Urias-Orellana v. Bondi

Thank you, Mr. Chief Justice, and may it please the Court: Deciding whether undisputed facts qualify as persecution under the law involves legal interpretation, not fact finding. Even the BIA agrees. It treats the issue as a question of law subject to de novo review. That should not change when a case reaches federal court. Indeed, the government concedes that courts must exercise independent judgment when establishing auxiliary legal principles for use in future cases. They concede that courts owe no deference in interpreting the INA's persecution standard. And courts have repeatedly established auxiliary legal principles on things like sexual violence, religious persecution, economic deprivation, 1 and beyond. But courts did not establish those principles by pondering the term "persecution" in the abstract. They interpreted the law by applying the persecution standard to particular sets of undisputed facts. The government wants you to hold that de novo review applies anytime a case supposedly would make new law in some sense but substantial evidence review when it wouldn't. There's no principled way to draw that line, and this Court has never tried in similar circumstances. For instance, de novo review applies to all fair use determinations made on undisputed facts and all antitrust determinations made on undisputed facts, not just some special subset. The government's contrary rule here would invite further confusion and collateral litigation. Better instead to just require de novo review and let judges get to judging. The courts of appeals are well equipped for the job. Several of them already have experience reviewing de novo the B -- the BIA's 1 persecution determinations, and they generally review the analogous question whether undisputed facts demonstrate torture -- demonstrate torture under the Convention Against Torture. The sky has not fallen. Under both the INA's text and U.S. Bank, courts should decide for themselves whether undisputed facts establish persecution under the law. Section 1252(b) explicitly provides for legal deference on closely related issues but not this one. And decades of experience confirms that courts perform crucial legal work in applying the INA's persecution standard. Deference to the BIA is unwarranted. I welcome the Court's questions.

Because Elias-Zacarias concerned a different element of the asylum eligibility analysis. The -- Elias-Zacarias, the disputed issue there was about the persecutor's subjective motive, their intent. That is a classic pure question of fact. And so Elias-Zacarias was absolutely 1 right to review that factual determination for substantial evidence. The statute before the Court in Elias-Zacarias, which was the predecessor to Section 1252(b)(4), also -- it didn't say the asylum eligibility determination across the board. It applied only to findings of fact. And that's what this statute says as well. So we think the government is just over-reading that decision. And the government doesn't really --

No, Justice Sotomayor. The -- the dispute in that case was about whether the -- as you said, was whether the persecution was done on account of the 1 non-citizen's political opinion or something else. That is a question about what was in --

The --

No, Justice Sotomayor. The -- the facts of the case, what was said, what was done, as Justice Gorsuch pointed out in the earlier argument, subjective intent is something that you infer from facts on the ground. So, in a murder case, you might know exactly what happened, who showed up when, who fired first. But I don't think anyone would say that the facts are undisputed in that murder case if we haven't figured out whether the defendant had the intent to kill or not. And that was the same issue in Elias-Zacarias.

Because that's actually not the question in this case. This is a -- this is a past persecution case. The -- under the INA, a non-citizen is presumptively eligible for asylum if they have suffered persecution in the past. It's not this forward-looking inquiry about whether they have a well-founded fear of persecution going forward. That's a -- that's a different kind of inquiry. And I agree with Your Honor that it goes to -- it goes to their state of mind. It involves the calculation of a future probability. And so the case for a substantial evidence review would be much stronger there. But, here, where -- where -- the facts of what happened to my client are undisputed. The -- the immigration judge took his testimony as credible and true and found that the -- that the death threats he experienced were -- were indeed credible and menacing but nevertheless held that, under the law, they did not rise to the level of persecution.

No, Your Honor. First of all, the BIA doesn't view it that way. If this were all about, you know, determining whose -- whose testimony to believe or not, figuring out what happened, deference would be owed to the immigration judge, and the BIA would do the same. That's not how the BIA views this issue. And the reason for that is credibility is not in dispute here. My client was found credible. Whenever you have a set of undisputed facts, credibility has been resolved one way or the other. So we know exactly what happened to my client, and the sole dispute in this case is what is the legal effect of those events. Did it qualify as -- as "persecution" within the meaning of Section 1101(a)(42), or did it not? That boils down to a legal 1 inquiry.

No, Your Honor, (c) is about eligibility for admission. That's a distinct inquiry for eligibility for asylum.

No, Justice Jackson. I mean, number one, the government has not made this argument. They've conceded that (c) and (d) do not apply. They do not capture this case --

Fair -- fair enough, Justice Jackson.

So I'll -- I'll start with what I agree with in what Your Honor just said, that we do think Section 1252 establishes a reticulated scheme for judicial review that was meant to be comprehensive or -- or close to it. Where I disagree respectfully with Your Honor is that admission somehow encompasses asylum. It does not.

Yes, I do, but I think we've shown that because, again, in (D), in 1252(b)(4)(D), it carves out entitlement to asylum, the second step of the asylum determination, but it doesn't say anything 1 about -- about the first step. And, again, (c) -- and -- and perhaps we just -- we just disagree on this, but admission is a distinct inquiry from -- from asylum.

No, Your Honor, that's -- that's incorrect. You do not need to be admissible to be eligible for asylum. Those are -- those are distinct statuses. My clients have conceded removability, they've conceded they were not admissible, but they are claiming nevertheless that they are eligible for and ultimately entitled to asylum. Those are distinct inquiries, and, again, I understand Your Honor is not the government, but I think that is why the government to their credit did not make this argument because that -- that's well established in this Court's cases.

That -- that's correct. But we think the inference --

Fair enough.

No. We -- we disagree with that, Your Honor, because, again, that -- that sentence you just read from Elias-Zacarias is correct as to what that case was -- was about, this inquiry into causation and to the subjective state of mind of the persecutor and whether the non-citizen feared --

Be --

Because, for example, I'll analogize this case to -- to fair use or antitrust conspiracies, the -- which often go to jury verdicts like it did in -- in Google versus Oracle, in that instance, Your Honor is correct the -- the reviewing court would construe the record in favor of the verdict. But it would not then ask whether any reasonable fact finder could have reached the result rendered by the jury. It would review the determination de novo, and that's because the inquiry involved in those instances, despite being record-intensive, you have to look at, you know, the particular facts of the case, as this Court did at length in Google versus Oracle, for example, it's still a fundamentally legal inquiry or at least a primarily legal inquiry under U.S. Bank. And -- and we know that because, as, again, decades of judicial experience --

Respectfully, Your Honor, no, you do not do that in the fair use context, in the antitrust context, in Icicle Seafoods, where the question was whether the particular activities of a group of workers qualified them as seamen under the Fair Labor Standards Act.

The -- the -- the -- the ultimate question of whether an antitrust conspiracy existed, whether there was anticompetitive conduct, that is ultimately reviewed de novo.

I -- I don't believe so, Your Honor. In Google -- Google versus Oracle, I mean, that one, that's a decision from this Court, held the exact opposite. It said we do not ask whether any reasonable jury could find that there was fair use. We determine -- we're -- we're construing all of the subsidiary facts in favor of the verdict, but as to the ultimate question of whether those facts demonstrate fair use under the law, the Court did not ask whether any reasonable jury could have found fair use. It asked do we, exercising independent judgment, find fair use.

Because I think that's what the First Circuit essentially did in this case. It said we're taking the facts as given and we're reviewing the BIA's interpretation of the law for reasonableness. That's -- and that's the inquiry that we are asserting in this context is incorrect because the underlying question is a primarily legal one 1 that should be reviewed de novo by a reviewing court. That's why the BIA -- the BIA itself under -- I think under that reasoning should be deferring to -- to the IJ on -- on this question. The IJ is the one who sat through the evidence, they've heard the testimony in the record, they're closest to the facts. That's an inquiry that, you know, under U.S. Bank, would normally -- would say, if this is a factual inquiry or -- or a primarily factual inquiry, we should defer to the -- to the fact finder. The BIA does not do that. It recognizes that the inquiry is primarily legal and it warrants exercise of independent judgment by the BIA as the appellate body and the --

So we -- we disagree with that, Your Honor, and I -- I think for -- for two primary reasons. I think the persecution inquiry in itself, persecution, that's not a commonplace word like are two people acting as strangers like in U.S. Bank or are people going to have a hardship like they -- like in Wilkinson. It's -- it's -- it's a term of art, it's the product of a treaty that this country signed.

So I think just one -- one quibble with the -- with the question, Your Honor. We actually do dispute the propriety of -- of that legal rule. It's not current -- it's not presently before the Court. But, when we were stuck --

To an -- to an extent --

Yeah.

We disagree with that, Your Honor, because --

That -- I don't -- that's true to an extent, but I think the same thing is true in the fair use case. I mean, this Court's opinion in Google versus Oracle was looking at the nature of declaring code. It got extremely in the weeds, credit to Justice Breyer for doing that, but it -- it still involved like any mixed question is going to involve a close look at the factual record. But I think what the inquiry in U.S. Bank calls for is asking, number one, is this 1 standard something that people have a -- have an ordinary common understanding of? I think the answer to that is no. And in addition to that, in applying this standard, is the stock judicial job just to say, oh, persecution, let's look at the facts and determine whether it's persecution? That's not what's happened in the lower courts. And the government doesn't dispute this. Lower courts have developed a slew of auxiliary legal principles related to threats, as Your Honor pointed out, but in a whole bunch of different areas, religious persecution, sexual violence. I -- I -- I -- I could go on. And what U.S. Bank says is that when that is the case, you don't necessarily look at any given case that comes before -- before the court and say, look, is this -- is this a boring case that's kind of, you know, run of the mill, or is this a case that raises some really interesting new issue? No, when it -- when the U.S. -- when courts are developing auxiliary legal principles, all decisions about that mixed question are reviewed de novo. We don't look at a fair use case and say, oh, is 1 this a -- is this a boring one or is this a really interesting one presenting novel issues? And I think that's the -- the problem with the government's position. I do want to say one thing, to the government's credit, they've made, I think, a significant concession here that departs from the status quo in a lot of circuits, is that they have at least acknowledged that courts owe no deference to the BIA in interpreting, you know, the law at least in some kind of purely legal sense and in developing these auxiliary legal principles. And I think were the Court to announce that and reaffirm it, you know, with -- with a megaphone, I think that would do a lot of good and it would change how a lot of circuits address these cases. And we think, in our case, it would warrant a remand to assess whether this requirement of --

No, Justice Alito, we don't think that. We think that under this particular statute, when -- when that inquiry looks a lot like an admission eligibility decision or an asylum entitlement decision but isn't one of those like we have here, that's a situation where, yes, the -- the application of law to fact would be de novo.

Because, when courts make that decision, they are building out a decision -- a body of decisional law that guides future cases. If you look at the First 1 Circuit's decision in this case, the IJ's decision in this case, they are analogizing to cases that came before.

No, Your Honor. We -- we would agree that negligence would be something that's reviewed deferentially. But I -- I think there's actually a helpful --

Because I think negligence, number one, the inquiry there is basically what should a person have done given the circumstances. The light was red. Should they have braked or accelerated? People have an ordinary sense of that. I think the negligence example is a helpful one because you can contrast it with the duty-of-care inquiry, which -- and as -- you know, as this Court well 1 knows, duty is traditionally for the judge; negligence is for the jury. And in the duty context, you do have to get immersed into the record to establish whether there was a duty of care. But, if you ask someone on the street, hey, was there a duty of care in this circumstance or that, it's a -- they're not going to be able to answer that as readily. And courts build out, again, a host of legal principles for different kinds of situations in the duty-of-care context. And that's what happens when courts apply the persecution standard. They're taking a term of art that's not commonly understood. They're building out these legal principles that courts can then apply and flesh out. Yes, at this kind of, you know, medium-high level of establishing auxiliary legal principles, as Justice Kagan was explaining, but then also refining what those auxiliary principles actually mean on the ground. What does it mean for a death threat to be sufficiently menacing to qualify as persecution or not? And, again, that is why the decisions in this case, like in so many others, are 1 looking to these prior decisions to figure out what is the mode of analysis that is applicable to these facts. And if you -- if you go to -- many circuits, in fact, to help their staff attorneys and law clerks, have these very detailed outlines explaining prior -- decisions that came before so that it's -- it helps them assist their judges in making correct and efficient decisions that are -- can be squared with the cases that came before them. But the irony in this context is that a lot of those decisions, the ones that are at least denying relief, they're not holding that the BIA's decision was correct as a matter of law. They're just saying that it was reasonable. So you have courts, you have the agency, they're all looking for guidance on this question. They're looking for prior -- they're looking for precedent to -- to -- to use in making their determination. But they don't have a clean set of precedent to do that. They don't have precedent where courts have come in and said this is, in fact, the correct interpretation of -- of persecution as a matter 1 of law. And that is really the problem here. And, again, courts are already doing the -- the kind of inquiry that we're suggesting here under the Convention Against Torture. What -- in most courts of appeals, if the facts are undisputed and the sole remaining question is whether those facts demonstrate mistreatment rising to the level of torture within the meaning of CAT, they review that determination de novo and they're doing just fine. And that's correct because it builds out a body of precedent that courts can use effectively for future cases. And that is the kind of judicial work that, under U.S. Bank, we think would counsel in favor of de novo review. And it's the kind of judicial work that makes clear that this inquiry is not a finding of fact within the meaning of Section 1252(b)(4).

Certainly, Justice Kavanaugh. And, again, we're not here telling you that the -- the BIA's interpretations of these -- of these regulations dictate the meaning of the statute. But what's telling is that the regulation on which the BIA relies uses the exact same words as Section 1252(b)(4). It says findings of fact. And the BIA has said that this exact inquiry is not a finding of fact. And on top of that, the BIA, despite recognizing that immigration judges are in the room, they're hearing the witnesses' testimony, they found the evidence, the BIA is still viewing that as something that it is institutionally better positioned to resolve, just like a court is.

Not explicitly, Your Honor. We've obviously resisted the finding of persecution here. We think that argument --

Our -- our briefs --

We did not make that argument explicitly before the First Circuit. I think it's fairly subsumed in the question of whether do these facts as they are demonstrate persecution. And, again, on remand, we would make that argument.

I -- yes, I think --

I -- I do think that is the government's position, but, again, below we were up against a substantial --

No, Your Honor, because, in the First Circuit, they were -- they cabined their review entirely to -- to deference to the BIA. And, like many other courts of appeals, they have not taken the same position as the government that a question like that --

-- would be reviewed de novo.

Correct.

Not in the first instance, but remand so that we can -- we can advance it. Maybe the government has a waiver argument. Maybe they -- they argue that we didn't waive it because there was First Circuit precedent on point, but we need to go en banc. You know, we'll have that fight on remand.

But it's not something this Court should resolve in the first instance.

Yes.

Well, two points there. I think the question that Your Honor finished with of whether there's, like, a categorical rule that something would qualify as being on account of persecution, I don't actually think that the -- the court deferred to the BIA in that regard. It analyzed that -- that issue -- that legal issue for itself. But the underlying question of what was in the persecutors' heads, why were they doing this, that is a classic finding of fact. 1 And, again, the statute that Elias-Zacarias was analyzing, like this one, did not say --

I think that's right, Your Honor, because, again, the -- what Elias-Zacarias was analyzing, and the opinion says -- makes this clear, is that the statute makes motive central. And the -- what the non-citizen had failed to do in that case was come forward with sufficient evidence to compel the conclusion that the motive of his alleged persecutors was on the basis of political opinion. That was the core -- that was the molten core of the dispute. And the court correctly viewed that as a factual issue going to state of mind and it reviewed that deferentially.

I think there are a few lines that are -- are a little bit more broad, but, again, I think that's exactly why this Court has cautioned multiple times we do not read judicial opinions like statutes. We take them in context. We look at what was the actual question presented.

I think it would be a problem for us.

But, again, we would encourage the Court not to read its opinion like a statute but read the statute here like a statute. And that's where the government has some problems.

I think it's going to be tricky. I mean, again, I do want to commend the government for making that concession because that is not how courts -- many courts of appeals are viewing this issue. They are deferring to the BIA on those kinds of questions. We've cited some cases in our reply brief doing just that. So this Court again takes --

That's correct. That's correct.

Okay.

No, Your Honor. That preexisting practice was not, with respect, as Your Honor described. It's true that some aspects, like motive, as was at issue in Elias-Zacarias, were reviewed deferentially as they should be, as they still are today, but if you look at the actual component of the asylum eligibility inquiry that is at issue in this case, what rises to the level of persecution, that inquiry, the government -- the government has an extended string cite on -- on page -- Footnote 2 of its -- of its brief, and those cases are not, by and large, not addressing 1 that issue. They're addressing other issues. They're just referring to --

Yes. In our reply brief -- forgive me, the -- the -- the page number is escaping me -- but we've cited a couple of cases that go the opposite direction. And the point here is that the government asked you to grant cert in this case because they agreed with us that the courts were so confused about the appropriate standard of review as to this narrow subcomponent of the asylum eligibility determination. I think it's a -- with respect to my friends, it's a little bit rich to say that that confusion that is so entrenched and significant that they supported cert in this case can somehow be resolved on a snapshot consensus at a given point in time. That's just -- that consensus did not exist.

No, Your Honor. Point number one, asylum has actually been kind of like singled out for special treatment by Congress. If you look at Section 1252(a), even the discretionary determination about whether to grant asylum, that's -- judicial review over that determination has been preserved even as Congress stripped away judicial review of most other discretionary determinations. And I -- I think what that tells you is that even as -- even as Congress is -- of course, was limiting judicial review in many respects, it viewed asylum as different and something that needed to be preserved in 1 significant part. And I would also just come back to the fact that we're not vouching for a change here. We're -- we're saying that Congress crystallized in Section 1252(b)(4) what had already been the case, is that a determination about whether undisputed facts legally qualifies as persecution is not a finding of fact. The -- those words, they were in place before IIRIRA. They're in place now. And the plain meaning of those words does not encompass this determination.

Thank you, Mr. Chief Justice, and may it please the Court: Let me just start by picking up on questions about, like, what the inquiry is that the BA -- BIA is making here and what our position is on what the standard is here. So the courts of appeals -- appeals have broadly 1 agreed that "persecution" means really extreme suffering. So you kind of plug that in kind of like the Court did in U.S. Bank. So what the agency is asking here is looking at the evidence that the Petitioner has provided, is this extreme suffering? How much suffering has there been? And that is, one, a question that requires lots of weighing of facts and drawing of inferences. So, in this particular case, just look at the facts here. So, on the one hand, Petitioner's half-brothers were shot because of a dispute between their father and a sicario -- the half-brother's father and a sicario. But then, on the other hand, his mother and his stepsister and his stepfather were never threatened or harmed. The half-brothers appear to have been fine when they moved away from the hometown. Petitioner appears to have been fine when he moved away from the hometown and he lived for years in peace in other parts of El Salvador. Then -- then Petitioner talks about the threats, but, on the other hand, the 1 threats demanded money. So whether you infer from that was it just intimidation, were they even connected with the sicario, I mean, these are all questions of the kind that go to juries. The juries draw inferences. They can cut in different questions. You have to weigh the facts and inferences. What we're saying here is Congress wanted substantial evidence review to apply so that there is deferential review of the weighing of facts and inferences by the agency. And we think it's very clear for several reasons. Let me just do three quick points. One, we think the text and the history point exactly in that direction. So, one, Congress in Section 1158 talked about the question as one that is factual for the trier of fact. Two, this Court in Elias-Zacarias beforehand had said that that was a factual question and applied substantial evidence review. The courts of appeals had all done that. And then the substantial evidence provision at issue here codified language from 1 Elias-Zacarias. So, clearly, Congress was aware of it. And there's no way, no realistic chance that when Congress was overhauling standards of review in IIRIRA and affirmatively overturning what courts of appeals were doing in certain respects and then codifying Elias-Zacarias that it was either unaware of that practice or silently departing from it. It's just not realistically possible. And the second point I'll do even more quickly. This Court's standard approach, even if you thought the statute were less clear, the standard approach that the Court uses across a variety of contexts points in exactly the same way. The work being done here is primarily factual. And just last year in Wilkinson, this Court applied that approach to the INA and said that application of extraordinary hardship is going to be a primarily factual inquiry that is subject to deferential review. And the last thing I'll say is Petitioners' contrary arguments, they really kind of all boil down to the idea that every mixed question in the INA is going to be 1 subject to de novo review. I think this Court both in Guerrero-Lasprilla and -- and Wilkinson said that wasn't going to be the consequence, and -- and I think that would be a sea change and irreconcilable with IIRIRA. I welcome the Court's questions.

So I think, here, the ultimate fact is -- is the one that's at issue, is how much suffering was there? So, for example, I'll point the Court to, like, the decision in TSC Industries about materiality. There are all kinds of situations in which juries get a question like, was the statement material or pain and suffering or negligence? And you have historical facts that -- that could be undisputed, like, here, there were these threats and we know what was said. 1 But then what do you infer from that? How much suffering was there? How do you balance the fact that he was fine when he moved away from his hometown versus the fact that he was threatened several times and the -- the threateners asked for money? And that -- that kind of weighing is one for the BIA. Now, if Petitioner succeeds in saying -- in the argument that no reasonable fact finder, no matter how you weigh the evidence against him, no matter what inferences you draw against him, no reasonable fact finder could have found that there was no persecution, then he would win under substantial evidence --

That -- the substantial -- that question, when you draw all of the inferences --

-- yeah.

And that's what we say should be done.

That's correct, and that is our position.

Yes, that -- that's right.

And -- and just to pick up on Justice Kavanaugh's earlier question about how that would work, I think this is very similar to what courts of appeals do in all kinds of contexts and, in fact, what they are doing under the INA. So think of the -- the example of a district court. There are all kinds of decisions that district courts make that are subject to abuse of discretion review or clear error, and all the time litigants come and they say, okay, so here are my arguments on abuse of discretion and clear error, but 1 setting aside the way that the district court cashed out my particular facts, the district court used the wrong standard, and that is a legal question and you have to correct the legal standard de novo. And the courts of appeals distinguish these things all the time.

No, Your Honor. I think you -- you could look at the way that this Court talked about it in U.S. Bank. Not all mixed questions are alike.

With respect to what? I'm sorry.

So I don't -- I don't have any examples to give you in that respect. I think a lot of what is done is going to be factual in the immigration space. If there are -- you know, if there are legal questions that turn on interpretation of a particular definition like a reticulated definition and the -- the primary work being done is interpreting, looking at a statute, looking at structure, looking at legislative history, and so on, which is not what's at issue here, I mean, then it looks more primarily factual. The other thing I'll say is one of our arguments is that Congress treated this as factual. So, if there were evidence that Congress treated something else as -- as more legal, then that would cut the other way. But, I mean, here, if you look at Section 1158, Congress said --

So several --

So I -- I think what we're saying about 1158 is the thrust of the way Congress thought about the refugee question, is someone a refugee, is that it's -- it's primarily factual. And so the way that it described the way that an applicant sustains the burden -- that's the title of the provision -- to show that they're a refugee was by saying that they have to satisfy the trier of fact not only that their evidence is credible, not only that it's persuasive, but that it's sufficient to show that and then goes on to say: And the way that the trier of fact 1 does that is by weighing the testimony, weighing the other evidence. And that, combined with Congress's codification of Elias-Zacarias taking that same language and plugging it into the substantial evidence review provision, combined with the history before that, I mean, I think it all just makes clear that the way Congress was thinking about it is these are primarily factual inquiries that the IJ is going to make, that we want to have the -- the attorney general take care of. And then --

Yes. I think most -- every court of appeals essentially has said persecution --

No. I think that's -- that's the kind of inquiry that's 1 going to just -- you look at the totality of the circumstances. I think it's kind of a -- a concept that most people can grasp, and you're going to have to in each case just look at all the facts and say, was this extreme enough? How much suffering was there here?

So that's because, like this Court has said several times, when there's a mixed question, it's not just the historical facts. It's what inferences you draw from them and how you weigh them. And that's traditionally the province of the jury, like this Court said in TSC Industries, where the Court said it's not proper to do summary judgment for the materiality question when, even if you take the historical facts, there's going to have to be infer -- inferences drawn and weighing. And the other thing I'll say is that it's not the case that the legal principles go 1 totally undeveloped. I think that's totally belied by experience over decades. There -- there are lots of times when courts of appeals in the course of applying substantial evidence review will either say, one, petitioner wins, so, as a matter of law, no matter how you weigh this, you know, then you have this rule of law, but also, in the course of saying that, they'll provide broad guideposts, like, for example, in general, mere threats are usually not going to be enough or --

That's -- that's right. So that -- that is the -- it's exactly in the same way that you would review a jury verdict.

That's right.

Yeah, I think the Court has said that when the background -- I should say just to back up to sort of make clear what the inquiry is here, in the U.S. Bank, sort of how do we decide which bucket of standard of review to put something in, there are a couple of different important considerations. The lodestar is what kind of work is there, primarily factual or legal. But then there's also you consider the background, what's the tradition been here and whether it's constitutional or not. With those kinds of questions, the ones you're referring to, I think the -- the Court has looked to the history of this is that it developed de novo in -- in, like, a common law sort of way.

Right. It developed that way. And so, if -- if Congress has then talked about it, we presume that, you know, 1 Congress was treating it that way. That's not the case with these kinds of determinations. The tradition cuts exactly the opposite way.

That's right.

So, as Petitioner, I think, admitted, they didn't make that argument about the legal standard. So the only question was applying the First Circuit's own precedent --

Applying the standard that the circuit had already announced, yes.

No, the First Circuit definitely applied the standard. Everybody applied that standard from First Circuit cases, the B -- the IJ, the BIA, and the First Circuit itself.

I -- I think Petitioner agrees that the First Circuit applied that rule, which is substantial 1 evidence, which is Elias-Zacarias.

Yeah. So I -- I think understanding what the regulation actually says, which is different from what my friend --

So if I could give you a --

-- little background, it'll make more sense.

Yes. I think the history is very helpful here. So, before 2002, the BIA was reviewing everything de novo, including facts.

Oh, okay.

So --

So here's where I was going. It's not that it's a question of law. The reg -- what the regulation says is we're going to stop the de novo thing, we're going to do clear error for facts, but we don't want that to mean that judgments and questions of discretion are not subject to de novo review, and it said judgments include decisions about whether the applicant has shown past persecution.

Yeah. Well, my answer is because it's a question of judgment and both the BIA and the IJ share the responsibility to exercise --

Because judgment -- the question about weighing --

So the question about weighing facts and determining whether it's persecution is going to involve judgment. And both the BIA and the IJ have expertise in looking at recurring fact patterns and seeing all kinds of -- of -- of different versions of these cases, far more cases than any court of appeals is ever going to see. And as this Court recognized in Ming Dai and Orlando Ventura, in many cases, the agency has expertise here to apply to -- to deciding those questions. So the AG decided, you know, we'll do clear error, but -- but I'm going to reserve more review for these judgment questions. That doesn't mean that Congress wanted courts of appeals engaging in that kind of judgment in the first instance, and -- and as was already 1 said and admitted, the way that the BIA engages in review, the way that the AG designs internal review has -- has no bearing on that. For example, the BIA reviews questions of discretion de novo. No one thinks the court of appeals should do that. It reviewed facts de novo before 2002. No one thinks the court of appeals should be doing that either.

Yes, I do. I think they were doing that before IIRIRA. I think they're doing that now. I think it's standard practice where, if a petitioner says, look, regardless of the way my facts cash out, the BIA stated the wrong standard and used -- it used a standard that's incorrect. That is a question that courts review de novo.

So -- so there -- 1 there have been questions that have come up. So, like, one that -- that happened earlier was is evidence that there's lots of crime in my country and I'm very likely to be a victim of crime, is that relevant to persecution? And the court said no, that's -- that's not relevant to persecution.

So, like, is it ever relevant, right? Is it ever going to be relevant to persecution? Legal questions, it's not like look at the facts of my case, it's like this is the standard. And there are -- there are other questions like that where questions about whether the standard should be lower in certain circumstances and -- and so on. So the other thing I just want to emphasize is that this Court in Wilkinson really already talked about this some and cited the U.S. Bank sort of formulation and said -- and applied it there and said application of the term "extreme hardship," which is kind of very analogous to the question of extreme 1 suffering, that that is going to be something that is primarily factual and reviewed deferentially. I think the Petitioners' contrary arguments are really saying the thing that this Court was resisting in both Guerrero-Lasprilla and in -- in Wilkinson, which is that every mixed question no matter how factual it is in the INA is going to be subject to de novo review. I mean, that really would be a sea change and is really irreconcilable with IIRIRA. The other thing that I'll note, as, Justice Jackson, you pointed out, I mean, the thrust of IIRIRA was Congress was coming in and then overturning things that courts of appeals were doing. I don't think there's -- there's any realistic chance that you could look at IIRIRA and think that Congress was silently departing from the substantial evidence review consensus for persecution questions and asylum eligibility generally in -- in adopting IIRIRA. Unless there are any further questions?

Thank you, Mr. Chief Justice. Just a few points. The first one is that our position is not that every mixed question in the immigration context would be reviewed de novo. Under the text of the statute, our argument is only that mixed questions that are -- that resemble the determinations at issue in (c) and (d), admission eligibility and entitlement to asylum, would get de novo review because, if the -- the language Congress could have used is right there, they didn't use it. We think that inference is strong as to this particular question, but other things, such as equitable tolling in Guerrero-Lasprilla, other kind of docket 1 management determine -- determinations, for example, those would be reviewed for abuse of discretion. Of course, if the government came in and tried to say that those are findings of fact, we would again disagree, but we wouldn't dispute that they are entitled to deference, just a different kind. And then, if you are looking beyond the text just to U.S. Bank, that -- that analysis is really limited to just this particular mixed question at issue here. And as U.S. Bank, you know, makes clear, the Court has to look at each particular mixed question that comes before it and make the call. So that -- that -- a decision on those grounds would be even narrower. I -- I think it's telling that the government couldn't come up in response to I believe it was Justice Kagan asked what about -- would any mixed question in this context get de novo review, and they couldn't name one. I think that's because, if this one doesn't get it, it's hard to think of an example. Persecution, again, it is a term of 1 art. It's not something that ordinary people understand right off the bat. It's the product of -- of -- of a major international treaty, and we have decades of experience where courts are coming up with the exact kind of auxiliary legal principles that U.S. Bank talked about as a -- as a -- as a marker of a mixed question that deserves de novo review, even though it can -- yes, some aspects of the decision can be -- can be fact-intensive. They involve a close look at the record, but courts are also doing what courts do when they analyze this mixed question and that's what separates this particular mixed question from others. Second, Justice Kavanaugh, to your -- your question about the BI -- the BIA regulations, the government is trying to couch this determination as a question of judgment versus law. They don't cite anything for that. We have BIA decisions calling it a question of law. But it really doesn't matter because the fundamental point is that the BIA agrees that it's not a finding of fact, which is the exact language we have in this statute. And the BIA agrees that the IJ's 1 institutional advantages of being closer to the evidence don't justify deference to the IJ on this particular question. So this debate about judgment versus question of law is just -- is just immaterial. Fourth, on Elias-Zacarias and the history, Justice Jackson, your question, if you look at page 12 of our reply in Note 2, we cite examples of cases analyzing this exact mixed question that predate IIRIRA that were reviewed de novo. The case law as it is now is a muddle, so I'm not going to pretend that there is some consensus going in our favor, but we don't need to show that. The government is trying to tell you that there was a consensus that was somehow overruled by Congress. That's not the case. And on -- on Elias-Zacarias more broadly, we would just reiterate that the actual dispute in that case was about motive, it was about a subjective state of mind, and the decision needs to be read in that context. And if it's read as the government says it is, number one, the government concession doesn't make sense because 1 Elias-Zacarias, if you take a couple lines at face value, says the entire asylum eligibility determination must be reviewed deferentially, but they're admitting that even legal questions or even some amount of mixed questions about these auxiliary -- auxiliary legal principles are reviewed de novo. So that just can't be right. And we think trying to draw that line about where do interesting cases that are creating auxiliary legal principles, where do those end and where do the fact-intensive ones start, you just can't draw that principled line and this Court has never tried to do that when it's analyzed other mixed questions like in -- like in U.S. Bank, like in Google versus Oracle and other cases. But, finally, even if the -- if -- if the Court does try -- try to draw that line and accept the government's position, it's not the case that courts of appeals are doing what the government says is correct. Some are, but some are not. If you look at page 3 of our reply brief, we cite a Tenth Circuit decision and a Fourth Circuit decision that said, look, we 1 knowledge that the BIA has created this legal rule, but the Tenth Circuit said we can't disagree with that legal rule unless, you know, we could overcome the substantial evidence review. And the Fourth Circuit dis- -- ultimately disagreed with the BIA but only because it found that this requirement of showing significant actual harm was manifestly contrary to law. So it too was deferring on what even the government admits is a legal question that should be reviewed de novo. So, if the Court -- even if it sides with the government here, we believe that is a -- that is an important change. It's clarified the standard of -- of review. That will matter going forward, and it should matter in this case, and we should be entitled to vacatur of the judgment below at least on those narrow grounds.