiguity about what -- how much ambiguity, the question that justice thomas asked, it doesn't take away from the basic premise of chevron, which is a reasonable interpretation within the bounds of -- of common statutory interpretation should be given deference. martinez: right. but i do think the ambiguity trigger introduces a whole kind of threshold question that's very hard to apply neutrally. i mean, you have great judges, judge kethledge, i think, was referenced. he doesn't -- he never found a case that required him to go past step one. just silberman, another great

judge, said that in most cases he thought the statute was ambiguous. and if there's that much disagreement, then i think that's a sign that chevron really isn't workable. and this court has tried to rein in chevron in numerous ways, but i think that what all of those efforts show is that you kind of need a secret decoder ring to figure out what the law means under this court's approach. you have to do step zero. you have to apply need. then you have to do a robust step one inquiry, taking into account footnote 9 and taking into account, you know, how much ambiguity is needed. in this -- in the d.c. circuit, you have to do step one and a half, where you have to figure out whether the agency recognized that the statute was ambiguous. der kisor, there's maybe a step three that says you turn off deference when the agency's operating outside of its area of expertise. and then overlying all that you've got the major questions doctrine. and so i think, if -- if -- if that's kind of what -- justice sotomayor: well, that's the court's creation. martinez: right. but it's the court's creation because it's trying to solve the fundamental problem, which is that chevron is doing something very weird. it's taking interpretive

authority that belongs to courts and it's giving it to agencies. so all of these bells and whistles are efforts to kind of claw it back to address the symptoms, but i think it's time for the court to address the disease, the underlying problem, which is chevron itself. chief justice roberts: justice kagan? justice kagan: mr. martinez, i want you to think of this from congress's perspective. so i was thinking what is the next big piece of legislation on the horizon and who knows, don't have a crystal ball, but i'm going to say -- i'm going to guess that it's artificial intelligence. so let's imagine congress enacts an artificial intelligence bill and it has all kinds of delegations, maybe it creates an agency for the purpose or maybe it uses existing agencies and it has all kinds of delegations to that agency or agencies about how to regulate artificial intelligence so that this nation can capture the -- the -- the opportunities but also meet the challenges of that. and then, just by the nature of things and especially the nature

of the subject, there are going to be all kinds of places where, although there's not an explicit delegation, congress has in effect left a gap. it has created an ambiguity. and what congress is thinking is, do we want courts to fill that gap, or do we want an agency to fill that gap? when the normal techniques of legal interpretation have run out, on the matter of artificial intelligence, what does congress want, mr. martinez? martinez: i think congress wants courts to interpret the best interpretation of their -- justice kagan: congress doesn't know -- kagan: -- apply the best interpretation -- justice kagan: -- what that answer means. congress knows that there are going to be gaps because congress can hardly see a week in the future with respect to this subject, let alone a year or a decade in the future. and congress knows that there are going to be things that it writes that it's just not going to be clear how this will apply or what it will mean with respect to countless factual

situations that this country will have to address. does the congress want this court to decide those questions, policy-laden questions, of artificial intelligence? martinez: i -- i don't think congress wants the court to do policy. i think congress wants the court to do its ordinary function, which is interpret the law and figure -- and apply the best understanding of the law. and i think that the implication of your question is that this is some sort of intentional delegation by congress that chevron deference is -- is this implicit delegation. but i -- i don't think that's right. i think many people, including a very insightful article that -- that you wrote 20 years ago, make clear that this is fictional. this is delegation of a fiction. justice kagan: fictional just means -- is like academic speak for presumed. we are indeed presuming congressional intent. the congressional intent, you know, the -- the delegation that's not explicit on the face of this statute, but what we're thinking is congress knows things about different institutions, about what they know, about what they're competent with respect to, and congress knows that this court and lower courts are not competent with respect to

deciding all the questions about ai that are going to come up in the future. and what congress wants, we presume, is for people who actually know about ai to decide those questions. and also, those same people who know about ai are people who, to some degree in some way, are accountable to the political process. they have constituencies. they have fact-finding abilities. they are obligated to go consult with people. they report to a president, who needs to be elected. in all kinds of ways, both with -- with respect to expertise and with respect to their connections to the public and to other policymaking entities, those are the people congress wants to decide questions about ai. we don't even know what the questions are about ai, let alone the answers to them, we being the court. martinez: justice kagan, i think, if we're trying to figure

out what the -- what the reasonable thing to infer that congress has presumed, i think the far more reasonable presumption and the one that's most consistent with our constitutional structure is that congress is going to presume that courts are going to do law and not policy, they're going to pick the best interpretation and enforce the best interpretation as to this statute in the exact same way that they would do it with respect to any other -- any other statute. and i this think case actually -- you know, ai is a trickier example -- 19 example. i mean, this case, you know, whether it's -- it -- it was a correct interpretation or not a correct interpretation of chevron is really not the issue that we're deciding here. the issue we're deciding here is more like that, is more like the countless policy issues that are going to confront this country in the years and decades ahead. will courts be able to decide these issues as to things they know nothing about, courts that are completely disconnected from the policy process, from the political process, and, you know, that just don't have any expertise and -- and experience

in an area, or are people in agencies going to do that? martinez: i -- justice kagan: that's what this case is about. martinez: -- i think the constitutional answer is that congress needs to set the rules with respect to ai. it can delegate some policymaking discretion to agencies, but once the law is written and the interpretive function has begun, then that job is -- is for the courts. and i think this case actually really is a good example because i think the problem with chevron is that, like, no one really -- i mean, i'm curious to see what the solicitor general will say about this, but does anyone really think that congress was presuming that the agency would get to decide the question of who pays for the monitors? justice kagan: ok. i have one last question. do you think that congress could codify -- codify chevron? martinez: i -- i don't think so because i think that -- that a statute that codifies chevron would say, essentially, that the interpretive authority has been reallocated from the court to the agency. i think that -- justice kagan: congress -- martinez: -- interpretive authority -- justice kagan: -- cannot decide that in cases -- after all the statutory tools have been used and there remains a gap or an

ambiguity, congress could not decide that it wants people who know something about something to decide the questions that will be left over? martinez: i -- i think that gives away and -- and would -- would take away from courts and give to agencies core judicial interpretive authority. i don't think congress could do that. in the same way that congress couldn't tell the president how to exercise the veto power or the pardon power, it can't tell courts how to do interpretation and to defer to someone else. justice kagan: thank you. chief justice roberts: justice gorsuch? justice gorsuch: do we have to decide that constitutional question? martinez: i think it makes sense to decide the constitutional question. i think you could -- justice gorsuch: that wasn't -- do we have to? martinez: i think you could resolve this case under the apa, and we would certainly welcome an -- an interpretation of the apa that comes out our way, especially if it's informed by constitutional avoidance principles that i think have a lot of salience here. justice gorsuch: are -- does anything in your argument suggest or depend upon the idea that judges should make or decide policy questions about ai or anything else?

martinez: no. we -- we a hundred percent agree that judges should not do policy. we just think that they should do law. and that's in -- chevron is about legal questions. justice gorsuch: then there was some question about past decisions, and as you pointed out, this court's moved away from using legislative history to some degree in favor of text, and we've made other changes in our interpretive approaches too without congress's intervention, for example, in sovereign immunity contexts, returning to the clear statement rule that had preexisted this court's jurisprudence for 200 years, and then we wandered off into legislative history and circled back around and corrected our own mistake. we had to deal with the question of what to do with those precedents, and our answer was to leave them alone from -- from those ancient regimes, as we -- martinez: right. justice gorsuch: -- called them. are you asking us to -- to do anything different when it comes to chevron? martinez: no, and if i could just explain what -- how i think the world would look with respect to the old cases. i think stare decisis would apply to the holdings of those

old cases. i don't think that -- that anything would change. you know, stationary source would still mean what it meant when -- when the court issued that bottom-line interpretation. and so i don't think that this would -- a ruling in favor of our side would -- would require or entail overturning any of those old cases. i think what we really care about is prospectively, both with respect to the fishing regulation here but also with respect to other cases that come forward to the courts, making sure that courts are the ones doing the interpreting and not agencies. justice gorsuch: thank you. chief justice roberts: justice kavanaugh? justice kavanaugh: several questions. first of all, on skidmore, there was reference to skidmore deference, and i guess i don't think that's the right term, that it's respect or pay attention to, but i think, if we throw the term "deference" into skidmore deference, we're going to walk into another problem -- martinez: some -- justice kavanaugh: -- like the one we have with chevron deference. martinez: some might say "deference" is ambiguous. i think that --

[laughter] -- that it's imprecise. i think the better way -- i think oftentimes, when people say "deference," what they mean is that if you think the answer is x, you should defer to someone else's answer, which is different. i don't think -- i think absolutely that that would be inappropriate. so i would not use "skidmore deference" because i think it -- it runs the risk of -- of giving that implication. i think that, really, we're talking about very serious consideration of the points that the agency makes, but, ultimately, you have to be persuaded. and if you're persuaded, then that means that you've concluded that the agency has the best interpretation and then you just apply the normal rules. justice kavanaugh: right. i thought skidmore was about the power to persuade, not the power to control. martinez: exactly. justice kavanaugh: yeah. martinez: we -- i agree with that. justice kavanaugh: ok. on the constitutional issue that justice gorsuch and justice kagan were raising, you have lots of arguments here, and mr. clement does too, for overruling chevron without reaching the constitutional issue. so i guess why -- why would we reach it? if -- if we agreed with you on overruling chevron on other grounds, i don't see the need to

address the hypothetical that justice kagan raised about congress passing a chevron-type regime. martinez: i think three things on that. like i said earlier, we would certainly welcome overruling chevron, especially under the apa and especially if informed by constitutional avoidance principles. but i think there are three reasons why you should consider going beyond that to the constitutional holding. there are going to be some cases that, as a technical matter, section 706 of the apa wouldn't -- doesn't apply. and so, if it's an apa holding, it may be that in those cases there might be lingering uncertainty about whether deference should -- should apply to cases that aren't technically under section 706. i think the second thing is that a lot of the analysis in figuring out what the duty under the apa to interpret the law, i think a lot of that analysis really overlaps with the constitutional points. and i think, if you -- if you get to a place where you agree with us on the apa, it's not that far, not that different to ultimately agree with us on the constitution as well.

and then, finally, i would just say that although, of course, this court often prefers to rule on non-constitutional grounds, i think it's also recognized in cases like pearson versus callahan that there's going to be a value and a benefit to the judicial system to providing clarity about what the constitution means. i think -- i would respectfully submit this is one of those situations. justice kavanaugh: on the question of how much does chevron matter on the ground, i think you addressed this a little bit by citing judge silberman, but do you want to elaborate on that? i mean, are -- there are cases, i assume, that get to chevron step two pretty regularly. martinez: very regularly, your honor. it happens all the time. and i think, if a case like this one or two cases like these two can get to chevron step two, i think that suggests that it's really hard to figure out how chevron step one is supposed to work. i mean, the digital realty case is another great example. that's a case where there was a statutory definition of the term "whistleblower" that required the person to have gone to the sec and -- and, you know, submitted a -- a complaint, and

the government and the lower court concluded that that was ambiguous and that it might actually apply, it was reasonable to read the statute to not require a report to the sec. so i think there are cases, there are examples like these that come up all the time, and, you know, thankfully, this court doesn't have to intervene every single time, but the reason that the problem is there is because you've told lower courts how to do their interpretation. and as long as that instruction is out there, there are going to be a lot of cases that get it wrong, and you're not going to want to be in the business of sort of error correction on each one. justice kavanaugh: on the question of how congress can operate without chevron, i just want to elaborate on -- have you elaborate on that a little more. my understanding is congress oftentimes will use terms like "the agency can regulate reasonable limits" or "appropriate limits," and that gives, under state farm, a lot of discretion to the agency to make choices to do what justice kagan was talking about, to think about the world as it exists five years from now or 10 years from now and not have to worry about going back to congress.

so the question really is for congress and its drafting choices, i think, what kinds of broad, capacious terms it uses, as opposed to using more defined terms or statutory terms -- usual kinds of statutory language. yes, it can't rewrite that. at least that's how i thought congress could operate in a world where chevron does not exist. martinez: i -- i think that's exactly right, justice kavanaugh. and i think that, like i said earlier, in -- in those situations, the court's job is basically figuring out what the best interpretation of that word is. and in many cases, maybe most cases, those types of capacious words are basically -- the best understanding of those words is that congress is, in fact, conferring the discretion on the agency. that's very different from chevron, where, instead of having any sort of language like that or express language conferring a delegation, you're -- you're basically applying

this fictional implied delegation that -- that is triggered by ambiguity, which is like -- you know, frankly, it's -- it's -- it's not -- it's fictional, it's made up. and so i think a world in which congress, when it wants to delegate to agencies, needs to be express and use language like that or other language, i think is a better world from the perspective of -- of article i and from article iii. justice kavanaugh: thank you. chief justice roberts: justice barrett? justice barrett: mr. martinez, i want to return to the question that justice sotomayor raised about stare decisis. so you said that overruling chevron wouldn't have an effect on the many cases that have gotten to chevron step two and then deferred to the agency. you said -- am i -- did i understand you correctly? martinez: those bottom-line so isn't it inviting a flood of litigation even if for the moment those holdings stay intact? martinez: so i would say the bottom-line holdings in those cases, i would just quibble slightly, i would -- i would describe the bottom-line holding as being that the agency's action was lawful. and so that's the bottom line. i think it's true that people could come and say, look, the interpretive methods have changed since this bottom-line holding was issued and we think that -- that, you know, a different result now should apply.

and -- and that's why courts consider requests to overturn precedent. but i just think that they would apply the same standards that they would apply to other stare decisis inquiries, and i think it would be the rare case that would require -- that -- where a court would say this -- this decision not only isn't the best interpretation, but it's like so bad and so practically important that we're going to overturn our own precedent. so i think that would be the safeguard. justice barrett: so, when you say that the bottom-line holdings, you -- you've kind of changed the level of generality, right? if you say the bottom-line holding is that the agency's interpretation is lawful, you think it's not open to people to come back then and say, well, it's actually not lawful, this is wrong. the court got it wrong because the best interpretation isn't the agency's. martinez: i -- i think litigants could make that argument, but i think they would have to overcome the normal stare decisis test, which is very hard to overcome, and so they would probably have to show that it's really wrong and really practically important.

and i think most courts, and i imagine this court, is -- is going to find that that threshold is -- is met, like, almost -- very rarely, maybe almost never. and so, as a practical matter, you're not going to be upending, you know, those -- those bottom-line decisions -- justice barrett: ok. martinez: -- even if you let people in theory come and challenge them, which they can do now. justice barrett: so let me ask you -- you -- you just referred to the, you know, serious stare decisis threshold, you know, that would have to be overcome. martinez: yeah. justice barrett: so let's talk about the stare decisis threshold here. why is it different here than it was in kisor? you know, in kisor, the court declined to overrule auer and the part -- the opinion that was for a majority of the court was

largely it was on stare decisis grounds. so why would a different result obtain here? martinez: i think my first answer is that the chief justice's opinion suggested it might be different and i think the reasons why it's -- it's reason -- it's -- it's -- it really is different is because there are important differences between chevron and auer. the most important that i think plays on the reliance question is this idea that chevron allows and -- and almost like a feature of chevron, not a bug, is that it encourages and allows agencies to flip-flop. and so the reliance consideration with respect to chevron is -- is much, you know, weaker for -- for -- for the government's side because the agency is allowed to flip-flop all at once, whereas, with our deference, the idea is that the agency -- it's going to be very hard for the agency to flip-flop. so i think it's more important to correct chevron because it's -- it has that mistake that auer doesn't. there are other differences. you know, chevron is problematic because it lets agencies say what congress intended or what congress's meaning was, as opposed to just saying what they themselves meant with the regulation that they themselves enacted. so i think the -- the kind of -- you know, the deference makes more sense when you're deferring to the entity that actually created the provision in question as opposed to deferring to their interpretation of -- of a provision that was created by congress. i think, in addition, you know, chevron is not limited to agency expertise. auer is limited to agency expertise. so auer is -- is narrower. and then, finally, i do think there's a difference even with respect to the apa where i think the apa more clearly puts constitutional interpretation and statutory interpretation on equal footing, and that might play into the analysis. you know, this court, the plurality in -- in kisor sort of emphasized that -- that the apa was enacted after seminole -- a

year after seminole rock, and so maybe that was a basis to think that -- that congress was okay with something that looked like auer deference. but that's not true here. chevron came many years after the apa. so i think there are a lot of differences that really flesh out, i think, the important point that the chief justice was making, which was that the analysis there doesn't automatically transfer over to chevron. justice barrett: thanks. chief justice roberts: justice jackson? justice jackson: so i've heard you say several times that you agree that judges should not be doing policy, they should be doing law. and i guess i too agree with that, and my concern is that it's actually not as easy as it seems to distinguish between the two and -- and that it appears in a lot of your answers that you sort of say, well, you come up with the best answer, it's a legal question. but i'm not so sure it's a legal question as opposed to is it the best under the sort of policy regime.

and i think that there's a real separation-of-powers danger here to the extent that you're saying that the judges are deciding whether or not this is something the agency should do or not, whether this is a legal question or not. you know, there's the old saying that when you're a hammer, everything looks like a nail. and i'm concerned that judges are going to look at all of the questions related to a statute and call them legal if we don't have something like chevron that requires judges to be actually thinking about their proper role relative to this issue. so how can you assuage my concern in that regard? martinez: so i think two points. i think the first point i would make on the distinction between law and policy and how they kind

of maybe seem like they blur together, i think that -- that there are just so many instances in which a court can get a question that comes before it that maybe it involves an agency regime, but the agency hasn't acted yet. and i think the court in that circumstance just does its best. it doesn't have guidance, it doesn't have instructions from the agency. it does its best. and i think, when it does its best -- justice jackson: but does it have to, mr. martinez? i mean, there are -- there are other regimes in which a court is presented with a question and it identifies it as a policy question that it cannot answer. so what i'm saying is that it's not necessarily true that just because the court gets an issue, it automatically says, oh, this must be legal, i have to act. martinez: but, if -- if the court got -- just to go back to justice kagan's hypothetical, the question of what -- what is a dietary supplement and the agency hadn't acted, i think the court would absolutely give meaning to that. and i don't think the court would think that what it's doing is making policy. justice jackson: well, let me give you a -- a particular example, all right? in the food and drug and cosmetic act situation, new drugs can be approved only if an adequate -- "adequate and well-controlled investigation" shows that the drug will have its attend -- intended effect. this term, what is an "adequate and well-controlled investigation," is it your view that congress wanted the courts to decide what it means for a study to be adequate or well-controlled? i mean, how would a court go about

determining whether that's something it's supposed to be doing or the agency is supposed to be doing? martinez: i think that the -- the court would -- would do exactly the kind of analysis there that it would do if it had that exact same statute without the agency acting. and i think what that means is the court would go in and it would do everything that -- that we all agree happen -- should happen under step one. i think the only difference is that if, after doing that step one analysis, the court concludes that there's a better view and a less better view, then the court should just go with the better view. justice jackson: but when -- when does the court decide that this is not my call? martinez: well, i think at the -- justice jackson: i guess that's the part that's dropping out for me in your analysis. you just say, you know, we do a step one analysis and then the court makes the interpretive decision about what this means. and i guess -- martinez: i -- i -- i don't think the court ever says that it's not my call if the question in front of it is a question of statutory interpretation, because i think that's a core job --

justice jackson: so every statutory interpretation question is one of law that a court can decide, you're saying? martinez: yes, and that -- justice jackson: there's never a statutory interpretation question that is one of policy that you see congress may have been intending the agency to answer? martinez: i think, by definition, if we're talking about interpreting a statute, then you're talking about a legal question in the same way that if you're talking about interpreting the constitution, then you have a constitutional question. no one would say that you would apply deference there. justice jackson: so there's never a world you -- maybe we just differ on this. i'm worried about the courts becoming uber-legislators, that when we have a policy -- so one way that some of the experts have looked at this, some of the legal -- legal scholars have looked at this, is that they say, when there's an ambiguity, there are actually different kinds of ambiguities. so you might have a situation in which there's a statutory term and it's ambiguous in the sense that there are several reasonable meanings of what "stationary source" might mean, for example, several different ways that you could define that.

when you get down to that level of analysis, the question is, who's going to make the choice as between what those meanings are? and i hear you saying there might be a best choice, but i guess, if we're talking about a policy question, there are several reasonable meanings, why should the court be the one to make that determination? martinez: i -- justice jackson: and -- and couldn't we be in a world where congress intended for the agency to actually decide which choice is best? martinez: i think where i -- where i would just sort of disagree is what you said at the end when you sort of assumed that it was a policy question. i would just say that if it's -- if the question is the meaning of a statutory term, that's an interpretive question that's a legal question and would be treated as a legal question if you got that exact same question before the agency had acted. justice jackson: all right. let me ask you one more thing about practical implications.

so let's say it is, you know, a legal question, as you have analyzed, adequate and controlled investigations. if i'm an agency and i'm trying to be responsible, how is this going to work as a practical matter? is the agency going to go to court every time it gets one of these undefined terms in a statute and seek, you know, a declaratory judgment as to the meaning of "adequate and controlled" -- and "well-controlled investigations" before it goes forward with its policy? martinez: no. justice jackson: all right. so the agency can come up with its own definition and implement it and then wait to be sued with respect to that, and -- and every term undefined in a statute we're going to have litigation about? think what the agency has to do is what everyone else has to do, which is try to figure out what the -- what the law means and then act accordingly, and if someone challenges that, then that'll get sorted out. if there's a -- a stat -- a legal question, a statutory interpretation question, then that'll get sorted out by the courts. but the agency isn't, like, paralyzed -- justice jackson: what do we do about the -- the chaos that we talked about in -- in the city of arlington case that comes from perhaps having different courts, right? we have 11 different, you know, jurisdictions that have legal authority.

so something like the definition of "adequate and well-controlled investigations," you say the courts will sort it out. well, first of all, it will take years perhaps for the courts to sort it out. what is the agency supposed to be doing in the meantime? and different courts from all of these different jurisdictions could actually have a different view, as justice sotomayor pointed out, of what "adequate and well-controlled investigations" are supposed to do, so -- means. so isn't it sort of impractical and chaotic to have a world in which every undefined term in a statute is subject to litigation if you're trying to govern? martinez: well, i -- i don't think it's impractical. i think that to the extent that justice kagan's questions sort of indicate that there's actually a relatively small set of cases in which chevron's going to make a difference, you're going to have that same problem with respect to the cases that maybe 20 years ago under a looser approach to chevron wouldn't have gotten deference.

justice jackson: wouldn't you have more of a problem in a world in which we've gotten rid of chevron because it's going to give incentives to parties to raise legal issues that they wouldn't have raised before? martinez: i -- i don't think it's a problem to -- to have parties, if they think an agency is overstepping the boundaries and if they're right that -- justice jackson: no, i understand, but, under a chevron regime, right, if that's the background rule, then you're going to have parties thinking twice before going down a litigation road with respect to a term because they're going to say, at the end of the day -- martinez: right. justice jackson: -- the agency has a reasonable interpretation, that's what the court's going to find, so it's not any -- martinez: right. u're -- you're going to have parties being less likely tohallenge agency action that is unlawful under the best interpretation of the statute because they know that when they go into court, the judge is not going to apply its independent neutral judgment and instead is going to tilt the scales and defer to the agency.

justice jackson: thank you. martinez: and -- chief justice roberts: thank you, counsel. general prelogar. oral argument of gen. prelogar: mr. chief justice, and may it please the court: the chevron framework is a bedrock principle of administrative law with deep roots in this court's jurisprudence. overruling a precedent is never a small matter, but overruling a precedent as foundational as chevron should require a truly extraordinary justification, and petitioners don't have one. they say that article iii requires de novo review of all statutory interpretation questions. but that's flatly inconsistent with precedent going back to the marshall court and with the traditional limits on mandamus jurisdiction, which governed most judicial review of executive action in the early republic. they've said that chevron violates due process. but the application of deferential standards of review doesn't constitute impermissible bias. and they contend that the apa requires de novo review. but that theory is inconsistent with the statute's history and the way it's been understood ever since its enactment, including in the more than 70 cases in

which this court has relied on chevron to sustain an agency's interpretation. on top of all that, reliance interests in this context are at their apex. congress, agencies, states, regulated parties, and the american public have all relied on chevron and the regulations upheld under it to make important decisions that could be upended by overruling that framework. thousands of judicial decisions sustaining an agency's rulemaking or adjudication as reasonable would be open to challenge, and that profound disruption is especially unwarranted because congress could modify or overrule the chevron framework at any time. congress has many times considered proposals to do so, but it's never taken that step. instead, congress has legislated for decades with chevron as the background rule informing the degree of discretion that congress has chosen to confer on federal agencies. just five years ago in kisor, this court declined similar calls to

overrule the auer deference doctrine based on many of the same flawed arguments that petitioners are making here. the court observed that it would be the rare overruling that would introduce so much instability into so many areas of the law, all in one blow. overruling chevron would be an even greater and unwarranted shock to the legal system. i welcome the court's questions. justice thomas: general, section 706 of the apa was not mentioned in chevron. how would you reconcile the requirements of -- on this -- on federal courts under 706 with your view of chevron? general prelogar: section 706 says that courts should decide all relevant questions of law and interpret statutes, but none of that is inconsistent with the chevron framework because 706 doesn't prescribe a universal standard of review to govern those kinds of statutory interpretation questions. and the courts are interpreting

statutes when they walk through the chevron framework. first, there's all the work that the court does at step one of chevron. that is using the tools of interpretation to identify whether congress has spoken to the issue in the case and, if so, chevron said that's the end of the matter. so, in that sense, in a step one case, the court has, of course, interpreted the statute. but, in a situation where, at the end of that interpretive process, the court is left with no conclusion that it's actually able to ascertain that congress has spoken, then, in that circumstance, i think the right interpretation of the statute is that congress left a gap or maybe created an ambiguity and simultaneously vested the agency with the important responsibility, pursuant to an express delegation, to administer that statute with the regulations that have the force of law.

and that's within -- tells the court what the relevant question of law that's left over to resolve is. it's whether the agency acted within the bounds that congress itself prescribed. so i don't think there's any fundamental incompatibility with section 706 and what chevron dictates about how to think about congress's delegations. justice sotomayor: can i say, counsel -- general, i know plenty of statutes where congress uses the word "de novo." it didn't here, correct, in 706? general prelogar: that's correct. justice sotomayor: i thought it, and i do think it, would be revolutionary to say that congress can't limit judicial review. aedpa is the quintessential question where we not only give deference to state court decisions, we say even if it got it wrong, if it didn't get it unreasonably wrong, we are superseding the court's ability to declare a violation of the constitution and give relief. so i -- i -- i think it would be radical to say that congress couldn't implement chevron. in fact, there is legislation to overrule chevron, requiring de novo review, that hasn't passed. there are statutes that basically don't -- say apply de novo review, correct?

general prelogar: yes. justice sotomayor: and there are statutes that require differential review explicitly to legal questions, correct? general prelogar: yes. justice sotomayor: besides chevron? general prelogar: yes. justice sotomayor: all right. so now we have -- we're now at 706. and my -- your adversary, your opposing counsel, said that he didn't see that much disruption from overruling chevron, that nobody would really bring up those old cases. do you have a view on that? general prelogar: i think that my friend, it -- it might be easy for him to say that because

he is not going to be involved in the endless litigation that i think would result if this court were to overrule chevron. i understand his point to be that all of the holdings in those cases will be secure because stare decisis will apply in those contexts. but the important thing to realize is that in those cases, as justice barrett's questions emphasized, the court has decided that what the agency did was reasonable. the statute has essentially been interpreted to vest the agency with discretion such that the agency's regulation is being held lawful or valid on the basis of reasonableness, and i think that that means that litigants will come out of the woodwork seeking to open those decisions and contending that they didn't actually address what they now say is the relevant question, not whether the agency's interpretation is reasonable or whether the regulation can be upheld on that basis, b how the statute should be interpreted without granting any deferce to the agency's interpretation. chief justice roberts: counsel, i'll ask you the same question i asked your friend. you began by saying chevron is foundational. we get a lot of statutory interpretations from agencies, and i don't know whether it was 14 or 16 years, we haven't relied on chevron over that time. i -- i mean, have we overruled it in practice even if we've let the -- had to leave the lower courts to continue to grapple with it?

general prelogar: no, i don't think so, mr. chief justice. it's been eight years since this court relied on chevron at step two, but there's no case that my friends have been able to point to where the court has said that a statute was ambiguous or left a gap and chevron would otherwise apply, but the court is not going to defer in that circumstance. i think that that -- chief justice roberts: no. but, i mean, that's simply a function of the fact, when -- when we go through the work of trying to interpret what a statute means, when we get to the end, that seems to be the right interpretation, and -- general prelogar: i agree. those are step one holdings. so i -- so i think that they are consistent with the chevron framework. and the fact that this court hasn't had a step two case in recent years in no way indicates that in those cases where congress is, in fact, leaving ambiguities or gaps, chevron no longer sets the right ground rule for understanding the scope of the delegation. justice kavanaugh: can i ask you

about what i see is an internal inconsistency in chevron itself? it relates to footnote 9, which is -- instructs that a court should use all the traditional tools of statutory interpretation before getting to step two. my concern about that or my confusion about that is, if you use all the traditional tools of statutory interpretation, you'll get an answer. and we know that because, in cases where we don't have an agency involved and we use those same traditional tools, we get an answer. so how do we deal with footnote 9, which seems to suggest that you'll never get to step two if you follow footnote 9 by what it says? general prelogar: so what the court said in footnote 9 is that the court should use all of the traditional tools to ascertain whether congress had an intent on the issue. and that, of course, is an important part of this framework because, if congress actually spoke to the issue, then the agency doesn't have any discretion to act in a way that's contrary to congress's express direction. justice kavanaugh: do you think that's different from ascertaining what the statute means? general prelogar: i think that there can be a relevant difference and it touches on exactly what you were asking about in the context where a court has to do it without an agency. in that circumstance, i think it's absolutely right that the

court is ultimately going to keep working and decide how it thinks the statute should best be administered, even in the circumstance where there might be an ambiguity or a gap to fill. but what chevron recognizes is that there is a third option available. it's not just congress spoke to the issue and it necessarily authorized what the agency did or congress spoke to the issue and it prohibited what the agency did. there is a category of cases and statutes out there where, really, using all of the tools, the best interpretation of the statute is that congress didn't resolve it. it left that gap or ambiguity and coupled it with this express authorization to the agency to carry that statute into effect. this is congress and the agencies working together hand in hand to put into effect this -- justice kavanaugh: how would you define ambiguity or how would you, if you were a judge, say, yes, this is ambiguous or no, that's not ambiguous? general prelogar: so i would draw on what the court said recently in kisor where it said a statute is ambiguous when the court has exhausted the tools of interpretation and hasn't found a single right answer. and i recognize, justice kavanaugh, and you have expressed these concerns that

there are some limits of language here and it's not subject to precise mathematical quantification, but that's because i think it's a standard that inherently requires the application of judgment. and at the end of the day, what the court should be looking for and asking itself is, did congress resolve this one? do i have confidence that actually i've got it, i -- i understand what congress meant to say in this statute and it meant to proscribe a -- a uniform approach to stationary source, that it has to be plant-wide or it has to be a particular piece of equipment? but, in a circumstance like chevron itself with stationary source or some of the examples that the justices have been talking about with reasonable or feasible, i think you can get to the end of that process and a judge could say: i think, actually, the way -- the right way to understand this statute is that it's conferring discretion on the agency to take a range of permissible approaches. justice kavanaugh: do you -- do you think it's possible for a judge to say, the best reading of the statute is x, but i think

it is ambiguous and, therefore, i'm going to defer to the agency, which has offered y? general prelogar: no, i think that that would probably -- justice kavanaugh: that can't happen? i think that happens all the time. general prelogar: well, i think that there are two different ways in which courts use the term "best interpretation of the statute." so, if what you're asking me is, is there a world in which a judge could go through the rigorous step one inquiry, apply all of the tools, and say, i think there's a best interpretation insofar as i think congress spoke to the issue, but the agency's interpretation is it could be permissible, i recognize there's some doubt here, the answer is no. chevron does not require a court to ignore what is ascertained doing the step one inquiry. at that point, that is the -- the judge's conclusion that congress actually spoke to the issue and chevron is totally clear about this, give effect to it. but, if what you're asking me is, is there a world in which the court could get to the end of the step one inquiry, decide that congress hasn't spoken to the issue, and then say, if, in fact, the courts had been given the role of filling the gap, i would have done it differently, i would have exercised whatever discretion that congress left open in this statute in a different way, even looking to things like the overall objectives in the statutory

program as a whole, then yes, of course, in that circumstance, it's -- it's implementing congress's directives -- justice gorsuch: i mean, general -- general prelogar: -- for the court to not -- interrupt, but those are two different -- very different views about what qualifies as an ambiguity you've just given us. one is there is a better interpretation. i provide it as a court. the other is: well, yeah, but i'm going to defer anyway given whatever considerations you want to throw into the ambiguity bucket. and that's exactly the problem that your friends on the other side suggest have persisted in the lower courts for 40 years and why some judges claim never

to have found an ambiguity and other equally excellent circuit judges have said they find them all the time. and it's also why, i don't know, maybe a dozen or more circuit judges have written asking us to overrule chevron. and -- and -- and -- and -- and it also may be why one of your colleagues last year said i don't know what ambiguity means at this lectern. and should that be a clue that something needs to be fixed here, that even the federal government at the podium can't answer the question what triggers ambiguity? you've given us two different alternatives today, and so many lower court judges who just want to follow whatever we tell them to do faithfully can't figure it out. general prelogar: so there's a lot packed in there, justice gorsuch, and i want to respond to each of your concerns. first, i would draw from chevron and kisor in defining what is an ambiguity. it is when a court has applied the tools of construction and can't ascertain that congress had an intent on the matter. so i think that that is the core question for a court at step one of chevron, and if that's the circumstance, that would only ever move a court to applying deference at step two.

now i understand the concern you expressed that maybe lower courts are too reflexively finding that there's ambiguity at -- justice gorsuch: well, you gave us a second definition just a moment ago, and -- general prelogar: i was trying to -- to explain how i thought that sometimes -- justice gorsuch: some -- yeah. general prelogar: -- in the case law "best interpretation" -- justice gorsuch: yes. general prelogar: -- is used in two different -- justice gorsuch: right. general prelogar: i don't think that's a different understanding of chevron. justice gorsuch: well -- general prelogar: i think that's really a difference -- justice gorsuch: -- your -- your friend -- general prelogar: -- between step one and step two. justice gorsuch: -- your friend a year ago thought so and -- and -- and lower court judges think so. general prelogar: so let me respond to the concern -- justice gorsuch: so you agree -- general prelogar: -- about lower court judges. if you think that they are too readily finding ambiguity, i think the court could do in this case exactly what it did in kisor -- justice gorsuch: -- we done that, like -- like, 15 times over the last eight or 10 years, say, really, really, really, go look at all the statutory tools, and yet here we have a case, two cases, one in which one court found ambiguity and went to step two and another one which -- well, i can't tell what it did, but there's a pretty good argument it -- it tried to resolve it at step one. so, even in a case involving herring fishermen and the question whether they have to pay for government officials to

be onboard their boats, which may call for some expertise, but it doesn't have much to do with fishing or fisheries, it has to do with payments of -- of -- of government costs, we -- we -- lower court judges even here in this rather prosaic case can't figure out what chevron means. general prelogar: well, i do think that issuing a reminder to courts about the thoroughness -- justice gorsuch: another one? general prelogar: -- that's necessary at step one could make a difference in this context. and i can just share anecdotally on behalf of the government that we have canvassed the litigating components and looked at the lower court case law. and after kisor, lower courts granted auer deference far less frequently, so i think it can matter and that lower courts can get that kind of message if you're worried about it. but, justice gorsuch, the other point to add here is that if you are concerned that lower courts have different reactions in trying to implement chevron at step one, i think it's important to think about the alternative as well.

it's not as though, if this court overruled chevron, that's going to get rid of statutory gaps or ambiguities. justice gorsuch: no, it takes -- general prelogar: they will persist -- justice gorsuch: -- us back to skidmore, which justice jackson, the most ardent of new dealers, wrote and that persisted in this court for 40 years, more or less, after the apa. and the world seemed to continue on its axis just fine. general prelogar: but it's not going to create greater predictability or stability or consistency across judges. justice gorsuch: that's -- that's -- general prelogar: if anything, i think that -- justice gorsuch: -- an interesting thing to suggest, that chevron predicts stability, when the whole point -- i didn't see you mention brand x much in your brief. but i -- i'm sorry to go back there, but -- my good friend, but brand x is a recipe for instability, isn't it, because each new administration can come in and undo the work of a prior one. they're all reasonable. i mean, my goodness, the american people elect them. of course, they're reasonable people. (laughter.) justice gorsuch: and -- and -- justice sotomayor: that may be the first --

(laughter.) would have thought that chevron, at least as this court's understood it, is a recipe for anti-reliance. general prelogar: so i disagree with that characterization about brand x, and i think my friends have created, kicked up some dust about exactly what brand x does -- justice gorsuch: so you do -- general prelogar: -- and doesn't do. justice gorsuch: -- you do endorse brand x, the government does? general prelogar: yes. i think it is a logical follow-on of chevron, and here is why. as brand x itself recognizes, if the court has found at step one that congress spoke to the issue, there's no room under brand x for the agency to reverse the court or somehow change the underlying meaning of the statute. instead, the statute has been interpreted at step one and what congress says goes.

it's only in the category of step two cases where brand x comes into play, and in that circumstance, it's because the court in the prior case has understood the statute to leave a gap or an ambiguity for the agency to fill, considering a range of regulatory approaches. so, in that circumstance too, the meaning of the statute doesn't change. it remains a gap for the agency to fill at time two, and if the agency is running through all of the procedural hoops, which can be quite burdensome in this context, to change its regulatory approach, it is still acting consistently with the -- justice gorsuch: or not. general prelogar: -- with the discretion. justice gorsuch: or not if it -- if it issues an interpretive rule without notice and comment or issues an adjudication. it may or may not be that burdensome, right? so brand x also says that an agency can overturn a prior judicial interpretation. and i saw that as a circuit judge with respect to an alien who was allowed into the country under the tenth circuit's understanding of the law. and the government come back and says, no, you have to overturn your precedent, tenth circuit, and he's not allowed in the country. and we had to overrule our judicial precedent. do you think that's an appropriate understanding of the law too, that judicial precedents, maybe even precedents of this court, can be overturned by agencies? general prelogar: it depends on what the judicial precedent held. if it held at step one that that

statute was clear, then of course not. but brand x doesn't require that result. if the prior precedent held that congress didn't resolve the issue and had delegated to the agency the responsibility and role in administering it and filling the gap, including with the possibility of changing regulatory approaches based on things like change -- justice kavanaugh: but the reality -- just to pick up on that, the reality is -- you -- you say don't overrule chevron because it would be a shock to the system, but the reality of how this works is chevron itself ushers in shocks to the system every four or eight years when a new administration comes in, whether it's communications law or securities law or competition law or environmental law, and goes from pillar to post, like professor pierce wrote, and he had been a fan of chevron. now he's not because he says it's a source of extreme instability in the law. that's his -- his phrase. and it just seems like you just pay attention to what happens when a new administration comes in at epa, at sec, at ftc, you name it.

it's just massive change. that is at war with reliance. that is not stability. and so i think to hold up stability and reliance is a little tough given just watching how it operates every four years. general prelogar: well, let me give you a couple of different reactions to that. i think that that is a small sliver of cases or circumstances. and in the mine run case involving agency regulations, agencies themselves build on those regulations as a foundation. there's no evidence that agencies are out there flip-flopping left and right or doing so on a whim. and it brings me to the important point that to do -- justice kavanaugh: i don't think they're -- i'm sorry to interrupt -- and i'll let you finish. but i don't think they're doing it on a whim. i think they're doing it because they have disagreement with the policy of the prior administration and they're using what chevron gives them and what they can't get through congress to do it themselves, self-help,

and to do it themselves unilaterally, which is completely inconsistent with bicameralism and presentment to get your policy objectives enacted into law. general prelogar: but, justice kavanaugh, the premise i think that's embedded in that question is the idea that congress had spoken to that issue. and in a circumstance where congress didn't resolve it and, in fact, wanted the agency to have flexibility and a range of options, there's nothing inherently problematic or incompatible with our system of government to recognize that agencies can carry out those directives. and just look at "stationary source." you know, that was a circumstance where the court said, applying all of the tools, congress didn't have a view on it. it didn't want to foreclose a plant-wide definition. it didn't want to foreclose an equipment-specific definition. and i think it was entirely permissible for the expert agency to come in, take stock of the entire situation, and, yes, take account of the policy goals of an incoming administration to better account for the interests of the regulated parties and give them flexibility. that's just part of congress's design. justice jackn: after all, you know, taking into account the policyoals of the new

administration reflects a democratic structure where we have the new administration being elected by the people on the basis of certain policy determinations. i guess my concern is i suppose judicial policymaking is very stable but precisely because we are not accountable to the people and have lifetime appointments. so, if we have gaps and ambiguities in statutes and the judiciary is coming in to fill them, i suppose we would have a -- something of a separation of powers or policy -- excuse me -- separation of powers concern related to judicial policymaking. am i wrong to be worried about that? general prelogar: no. i think that that concern is valid, and i think it's valid along two separate dimensions, and one is to recognize that in these scenarios where we're at chevron step two, by definition, it's because the statute itself doesn't supply an answer and the

court can't ascertain that congress actually meant to resolve it. and in that circumstance, it's entirely sensible for congress to give the issue to an agency when it is charged with administering the statute and, of necessity, is going to have to fill the gap along the way. and congress could quite legitimately want the agency to draw on its policymaking expertise in figuring out the right way to fill the gap. justice jackson: what do -- what do you say to mr. martinez, who says we've already characterized that as a question of law because the court was involved at step one in making the determination, and so it seems a little odd -- i think i took this away from his presentation -- to suddenly say, when we're in a step two gap-filling world, now we're going to call it a policy question as opposed to a legal one? general prelogar: so i think you can still characterize it as a legal question while recognizing that in a circumstance, to borrow justice kagan's words, where the law has run out and congress hasn't actually spoken to the issue, the court, if it resolves that issue, is -- is going to have to draw on a set of considerations to inform its judgment. and i wouldn't call it

policymaking, but i do think it means that the court can't suggest that the answer it is giving is absolutely dictated on that precise issue by congress because, by definition, we're in a world where congress didn't speak to it. so the court will have to take account of a narrower range of circumstances, things like the overarching statutory objectives, to try to fill in the gap. but the point is that when congress has left that gap and charged the expert agency with the administration role, congress could have every expectation, and chevron says congress has the expectation, that the agency will fill the gap and that the courts will respect it within the bounds of reasonableness that always apply in this context. justice barrett: general prelar, most scholars of statutory interpretation consider chevron to be an interpretive canon, much lik clear statement rules, rule of lenity, judicially created. do you see chevron that way? and, if so, do you see it as different in kind from any of the other canons of interpretation that we apply? general prelogar: i do think it is different. i don't conceive of it as a canon. instead, i think that it is fundamentally rooted in -- in kind of setting the ground rules

for how all three branches of the government are operating together. and what i understand the court to have been doing in chevron is recognizing that there are legitimate reasons why congress cannot answer every question itself and why it will want to go hand-in-hand with an agency by charging that agency with administering the statute. and in that circumstance, it's the role of the court to give effect to that. so i think it's not just kind of an interpretive canon, but, rather, it really is grounded in the separation of powers. justice barrett: so is it dependent on a judgment about what congress would want, one that would have to be empirically tested? general prelogar: so i don't think that it's getting into congress's subjective intent, although, certainly, i think the primary rationale that chevron gave was its appraisal that this is, as an overarching matter, what congress would have intended when it comes to gaps. and i don't mean to suggest that this means that congress thinks about each and every gap it's creating in the moment. sometimes i think it does and it's clear when it says set reasonable rates. it knows that it's not itself prescribing what those rates will be in concrete

circumstances. it's leaving gaps and the agency has to fill it. but i think, even in the circumstance where congress doesn't know it's creating it at the time, someone's going to have to come in after the fact and fill it in, and it's either going to be the agency or it's going to be the court without deference. and in that circumstance, i think the court appropriately recognized congress would want for the agency to do it. justice barrett: and how do we know -- this is -- goes back to that question of what is the trigger of ambiguity that justice gorsuch was asking you. so think about a concrete example like pulsifer, which the united states is on the other side, pending before the court, turning on what "and" joins together. general prelogar: we think that one's clear. i'll just put it out there. (laughter.) justice barrett: so let's -- put aside the question of whether, you know, the department of justice and the executive can get to deference in interpreting criminal statutes. just erase that issue from the picture. is that the kind of question -- you know, judges below, very smart, very reasonable judges reached different conclusions about what that word in the statute meant.

is that the kind of question then, you know, thinking about brand x saying, well, it doesn't have to be the best, it just has to be, you know, a plausible reasonable one, is that the kind of statutory question that would trigger ambiguity and step two deference? general prelogar: so i think it's hard to speak in generalities about this. and i am struggling because, of course, the court has recognized that the department of justice does not get deference in the criminal context with respect to that particular issue. justice barrett: and it's that statutory structure in a communication would say to just try to address the overarching question is that, you know, i think that it's going to be kind of a specific exercise in every case, and i can't say here is the formula i can give you to know when the statutory interpretation exercise at step one runs out and the court should feel like, i don't have an answer, congress didn't supply one and when not. i think it's going to vary based on the statutory scheme. but, in each case, the court should conduct that inquiry, make it a thorough inquiry and take account of all of the

relevant aspects of interpretation that can bear on meaning and show that congress, in fact, did resolve it. that is the role of the court, and it's the role of the court likewise to enforce congress's directions. justice barrett: so that kind of question, putting aside the government's position in pulsifer, so maybe -- that's an unfair question to ask you, but that kind of question you think would be the kind of question that could -- you know, let's take it outside of what does the word "and" mean. you know, a question of statutory structure, the placement of a comma, you know, that kind of a thing, that is the kind of question that, depending on the circumstance, could trigger step two deference? general prelogar: i think it conceivably could. now i want to hold open and acknowledge that the court has said there are certain types of statutory questions that don't fit within the chevron framework because there are kind of statute-specific reasons to think congress wasn't giving this question to the agency. i think the major questions doctrine is a species of that. i'd point to the adams fruit case as well where it was a judicial review provision and the court said this wasn't something for the agency to do. but i think, in the mine run

case, yes, and -- and to the extent you're saying, well, it feels odd for it to depend on a comma or to turn on the meaning of the word "and," still i think the inference holds because, in that context, congress, if it, in fact, has left the ambiguity or the gap, recognizes that the agency is going to have to come up with an answer. justice barrett: except a lot of times congress doesn't intentionally leave the ambiguity or the gap, right? it's just limits of language, limits of foresight. general prelogar: yes. so i think a court ultimately, if it's able to ascertain that, although it's not perfectly clear in the statute, you can figure out what congress intended, give effect to that, that's step one. at least congress knows that if it's going to unintentionally create ambiguities or gaps, chevron is the stable background rule. it's been the rule for 40 years. this court acknowledged in city of arlington that congress, in fact, legislates against the background of that rule, and so it knows that with anything it's doing that's unintentional, that will trigger if the predicates are satisfied. justice kavanaugh: i'm sorry.

can i ask you about the phrase "law runs out." one way to think about that would be if you had the same statutory interpretation -- chief justice roberts: go ahead and finish, sure. justice kavanaugh: same statutory interpretation issue in a non-agency case, could the court decide it? and if the answer is yes, the court could decide it, then the law hasn't run out, so, therefore, you could ask yourself that question in an agency case. if this were a non-agency case, would we come to an answer on this case? and if so, you don't go to step two. what's wrong with that? and if that's not correct, because i don't think you're going to agree with that -- how would you define when the law runs out short of that, which i think is a problem, as you said, hard to speak in generalities about this. that's the problem. general prelogar: yes. so you predicted my answer. i don't agree that it's only in a circumstance where the statute would be incapable of the court issuing a decision at the end of day. of course, if a case comes to the court and it has to resolve it, it's going to have to do its level best. but what i meant by the law running out is that if the court

has walked through all of the tools of construction and interpretation and doesn't think that congress actually directly spoke to this issue, congress itself didn't resolve it, then the kinds of tools the court is going to have to use will be ones that sound in things like the overarching statutory objectives that congress revealed as part of its plan. and i think that in a -- a chevron circumstance, the insight of the court's opinion there was that the court doesn't have to go on and itself supply the answer when, actually, the best way to understand congress having not resolved it itself was to make the primary decisionmaker or the person with the primary role in the first instance to be the agency. chief justice roberts: thank you, counsel. justice thomas, anything further? justice thomas: just a -- a couple questions. you said that in an exchange with justice sotomayor and me that congress could require some deference when it came to questions of statutory interpretation.

and in 706, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, et cetera. could congress also require deference on the part of the court with respect to constitutional issues? general prelogar: so i think that that would raise distinct issues in light of the different history that would be in play in that kind of hypothetical. there has not been a longstanding history of courts deferring to agencies when it comes to interpreting the constitution, so i think there could be a unique article iii interest at stake there. but the history runs in precisely the opposite direction when it comes to statutory interpretation, where agencies themselves are charged with administering it because, as we've tried to explain, chevron was not an innovation, it was not something new. these principles of deference go all the way back to the very founding years of the republic. they're reflected in things like

mandamus practice, where virtually all executive action for the first hundred years of our nation's history was reviewed deferentially, and then it was continued in a long line of cases from this court recognizing specifically that in a circumstance when you have the executive administering the statute, congress could delegate and could expect for those delegations to be respected. justice thomas: i think mandamus is a little bit different and the other extraordinary writs in that you had quite a high hurdle before they became applicable, but we normally say that this court reviews questions of law de novo, and that includes statutory and constitutional. how would you distinguish that normal practice from what you're saying? general prelogar: well, i think it is more nuanced than that. i certainly take the point that the court reviews many legal questions de novo, but that's not invariably the case. there can be issues that arise under distinct statutes that set

forth more deferential standards of review. aedpa is a good example of that. there can be circumstances like mandamus where the nature of the action itself dictates a more deferential standard of review. and i just don't think it would be accurate to say as a uniform, across-the-board matter, de novo is the standard that always and invariably applies. that's inconsistent with cases from this court that were cited in chevron, going back to the early 1800s, things like edwards' lessee versus darby, where the court itself was recognizing that in a variety of contexts where you have ambiguity in particular and you have an expert agency charged with administering the statute, deference can be warranted. justice thomas: thk you. chief justice roberts: justice alito? justice alito: can you provide a concise definition of what "ambiguity" means in this context? general prelogar: ambiguity exists when the court has exhausted the tools of interpretation and hasn't been able to arrive at confidence that there is a right answer that congress spoke to the issue. justice alito: well, as justice kavanaugh's recent question

presented, in cases that don't involve an agency, we never say we have exhausted all of our tools of interpretation and we just can't figure out what this means. so that would seem to suggest you never get to step two. general prelogar: but the relevant question at step one is whether congress is, in fact, resolving it or delegating it to the agency. so i agree that in a circumstance where you don't have an agency, the court can't give effect to any delegation and, instead, the backup option in a situation where an agency would otherwise be available is the court has to do it, but i don't think that that undermines the very real on-the-ground possibility that congress is legislating and meaning to give the agency the gap. justice alito: well, i come back to the question of your definition of ambiguity. and what i heard you say the first time was it's when we've used up all our tools and we can't figure out what it means, then it's ambiguous. so do you want to provide an alternative definition?

general prelogar: so i think maybe the best way to try to clarify what the definition i'm trying to give is to use an example of something like a statutory term like "reasonable." justice alito: -- really would just like a definition so that all the courts that have to apply the regime that you're advocating will be able to apply it in the many different cases that come before them. general prelogar: the court gave this definition in kisor five years ago with respect to auer deference, and i think it's the right definition to use -- here as well. justice alito: what is it? general prelogar: when a court has used or exhausted the tools of interpretation and doesn't believe that it reveals a right answer. in that circumstance, chevron said the right way to think about that statute -- the real right answer there is a delegation. justice alito: but again, i think you -- you're running into the problem that we never do that in cases that don't involve an agency. general prelogar: because in those cases -- justice alito: so i think you've got to provide a different definition. now, what i heard you say at a couple of times during your argument was it's when we can't

figure out what congress intended. is that what you mean to say? general prelogar: that is the inquiry that chevron prescribes that you should be -- and this is drawn from footnote 9, which is another formulation of this, use the tools of interpretation to see if they reveal congress's -- justice alito: what do you mean by what congress intended? do you mean to say that you get to step two whenever we don't think that a majority of the house and a majority of the senate had an intent on the specific question that is before the court? then you'd always get to step two. general prelogar: no. so i don't think it's about individual legislators' intent. i think the court in chevron used the word "congress," but you're really looking at the statute and what the statute reveals about whether it's resolving an issue or not. justice alito: thank you. chief justice robert justice sotomayor? justice sotomayor: there hasn't been much discussion on why this is entitled to statutory -- to stare decisis consideration. there's been an argument by

petitioners that it's not really a holding of a case; it's a method only, and we have said in the past that a method that lower courts have to use is subject to change we can make without considering stare decisis. so could you address that argument? general prelogar: yes. and i think that petitioners have pointed to two relevant types of cases that they suggest just mean stare decisis doesn't apply here or it applies in particularly weakened form. first, they say the court has sometimes changed the interpretive tools it consults. things like legislative history might have been in greater favor, at least with some justices, before and maybe have fallen out of favor later. but i don't think that those provide a parallel at all, because the court has never distilled those kinds of interpretive tools into a governing framework. it's never, for example, dictated to lower courts you should be applying legislative history in all cases. and so i don't think that it has the same kind of roots in the

type of binding governing framework that chevron has, which really has functioned in quite a different way with respect to how you understand and implement congress's directives. the second case they've pointed to is pearson, which held, in the context of the saucier rule, that that was entitled to weakened stare decisis. but there the court said that is entirely a rule of internal judicial management about how courts decide issues and sequence their decision-making process. it doesn't have outward-looking consequences, and it would be foolish to require congress to step in to fix it. there, too, i think that the considerations run in precisely the opposite direction here because chevron is not just a binding framework about how courts conduct their business; it also gives notice to the legislature about how its statutes will be construed. and if the court got this wrong when chevron was decided and was wrong about legislative intent, congress is there at the ready and is perhaps the best part or institution in government to be able to correct it and actually say, going forward, what it

wants the ground rules to be. and the final thing i would say, justice sotomayor, is that these were precisely the kinds of considerations that the court took into account in kisor in applying the strongest form of stare decisis to auer deference. my friends have largely ignored kisor's analysis on this. this was the majority of the court where the court said congress can step in, these deference decisions are balls that are lobbed into congress's court, and there are big reliance interests at stake here because there are dozens in that case, here thousands, of decisions that could stand to be displaced and create chaos if chevron is overruled. so i think that, from a stare decisis perspective, that precedent counts as precedent too. justice sotomayor: there -- and you answered the reliance question, because one of the arguments on the other side is no one has -- well, the first argument, that the court hasn't applied chevron in how many years, and so nobody should have legitimate reliance interests. and the second argument against reliance is that no one should have reliance on a wrong

interpretation, basically. general prelogar: yes. and i think that those kinds of arguments are inconsistent with kisor and also inconsistent with what we know about what happens in the real world. there are agency regulations out there that have been on the books for decades. people have made investment decisions on the basis of that. people have decided what contracts to enter into on the basis of that. states in cooperative federalism programs have designed and invested the resources into their share of that program. and all of that could be thrown into disarray if now it can be subject to renewed challenge on the basis that that regulation was upheld answering the wrong question, not looking at whether it conflicts with some purportedly better interpretation of the statute. chief justice roberts: justice kagan? justice kagan: there's been a fair bit of talk, general, about how because you don't have a formula for saying when there's a gap or ambiguity so that you go to step two or because judges may have different tendencies, you know, which might be temperamental as much as

anything else, to find ambiguity, because of that there's going to be some variability. and it's hard to argue that it will be some variability, but could you talk about the variability in the alternative scenario? general prelogar: this is a really important point to focus on because, as i was trying to say earlier, in a world without chevron, it's not as though congress is always going to speak clearly and it won't leave gaps or ambiguities in statutes, genuine ambiguities where you apply the tools and at the end you are left with no certainty about what congress was trying to do. and in that circumstance in a world without chevron, what we'll see is what justice alito was suggesting, the courts will have to go on and try to answer the question. but there are 800 district court judges around the nation, and i think it's fair to say they will likely have different takes about what to do in that circumstance and what to give greater weight to and how to ultimately fill the gap in administering the statute. and that's going to create problems for a couple of different reasons. justice kagan: and those differences, to go back to

justice alito's earlier question, i mean, those differences were part of the impetus for chevron because those differences were looking awfully idealogical in nature, awfully partisan in nature. and chevron, all the empirical evidence suggests, dampens that kind of idealogical division between courts. general prelogar: that's right. there is good empirical evidence to support that judges have an easier time reaching common ground under the chevron framework and at least identifying when they can agree that congress did not itself resolve an issue, than they do when they have to ultimately go on and try to figure out what they are going to say is the bottom line of the best way to put the statute into operation. so i think that that is rooted in chevron, and it just reflects as well this uniformity concern, one of the basic justifications for chevron and one of the reasons why this inference of legislative intent is sound, because agencies can provide that kind of uniform rule for the nation, subject to the ground rules of course of judicial review under chevron. but i think that the alternative world where there's no chevron is that there will open up wide

disputes among the lower courts, maybe on these mine-run statutory interpretation questions in complex programs, things like medicare and medicaid, and i think that it could mean that regulated parties are subject to different rules in different parts of the country. you lose the uniformity value, and it diminishes the force of the political accountability value. so i think congress would have very good reason to think that agencies should do this and that courts should respect it within the bounds of reasonableness. >> justice gorsuch? >> you agree that courts, under the apa, have to review questions of law involving the constitution de novo? general prelogar: yes. i think there might be certain circumstances with respect to certain provisions where more deferential standards apply, but i certainly agree they don't defer to agencies. justice gorsuch: ok. and you agree that, elsewhere in the law, when posed with questions of law, courts review those de novo, generally speaking? general prelogar: i think that, in many contexts, it's de novo. certainly not in all contexts. justice gorsuch: the examples you gave, i think, were aedpa

and mandamus, right? general prelogar: yes. i think those are two good examples of situations where there are specifications of a standard of review that's more deferential. justice gorsuch: i wonder whether those have more to do with remedies, right? in a mandamus case, a court should say, or can say, what the law is. it just can't provide relief unless its conviction about the statute meaning is sufficiently clear. same thing in aedpa, that we require a heightened standard before relief is granted. same thing in sovereign immunity contexts. we may think the statute says the government's liable, but we impose a higher standard before we grant access to the fisc. general prelogar: so i acknowledge that i think that many of those doctrines do turn on limitations built into the writ or limitations on remedies. i don't think it would be right, justice gorsuch, to say that in the mandamus cases, what courts were traditionally doing is saying let me put aside what the executive officer did and just interpret the statute de novo and say what i think the right

answer is. and the right answer is the executive was violating the law, but not clearly outside the scope of the executive's authority. justice gorsuch: but he could do so just as we do in the qualified immunity context. there are two steps to that analysis. you can just go to the second one and resolve it and say, ah, it's not clear, so i can't provide a remedy. general prelogar: but i think, for petitioners to succeed on their article iii argument, they have to show not just that you can review de novo, but you have to. justice gorsuch: -- i'm not asking about article iii. i'm just asking about the apa and what it means. general prelogar: yeah. so sorry if i misunderstood. i do think, though, that what the history shows at the very least is there has been no fundamental rule in this country leading up to the apa's enactment that you have to review all questions de novo. and that's where the history of the apa really matters. this court has several times recognized the apa was a restatement of existing judicial practice when it came to review of agency statutory interpretations. and as we've explained, there are really deep roots here, a long line of precedent and history showing that courts will sometimes defer.

just as gorsuch -- justice gorsuch: you do point out cases like edwards' lessee and others where this court gave respect to the federal government's contemporaneous and uniform interpretation of the statute. and that's exactly what skidmore does. it gives respect to contemporaneous and uniform interpretations. but chevron, it doesn't matter whether it's contemporaneous and uniform. it could be novel and out of the blue and inconsistent with everything that came before and it still gets deference, right? general prelogar: i disagree with the idea that those cases stand for the more limited justice gorsuch: well, i'm -- reading from them, but okay. general prelogar: there are dozens of them. so i acknowledge that they use varying formulations, and maybe you can find some that look a little more like skidmore. i think i have a lot that look a whole lot like chevron -- justice gorsuch: let's say you don't. then what?

general prelogar: well, i think i -- i just have to dispute the look at gray versus powell, look at nlrb versus -- these are cases in the 1940s that were leading cases in administrative law. justice gorsuch: oh, i -- i put aside what happened in the '40s because it went back and forth and wound up in 10 you wanted to say it's a very old thing, and the old cases don't look anything like chevron. they look a lot like skidmore. general prelogar: i -- i disagree with that. some of them -- say you should give it controlling weight, it should tip the balance. they're not saying just pay attention to it if maybe it has the chance of persuading you. justice gorsuch: if it -- if it's contemporaneous and if it's uniform, right? general prelogar: and i just want to add as well -- justice gorsuch: i -- i have another question, though. chevron, you emphasize, is -- value-neutral and it'll sometimes favor industries that are regulated and sometimes favor the government. and i can certainly see that in scenarios where we talk about the flip-flop of administrations and new people leave -- come in and replace others and -- and there's a lot of movement from

industry in and out of those agencies. i think george stigler talked about regulatory capture. i don't worry in a chevron regime about those people. they can take care of themselves, okay? there is political account, fine. the cases i saw routinely on the courts of appeals -- and i think this is what niggles at so many of the lower court judges -- are the immigrant, the veteran seeking his benefits, the social security disability applicant, who have no power to influence agencies, who will never capture them, and whose interests are not the sorts of things on which people vote, generally speaking. and, there, chevron is almost always, i didn't see a case cited, and perhaps i missed one, where chevron wound up benefitting those kinds of peoples. and it seems to me that it's arguable, and, certainly, the other side makes this argument powerfully,

that chevron has this disparate impact on different classes of persons, and i wanted to give you a chance to respond to that. general prelogar: sure, and i have a couple of different reactions to that. one is to say that i, of course, acknowledge that the way that chevron operates, it gives effect to agency interpretations even in circumstances where that might be oppositional, some of the categories of individuals that you're identifying. but, if it does that, it does that in accordance with congress's intent and wishes because even my friend agrees that there are certain delegations that congress can make to agencies and certain gap-filling that agencies can do i don't think there is a kind of fundamental flaw. justice gorgeous -- gorsuch: it is in favor of the immigrant, in

favor of the veteran and the social security applicant, but because of a fictionalized statement congress wanted when it did not think about the problem, the government always wins. general prelogar: there are a couple of different ways to commit the concern. it is not just in the exercise of discretion the court with think something is fair and for the gap in that way, but rather the core things that actually the reason it is fair is because i have a sense that congress spoke to this. i can determine a based on all of the tools. justice gorsuch: it does not matter whether congress actually thought about it, and that there are many instances where congress did not think about it. and in every one of those chevron's exploited against the individual and in favor of the government. general prelogar: i do not think it is fair to treat that as an exploitation. congress has been aware. it can change chevron at any time. it can displace if you think that is being used in a situation not warranted. justice roberts: justice

kavanaugh. justice kavanaugh: i think the other side's argument suggests the basic analytical concern at the heart of chevron is that it treats law as policy. and if that is antithetical to our constitutional structure and the rule of law. and that is why the footnote nine question is so important. because if you use the traditional tools in a nonagency case and got an answer that suggests it is a statutory interpretation question. and you are saying, no, you can stop short of that in an agency case in some difficult to find -- define point and treat the rest of the case as a policy call for the executive branch. and that is treating what was a law question in an nonagency case as a policy question in an agency case, and it is the same question. so it is transforming law into policy. and that is very difficult, i

think, to accept, if you accept the idea that a premise of the rule of law is that the executive and the judiciary cannot just treat the laws passed by congress as mere expressions of policies that they can change. respond to that. general prelogar: i hear that concern. i think the way to address that is to readdress the principal in footnote nine. we agree that is an important principle. and to the extent there are agencies or lower courts that are effectively not giving the effect to congress's own enactments then a court can police that and put into affect the footnote 9 principal in a robust way with a rigorous analysis. that is the kind of instruction the court gave in kaiser. justice kavanaugh, i think it's not a different question in the agency context and nonagency context. what i understand chevron to be doing is figure out a congress spoke to this issue.

and if so, implement it. but hold open the possibility congress did not speak to the issue. and in that context, if congress has given the agency this primary, critically important role to administer the statute, that should give deference if the agency stays within the bounds congress set. in an nonagency case you don't have the agency to relying on. you might end up thinking congress did not precisely speak to this issue, but what is the best that i can do to figure out how congress would have resolved it or what is the interpretation most consistent with the overall statutory scheme here? the right way to resolve this case, congress would know that courts are going to have to do that in a context without an agency. it is still following the terms of the statute. i think it would be a fiction to suggest that what the court is doing there is following congress on explicit expression of the matter. justice kavanaugh are --

kavanaugh: i think it is important to distinguish statues that involve legal questions of statutory interpretation. there are tons of statues, let's go back to the a.i. example, that explicitly confer broad policy discretion on agencies. that is where state farm kicks in. and that is where we have always been deferential. general prelogar: correct. justice kavanaugh: you acknowledge those are two different kinds of statutes. one statute that says for example no one can catch more than 50 fish today. the next statute is basically the agency can define what a reasonable number of fish that can be caught in a reasonable day. the second statute confers broad policy discretion. do you agree those are distinct? general prelogar: i think that one is a clear bestowal of discretion on the agency. i think it shows congress can legislate in a variety of ways. justice kavanaugh: so you agree congress can legislate broad policy discretion to an agency or grant explicitly through

words like reasonable, appropriate? general prelogar: absolutely. i think the same question and sometimes come up in those context. congress has had to borrow from the chief justice's example, reasonable truck lengths, there isn't an agency interpretation. the court is going to have to do its best. that is actually meaning to create a zone of discretion. justice kavanaugh: that is the state farm question as i would see it. two more questions, i want to make sure the concerns of the other side get a chance to respond. some of this would be taking power and grant to get to the judiciary. there is a different conception of chevron that it is taking power from congress and shifting it to the executive and allow ed the executive in essence to unilaterally make policy without congress. and one of the concerns historically from the beginning of this country was unchecked executive power. and you hear presidents

criticized all the time, whether it is roosevelt, reagan, bush or obama, criticized for exercising unchecked power. the concern is about chevron ushering in aggressive assertions of unilateral executive power. and that is the concern that i think the other side has. not about the judiciary taking power, but the judiciary having taking it from congress and shifted it to the executive contrary to her usual concerns. general prelogar: i disagree with the characterization that chevron permits and congress is powerless to do anything. in the first instance congress has to make the delegation to the agency and the court can enforce that, and so congress knows to speak patiently -- capaciously when it wants to

bestow discretion, to speak plainly it wants to rain an agency in and resolve an issue itself. congress can change the rules of deference that apply in any context. there have been particular schemes where congress has said deference does not exist. do not apply or defer to this agency and not this other agency. congress is really in the driver seat here. justice kavanaugh: this is a technical point. most presidents would veto or get rid of the chevron deference. that's the technical points. last question, which is there was talk about democratically elected political branches. i want to get your agreement on something that i think you will agree on, the role of the judiciary historically under the constitution to police the line between the legislature and the executive to make sure that the executive is not operating as a king and is not operating outside the bounds of the authority granted to them by the legislature. you agree that is a proper judicial role? general prelogar: i agree with that, but i think chevron is consistent with that.

the court polices the executive at that one by ensuring congress his own choices are put into operation. if further polices the executive at step two as the court said in kaiser, reasonableness is a test that agency can bail. -- can fail. there is work to be done to make sure the agency doesn't transgress some outer boundary line that congress set. justice roberts: justice jackson. justice jackson: picking up where justice kavanaugh left off, that's in the court have to not only police the other branches, but itself as well? by that i mean to the extent that the other side raises the concern that, you know, they are treating law as policy. isn't there a concern that policy questions might be treated as law and that what chevron is doing is also helping the court to police its own determination in that regard? general prelogar: yes. and i think a way to illustrate is to think about a delegation like the deceptive practices as defined by the secretary. if there were a statue this

-- statute that said that, of course the court couldn't come in and say the secretary said what is a defective practice? i think there is a better way to think about the concept of what is deceptive. therefore, i am not going to override what the agency is done or not give any weight to it. congress has directed what you should do as a court is paying attention to what the secretary did. the secretary was given that role in administration. obviously chevron applies to circumstances that delegation is to identify the same basic idea where i think the courts rule -- justice jackson: why isn't the answer with the other side said, which is really make congress a that? in other words, it seems to me their argument is when we are policing this line between what is law and what is policy we should require congress to say the secretary gets to make this decision. and when it doesn't, i guess we look at it as a legal question that the court can decide? general prelogar: i think that

argument would have more merit if there weren't much water under the bridge in the fact that the court explained when or what identify this kind of delegation 40 years ago. petitioners talked about the reliance interest year and tried to diminish them. they did not talk about congress's own interests and reliance on chevron. i think at this juncture, to say that we are going to switch to default and make congress a discretion is in part would be to run to the detriment of congress's own reasonable expectations with respect to drafting. it also does not account for the category of cases where the language congress is using is infused with discretion. they agreed to terms like reasonable, appropriate, necessary. those are terms that require greater application to fill in the details. you cannot interpret them in a vacuum. so i don't understand how this idea of making congress say it could function in that kind of world. and in the final thing is congress has said something very important here, the agency shall

administer the statue with regulations or adjudications that have the force of law. that is part of the statue as -- statute as well. justice jackson: you think that really carries a lot? i've heard you focus on that many times when you are talking about a situation in which deference should be required. general prelogar: exactly, congress in each and every statue where this is going to be applicable or chevron deference will be available is going to have made that judgment in the statute to give the agency that responsibility and role in implementing the statute. justice jackson: let me just ask about whether or not going to the issue of ambiguity, which has come up many times, whether or not the court could clarify when there is a gap or ambiguity that allows for or requires the court to go to step two? and what i am thinking about is what i mentioned about previously with your friend on the other side, which is some scholars have actually identified different kinds of ambiguity. so in one scenario we have a

statute that uses a broad term and that term encompasses a range of reasonable meanings. there are three or four different ways that could be reasonably, you know, the meaning of stationary source, for example. but then there is also the kind of ambiguity in which a statute can mean only one thing, either letter a or b perhaps as of the way language is put forward in the statute. it is just unclear. whether it means a or b. i take the scholars to mean that really in the former scenario is the one in which we have a situation where chevron deference would be required. and could the court say something like that? let me clarify, i look at it as that is reducing to a policy choice. that once we are in the world of finding the kind of ambiguity where there are a number of reasonable alternatives in terms

of making this determination, then it is just going to be a policy choice as to which one congress wanted in some sense or which entity congress wanted to make that decision. general prelogar: so i think there certainly this court could provide more guidance to lower courts. and in particular identify the types of statutory issues and might clearly note discretion. there are going to be some easy calls on this. the types of situations where there might be multiple ways to implement and signal there really is a zone of discretion and the agency should have flexibility. my only concern with going down the road of saying there is some fundamental difference with respect to particular terms that might be subject to only two possible ways to be implemented is that there is kind of an endless number of statues out -- statutes out there and all kinds of varieties. i worry that it might side of certain context where congress actually is comfortable with i

the way of implementing that particular term, even if there are only two possibilities, and did in fact delegate that issue to the agency. so i wouldn't want some kind of bright line rule to diminish the court's ability to recognize and implement that kind of delegation. thank you. justice roberts: thank you, general. mr. martinez, rebuttal? mr. martinez: thank you, your honor. first of all, i think it's really important to be very clear about what chevron does. it takes the power to say that the law or what the law means. to say that the law means x, it takes that power away from courts and gives it to agencies. if then forces agencies -- forces courts to adjudicate the rights of individual litigants that are in front of them based on a version of the law that the courts themselves do not believe is correct. do not believe is the best interpretation. neither congress nor this court can create a doctrine for -- or legislative statute that effectuates that reallocation of interpretive authority. my friend on the other side that the purpose of chevron is to set

the ground rules on how the different branches of government should operate. with respect, i think the constitution says to ground rules and it makes clear the judicial power rest with the courts, not with agencies and certainly not with congress either. i think the apa reinforces that. the solicitor general tries to rescue or reconceptualize chevron by taking issue with our argument that under chevron if the court thinks the best interpretation is x it will have to apply y because the agency told it to. that is exactly what chevron says. it tells the court that it has applied interpretation that the court itself would not choose. interpretation that the court itself does not think is best. the solicitor general also described chevron as applying in the formulation that hurt a lot, it applies that the agency

didn't resolve the question, an innocuous raising. what is really meant by that is chevron applies in cases of ambiguity. and ambiguity has always been understood as a situation where reasonable people can disagree about what the law means. that just broadens the scope of deference. ambiguities are all over the place. courts resolve ambiguities all the time, that is core to the interpretive function. there is no reason to think just because congress is accidentally left an ambiguity in the statute that what it is trying to do is have it resolved by policy decisions made by an agency. justice barrett asked about the justification of chevron rather the intent justification is valid? i took my friend to concede the delegation is fictional. but nonetheless to say we should apply it anyway as a presumption. i do not think you can get the mileage that you need to get out of the intentional delegation theory after you have conceded it is fictional. the only reason it has weight is if it is actually what congress wanted to do.

if congress did not want to delegate it, then we should not be reconceptualizing how we think about statutory interpretation based on a fictional premise. here there is no reason to think congress actually wanted to delegate policymaking authority to agencies to resolve any ambiguity that arises in any statute administered by the agency. i think the government solutions to that problem is to propose a clear statement requirement on congress. hey, you can just legislate more clearly. ambiguities are unintentional. i do not think that works. i think that would impose a massive clarity tax that is unjustified. i think the precedents made very clear they were talking about remedies in those cases that expressly say if we were interpreting these legal issues in a different context where we were not limited by the limits on mandamus remedies we would

apply our best and independent judgment. the solicitor general is looking at text that requires us to interpret statutory provisions. and is saying that rule interprets statutory provisions and is consistent with chevron as she describes as allocating interpretive authority to agencies. the statutes say courts get the interpretation. chevron says agencies get interpretive authority and not court. these are inconsistent. finally with respect to the course correction ideal or the amended approach, i would respectfully suggest that you have tried to mend this and course correct over and over. that is why we have a chevron doctrine. it is overlaid with a lot of bells and whistles. it is very hard to apply in practice. i think in the real world if you try to mend without ending it you are going to put a lot of pressure on the major question doctrine. people will be coming to this court every become to the court every three or four years after you to adopt a new limitation, a new caveat. we would suggest the solution here i