United States Court of Appeals for the Fifth Circuit United States Court of Appeals ____________ Fifth Circuit FILED No. 21-30754 December 18, 2025 ____________ Lyle W. Cayce Clerk Richard Hershey, Plaintiff—Appellant, versus City of Bossier City; Bobby Gilbert, Individually and in his Capacity as Deputy Marshal; Daniel Stoll; David Smith; Tyshon Harvey; Eugene Tucker, Defendants—Appellees. ______________________________ Appeal from the United States District Court for the Western District of Louisiana USDC No. 5:21-CV-460 ______________________________

ON PETITION FOR REHEARING EN BANC

Before Dennis, Richman, and Ho, Circuit Judges.

Per Curiam: The petition for rehearing en banc is DENIED. At the request of one of its members, the court was polled, and a majority did not vote in favor of rehearing (Fed. R. App. P. 40 and 5th Cir. R. 40).

No. 21-30754

In the en banc poll, seven judges voted in favor of rehearing (Judges Jones, Smith, Richman, Duncan, Engelhardt, Oldham, and Wilson), and ten judges voted against rehearing (Chief Judge Elrod, and Judges Stewart, Southwick, Haynes, Graves, Higginson, Willett, Ho, Douglas, and Ramirez).

No. 21-30754

James C. Ho, Circuit Judge, concurring in the denial of rehearing en banc: Popular speech doesn’t need protection. It’s only when speech is unpopular that you need the First Amendment.

That’s why the devout look to the judiciary for protection. Religious speech is often unpopular speech, as people of faith have known for thousands of years. See, e.g., John 15:18–19 (“If the world hates you, you know that it has hated Me before it hated you. If you were of the world, the world would love you as its own; but because you are not of the world, but I chose you out of the world, because of this the world hates you.”).

It should go without saying, then, that the freedom of speech secured by the First Amendment includes religious speech. It protects not only the right to pray, but to preach—not just to worship, but to witness—to exercise your faith by evangelizing your faith. Our Founders secured these rights even if—indeed, especially if—the government doesn’t want you to exercise them.

So it should be obvious that believers have the right to share the good news with others. And the obviousness of that right should have been enough to defeat qualified immunity in this case, without the need for a factually identical case saying so. Our now-Chief Judge made this point in Morgan v. Swanson, 659 F.3d 359, 414 n.30 (5th Cir. 2011) (Elrod, J., dissenting in part). But a majority of our en banc court refused that view in Morgan. And a majority of our en banc court affirmatively rejected it in Villarreal v. City of Laredo, 94 F.4th 374, 395 (5th Cir. 2024). I detailed all of this in my dissent in Villarreal, 94 F.4th at 413–14, and again in McMurry v. Weaver, 142 F.4th 292, 304–07 (5th Cir. 2025) (Ho, J., concurring), and Hershey v. City of Bossier City, 156 F.4th 555, 557, 558–60 (5th Cir. 2025) (Ho, J., concurring). But I’m duty-bound to follow our en banc precedents, whether I agree with them or not.

No. 21-30754

Governing precedent does not, however, foreclose municipal liability under Monell v. New York City Dep’t of Soc. Servs., 436 U.S. 658 (1978). To the contrary, the Supreme Court has unanimously held that “the need to train officers . . . can be said to be ‘so obvious,’ that failure to do so could properly be characterized as ‘deliberate indifference’ to constitutional rights.” City of Canton v. Harris, 489 U.S. 378, 390 n.10 (1989). 1 “[I]n the absence of training, there is no way for novice officers to obtain the legal knowledge they require. Under those circumstances there is an obvious need for some form of training.” Connick v. Thompson, 563 U.S. 51, 64 (2011).

The petition for rehearing en banc asks us to shield obvious violations of religious liberty under Monell. What’s more, it would allow municipalities to trample on religious liberty simply by deputizing private actors to do their dirty work. But see NRA v. Vullo, 602 U.S. 175, 198 (2024) (“the First Amendment prohibits government officials from wielding their power selectively to punish or suppress speech, directly or (as alleged here) through private intermediaries”); Marsh v. Alabama, 326 U.S. 501 (1946); Murthy v. Missouri, 603 U.S. 43, 79–80, 99 (2024) (Alito, J., dissenting).

That’s wrong, and I’m glad we’re not going along with it. I concur in the denial of the petition for rehearing en banc. *** My dissenting colleagues would grant the petition for rehearing en banc. In doing so, they claim that I’m wrong on three fronts. I’m wrong about our court’s qualified immunity jurisprudence. I’m wrong about our _____________________ Harris refers to “fleeing felons,” but nothing in § 1983 suggests that courts should favor the rights of criminals over the rights of law-abiding citizens. We made that mistake as to qualified immunity, see Villarreal v. City of Laredo, 94 F.4th 374, 413 (5th Cir. 2024) (Ho, J., dissenting), and I’m grateful that we won’t be repeating that mistake here as to municipal liability.

No. 21-30754

approach to municipal liability under Monell. And most importantly, I’m wrong that there’s a sincere concern about religious liberty presented here.

I’ll begin by responding to the two procedural justifications for shielding even obvious First Amendment violations from judicial review, before addressing my colleagues’ curious skepticism about the legitimate religious liberty concerns implicated in this case (which should be resolved in the first instance at trial in any event).

In doing so, I note that it’s become a regrettable but unfortunately common practice of judges to invent reasons to avoid addressing sensitive matters of conscience on the merits—whether it’s by concocting procedural problems or distorting the facts. I’m not the only one to notice this tactic.2 But I note it here because this case is Exhibit A in the use of these stratagems.

I.

Let’s start with qualified immunity. For any citizen who seeks damages from a public official for violating their rights, it’s not enough that their rights have been violated. Their rights must also be “clearly established” at the time of the violation. One way to make this showing is to _____________________ See, e.g., Christian Legal Society v. Martinez, 561 U.S. 661, 707–18 (2010) (Alito, J., dissenting) (detailing the numerous ways in which “[t]he Court provides a misleading portrayal of this [religious liberty] case”); Parents Protecting Our Children v. Eau Claire Area Sch. Dist., 145 S. Ct. 14, 14–15 (2024) (Alito, J., dissenting from the denial of certiorari) (“I am concerned that some federal courts are succumbing to the temptation to use the doctrine of Article III standing as a way of avoiding some particularly contentious constitutional questions.”); see also Tucker v. Gaddis, 40 F.4th 289, 293–97 (5th Cir. 2022) (Ho, J., concurring) (collecting examples from other circuits about misuse of procedural doctrines to avoid deciding religious liberty claims); U.S. Navy SEALs 1-26 v. Biden, 72 F.4th 666, 677–78 (5th Cir. 2023) (Ho, J., dissenting) (misuse of mootness to avoid deciding religious liberty challenge to vaccine mandate); Neese v. Becerra, 127 F.4th 601, 603–06 & n.1 (5th Cir. 2025) (Ho, J., dissenting from the denial of rehearing en banc) (misuse of standing and mischaracterization of facts to avoid addressing conscience objections to gender ideology mandate).

No. 21-30754

identify binding precedent with sufficiently similar facts that the violation should have been clear to the public official at the time of the incident.

But that’s not supposed to be the only way to do it. Under Hope v. Pelzer, 536 U.S. 730 (2002), and Taylor v. Riojas, 592 U.S. 7 (2020), a plaintiff can also overcome qualified immunity if the violation is so egregious that it should’ve been obvious to the official, without the need for a materially similar case. See Hope, 536 U.S. at 741 (“general statements of the law . . . may apply with obvious clarity to the specific conduct in question, even though the very action in question has not previously been held unlawful”) (cleaned up) (quoting United States v. Lanier, 520 U.S. 259, 270–71 (1997), and Anderson v. Creighton, 483 U.S. 635, 640 (1987)); Taylor, 592 U.S. at 8–9 & n.2 (summarily reversing our court’s grant of qualified immunity due to the “obviousness” of the constitutional violation) (citing Hope and Lanier).

Unfortunately, we didn’t follow this principle in Villarreal v. City of Laredo, 94 F.4th 374 (5th Cir. 2024). The en banc majority held that Hope and Taylor are Eighth Amendment cases that do not apply in other contexts, such as the First Amendment. Id. at 395. I dissented, noting the oddity of treating claims from incarcerated criminals more favorably than law-abiding citizens. Id. at 413–14 (Ho, J., dissenting). 3 Naturally, I wish my views had prevailed in Villarreal. But they didn’t, and I’m bound to acknowledge that fact in all intellectual honesty.

What I could not have fathomed is what my colleagues are saying about Villarreal today. Having won in Villarreal, they now dispute what it is _____________________ The Supreme Court vacated our en banc decision in Villarreal. 145 S. Ct. 368 (2024). But on remand, our court reinstated its earlier en banc decision. See Villarreal v. City of Laredo, 134 F.4th 273, 276 (5th Cir. 2025) (“[o]ur previous en banc majority opinion is superseded only to th[e] extent” necessary to respond to the Supreme Court’s vacatur regarding the substantive requirements of a First Amendment retaliation claim).

No. 21-30754

that they won. They deny that they held in Villarreal that “the obviousness exception does not apply in First Amendment cases.” Post, at 29.

So let’s just roll the tape. Here’s verbatim what the Villarreal en banc majority said about qualified immunity, Hope, Taylor, and obvious violations: Villarreal cites no case, nor are we aware of one, where the Supreme Court, or any other court, has held that it is unconstitutional to arrest a person, even a journalist, upon probable cause for violating a statute that prohibits solicitation and receipt of nonpublic information from the government for personal benefit. Under the normal standards of qualified immunity, no “clearly established law” placed the officers on notice of Villarreal’s First Amendment right not to be arrested.

Villarreal, however, relies on Eighth Amendment cases where the Supreme Court denied qualified immunity for deliberate indifference to unconstitutional prison conditions and declined to scrutinize the cases fact-specifically. See Hope v. Pelzer, 536 U.S. 730, 738–39 (2002) (“[T]he risk of harm [to the prisoners] is obvious.”); Taylor v. Riojas, 592 U.S. 7, 9 (2020) (per curiam) (“Confronted with the particularly egregious facts of this case, any reasonable officer should have realized that Taylor’s conditions of confinement offended the Constitution.”) (footnote omitted)); McCoy v. Alamu, 141 S.

Ct. 1364 (2021) (instructing the court to reconsider an Eighth Amendment case “in light of Taylor”).

Hope and its progeny express a general, but decidedly narrow, obviousness exception to the requirement that “clearly established law” be founded on materially identical facts. In any event, those cases are inappropriate templates for describing “clearly established” law in this context. In Morgan v. Swanson, 659 F.3d 359, 373 (5th Cir. 2011) (en banc), a case involving First Amendment free exercise rights, this court noted that Hope does not stand for the broad proposition that plaintiffs need not offer any similar cases to prove that an

No. 21-30754

officer should have been on notice that his conduct violated the Constitution. . . .

Consequently, we adhere to the general rule that for an asserted right to be clearly established for purposes of qualified immunity, it must “have a sufficiently clear foundation in then- existing precedent” that it is “settled law.” Wesby, 583 U.S. at (citation omitted). “The precedent must be clear enough that every reasonable official would interpret it to establish the particular rule the plaintiff seeks to apply.” Id. (emphasis added). The law is not clearly established if referenced cases are “materially distinguishable and thus do[] not govern the facts of this case.” Rivas-Villegas, 595 U.S. at 6. 94 F.4th at 395 (cleaned up).

It’s hard to misread this passage. Villarreal reads Hope and Taylor to apply only to “Eighth Amendment cases.” Id. Hope and Taylor don’t apply in cases “involving First Amendment free exercise rights.” Id. And the court justifies this result by citing our earlier en banc decision in Morgan. Id. Now I’ll quote verbatim what I said in response in my dissent: The Supreme Court has made clear that public officials who commit obvious constitutional violations are not entitled to qualified immunity. In fact, the Court has repeatedly reversed circuits, including ours, for granting qualified immunity for obvious violations of constitutional rights. See, e.g., Hope v. Pelzer, 536 U.S. 730, 741 (2002); Taylor v. Riojas, 592 U.S. 7, 9 (2020).

The majority responds that the standard articulated in Hope and Taylor doesn’t apply here, because those cases arose under the Eighth Amendment, not the First Amendment.

Ante, at 395.

But that would treat the First Amendment as a second- class right. Nothing in § 1983 suggests that courts should favor the Eighth Amendment rights of convicted criminals over the

No. 21-30754

First Amendment rights of law-abiding citizens. Nothing in Hope or Taylor indicates that those decisions apply only to prison conditions. And no other circuit takes the approach urged by our colleagues in the majority. To the contrary, nine circuits have indicated that the standards articulated in Hope apply specifically in the First Amendment context. See, e.g., Diaz-Bigio v. Santini, 652 F.3d 45, 50 (1st Cir. 2011); Nagle v. Marron, 663 F.3d 100, 115–116 (2nd Cir. 2011); McGreevy v. Stroup, 413 F.3d 359, 366 (3rd Cir. 2005); Tobey v. Jones, 706 F.3d 379, 391 n.6 (4th Cir. 2013); MacIntosh v. Clous, 69 F.4th 309, 399 (6th Cir. 2023); Kristofek v. Vill. of Orland Hills, 832 F.3d 785, 798 (7th Cir. 2016); Galvin v. Hay, 374 F.3d 739, 746– (9th Cir. 2004); Frasier v. Evans, 992 F.3d 1003, 1021–22 (10th Cir. 2021); Leslie v. Hancock Cnty. Bd. of Educ., 720 F.3d 1338, 1345–46 (11th Cir. 2013). See also Cheeks v. Belmar, 80 F.4th 872, 877 (8th Cir. 2023) (applying Hope to the Fourteenth Amendment); Atherton v. Dist. of Columbia Off. of the Mayor, 706 F.3d 512, 515 (D.C. Cir. 2013) (applying Hope to the Fifth Amendment).

So I would apply Hope and Taylor in the First Amendment context. See also Morgan v. Swanson, 659 F.3d 359, 412, 414 n.30 (5th Cir. 2011) (en banc) (Elrod, J., dissenting in part) (concluding that Hope applies to obvious First Amendment violations). Id. at 413–14 (Ho, J., dissenting) (cleaned up).

If my dissent mischaracterized the majority’s ruling, surely they would’ve taken me to task. But they didn’t. Because I didn’t. (Put it another way: If our colleagues agreed with my dissent that Hope and Taylor apply to the First Amendment, why did they vote against Priscilla Villarreal?)

People can debate, of course, who’s right about qualified immunity, Hope, and Taylor—the en banc majority or my dissent. But we should be candid about what our opinions do and don’t say.

No. 21-30754

Numerous religious liberty organizations and other public interest groups have filed amicus briefs criticizing Villarreal for refusing to apply the obviousness exception to the First Amendment.

Start with the Alliance Defending Freedom and the Dhillon Law Group, and their amicus brief on behalf of the Young America’s Foundation and the Manhattan Institute. Here’s what they said: Some applications of laws are so “obvious[ly]” unconstitutional, Rivas-Villegas, 595 U.S. at 6, or “egregious,” Taylor, 592 U.S. at 9, that qualified immunity dissolves without a factually analogous case on the books. Often this principle is associated with Hope v. Pelzer, 536 U.S. at 741, where this Court said “a general constitutional rule already identified in the decisional law may apply with obvious clarity to the specific conduct in question.” Accord, e.g., Anderson, 483 U.S. at 640 (rejecting the notion “that an official action is protected by qualified immunity unless the very action in question has previously been held unlawful”).

The Fifth Circuit en banc majority avoided this conclusion by throwing out the obviousness exception to “normal” qualified-immunity rules in the free-speech and free-exercise context. Villarreal, 94 F.4th at 395. It refused to accept that “the ‘obvious’ violation exception applies broadly to arrests that may impinge on First Amendment rights.” Id. at 392. Obviousness, the majority said, is “no more than a possible exception,” ibid., or one narrowly confined to “Eighth Amendment cases,” id. at 395.

This dispelling of the obviousness exception is mistaken. Eighth Amendment precedent has its oddities but qualified immunity isn’t one of them. Nothing in Hope or Taylor suggests that the exception applies only to claims of cruel and unusual punishment. Nor does confining the obviousness exception to that narrow context make sense.

Freedoms of speech, press, and religion are among our

No. 21-30754

proudest liberties, not “second-class right[s].” Villarreal, 94 F.4th at 413 (Ho, J., dissenting).

Brief for Young America’s Foundation and Manhattan Institute as Amici Curiae Supporting Petitioner, Villarreal v. Alaniz, 145 S. Ct. 368 (2024) (No. 23-1155), 2024 WL 2786483, *9–10.

According to ADF, Dhillon, and company, Villarreal is “dangerous” for religious liberty: “Left undisturbed, the Fifth Circuit’s ruling provides dangerous license for government actors to flagrantly violate the Constitution without recourse, even against the most established rights, simply because they invoke a novel factual situation never before specifically addressed by the courts.” Id. at *12.

Consider also what the First Liberty Institute wrote: The approach taken by the Fifth Circuit towards qualified immunity is exactly the kind of approach this Court castigated in Hope and its progeny. In Hope, this Court indicated that public officials who commit obvious constitutional violations are not entitled to qualified immunity.

See 536 U.S. at 740–42, 745–46. Although this principle was first articulated in the Eighth Amendment context, it has been extended to cases involving the First Amendment, by [the Supreme] Court and nine circuits.”

Brief of First Liberty Institute as Amicus Curiae in Support of Petitioner, Villarreal v. Alaniz, 145 S. Ct. 368 (2024) (No. 23-1155), 2024 WL 2058693, *10.

The First Liberty Institute concluded that Villarreal is an “insidious” ruling for religious liberty: “Whatever value qualified immunity has, it cannot be predicated on the distinctly un-American notion that our freedoms are only cognizable in the light of judicial pronouncements cast down from on high.” Id. at *11.

No. 21-30754

These are the views of the Nation’s most respected religious liberty law firms. And they aren’t alone. A number of other organizations have also filed Supreme Court amicus briefs protesting our court’s elimination of the obviousness exception in First Amendment cases. See Docket, Villarreal v. Alaniz, No. 23-1155 (U.S.); Docket, Villarreal v. Alaniz, No. 25-29 (U.S.); see also McMurry, 142 F.4th at 305–06 (Ho, J., concurring) (citing amici).

So they may be surprised to see what our colleagues have to say today. *** One possible explanation for their dissent is that our colleagues regret what they did in Villarreal (and before that in Morgan). But if that is so, I just wish they had expressed this regret to the rest of us earlier, when we could’ve done something about it. Because it’s not what they’ve ever said before today. It’s not what our dissenting colleague said in her opinion at the panel stage. And it’s certainly not what Defendants have urged in their petition for rehearing en banc. They said none of this until today’s dissent.

And notably, Hershey himself did not file an en banc rehearing petition asking us to overturn Villarreal—presumably because he and his counsel know how to read opinions and count votes.

But make no mistake: Now that everyone’s on record, I’d welcome an en banc petition from Hershey asking us to overturn the mistaken principles of Morgan and Villarreal, deny qualified immunity, and grant Hershey the opportunity to go to trial. Because it sounds like there are now the votes to do it.

That said, Hershey may prefer to preserve his Monell win, and leave it to someone else to invoke today’s surprising developments to try to overturn Villarreal. And that leads me to my final point on Villarreal. I’m beyond delighted by this surprising turn of events, to be sure. But it bears noting that

No. 21-30754

this is precisely the kind of “surprise switcheroo” that members of our court have criticized federal agencies for doing. Wages and White Lion Investments, L.L.C. v. FDA, 90 F.4th 357, 386 (5th Cir. 2024), rev’d, 604 U.S. 542 (2025).

I’d hold us to the same standard that we apply to others. As imperfect human beings, we’re allowed to change our mind, of course. But we should be candid, not caustic, when we do so. “When an agency changes its existing position, it . . . must at least display awareness that it is changing position.”

Id. at 381 (quotations omitted). At least one influential scholar has suggested the same. See Josh Blackman, A Fifth Circuit Disgrantle in the Alien Enemies Act Case, Reason.com, Oct. 1, 2025 (comparing the dueling en banc opinions in United States v. Abbott, 110 F.4th 700 (5th Cir. 2024), with later developments). 4 Finally, the dissent asks a question that deserves a fulsome response: “Why say ‘We are bound by this terrible opinion,’” but “oppose rehearing to fix the purportedly terrible opinion?” Post, at 32. “Wouldn’t it be easier to do the right thing and vote for rehearing now?” Id. My two responses to their question/offer: It’s deeply disingenuous.

But I happily accept.

First, it’s disingenuous. Defendants’ petition for rehearing en banc (obviously) doesn’t ask the court to reverse its qualified immunity win by _____________________ My colleagues deny the surprise switcheroo, and insist that we must read Villarreal “in the context in which it was decided.” Post, at 31. In their view, Villarreal wasn’t about refusing Hope and Taylor at all—it was about the “concededly untainted intermediary and the valid arrest warrant.” Post, at 30. But that’s wrong on both fronts.

It’s wrong because it can’t be squared with the verbatim language from Villarreal that I quoted earlier. And it’s wrong because the intermediary and warrant issues weren’t “conceded” at all—to the contrary, multiple dissents in Villarreal discussed in detail the tainted intermediary and invalid warrant. See, e.g., 94 F.4th at 403–04 (Higginson, J., dissenting); id. at 418–19 (Ho, J., dissenting).

No. 21-30754

revisiting Villarreal. Their petition is solely about Monell. The en banc poll is solely about Monell. And the poll closed long ago in any event—well before any of them ever told me (or anyone else, from what I can tell) that they suddenly share my opposition to Villarreal.

So they can theorize all they want that I somehow secretly like what happened in Villarreal. But personally, I don’t think it passes the laugh test.

See, e.g., Villarreal, 94 F.4th at 409 (Ho, J., dissenting) (opposing the en banc decision in Villarreal); McMurry, 142 F.4th at 302 (Ho, J., concurring) (opposing the en banc decision in Villarreal); Hershey, 156 F.4th at 557 (Ho, J., concurring) (opposing the en banc decision in Villarreal); see also Oliver v. Arnold, 19 F.4th 843, 843 (5th Cir. 2021) (Ho, J., concurring in the denial of rehearing en banc) (supporting relief in Villarreal); Mayfield v. Butler Snow, 78 F.4th 796, 797 (5th Cir. 2023) (Ho, J., dissenting from the denial of rehearing en banc) (supporting relief in Villarreal); Gonzalez v. Trevino, 60 F.4th 906, 907 (5th Cir. 2023) (Ho, J., dissenting from the denial of rehearing en banc) (supporting relief in Villarreal). I regret the increasing lack of candor that has come to pervade our court’s en banc proceedings of late.

But no matter. I vote for rehearing.

I’m happy to let bygones be bygones, and ignore from here on out what my colleagues may or may not have said before, in either Morgan or Villarreal.

I’m delighted to go back to what we said in my original panel majority opinion in Villarreal: Of course the obviousness exception of Hope and Taylor applies to the First Amendment.

So as a member of the panel, I vote for the panel to rehear this case.

After today’s order is out, I’m sure that the panel will deny qualified immunity in this case. And that will be that.

No. 21-30754

II.

Next, municipal liability. It’s established law that, to obtain liability under Monell, an extensive pattern of past violations by municipal employees is sufficient—but it’s not necessary. A single incident can also trigger liability, but only if there’s an obvious risk of a constitutional violation if the municipality provides “no training whatsoever.”

Our dissenting colleagues insist that that’s wrong. Indeed, they claim that it’s “a sea change from our court’s prior approach” to contend that “a political subdivision is liable under Monell when it provides ‘no training whatsoever’ on the First Amendment.” Post, at 27-28 (emphasis added).

But virtually every member of the dissent has previously agreed that “a political subdivision is liable under Monell when it provides ‘no training whatsoever’”—at least in other constitutional contexts. Id. See, e.g., Henderson v. Harris Cnty, 51 F.4th 125, 131 (5th Cir. 2022) (Monell liability available under the Fourth Amendment where “the government actor was provided no training whatsoever”) (quotations omitted); York v. Welch, 2024 WL 775179, *5 (5th Cir.) (again, Monell liability available under the Fourth Amendment where “the government actor was provided no training whatsoever”) (quotations omitted); Covington v. City of Madisonville, 812 F.

App’x 219, 225 (5th Cir. 2020) (Monell liability available where “it should have been apparent” that the complete “failure to train” would result in a Fourteenth Amendment violation) (quotations omitted); Peterson v. City of Fort Worth, 588 F.3d 838, 849 (5th Cir. 2009) (Monell liability available under the Fourth Amendment based on “a complete failure to train”).

None of these cases suggest a different rule for the First Amendment.

To the contrary, a municipality is liable if it provides “‘no training whatsoever’ with respect to the relevant constitutional duty.” Littell v. Houston Indep. Sch. Dist., 894 F.3d 616, 625 n.5 (5th Cir. 2018) (emphasis added).

No. 21-30754

So I don’t see a “sea change”—I see equal respect for religious liberty when we apply Monell liability to a municipality that “provides ‘no training whatsoever’ on the First Amendment.” Post, at 28 (emphasis added).

The dissenters wonder what might happen if we apply our established law to the First Amendment. They worry, for example, that federal judges might start forcing cities to train all of their employees on “the intricacies of commercial speech.” Post, at 28. Rest assured that no one is saying that.

Nor is anyone ordering municipal waste workers to be taught about Obergefell, or insisting that firefighters be instructed on tax law.

We’re only saying what the law already says—and what the dissenters once agreed: a municipality is liable if it provides “‘no training whatsoever’ with respect to the relevant constitutional duty.” Littell, 894 F.3d at 625 n.5 (emphasis added). If the training isn’t relevant, then it isn’t required.

And that gets us to what I believe is the heart of the dispute. The dissent seems to think that the problem with this case is that the “risk of a constitutional violation was infinitesimal.” Post, at 26.

I have a different view. Street preaching is not some infinitesimally rare and obscure practice. Nor is it one that municipalities never bother with.

People of faith have long sought opportunities to spread the gospel in the public square. Nearly a century ago, the Supreme Court observed that “[t]he hand distribution of religious tracts is an age-old form of missionary evangelism—as old as the history of printing presses.” Murdock v. Pennsylvania, 319 U.S. 105, 108 (1943). “It has been a potent force in various religious movements down through the years,” as people of faith “carry the Gospel to thousands upon thousands of homes and seek through personal visitations to win adherents.” Id. at 108–9. So anyone who is “rightfully on a street which the state has left open to the public carries with him there as elsewhere the constitutional right to express his views in an orderly fashion.”

No. 21-30754

Jamison v. Texas, 318 U.S. 413, 416 (1943). “This right extends to the communication of ideas by handbills and literature as well as by the spoken word.” Id. It’s a religious practice that dates back thousands of years. The Bible speaks of the Great Commission, in which Jesus calls on the disciples to go forth and spread the word to others in every nation. Matthew 28:16–20. “[O]pen-air preaching was a preferred method of Jonah, Jeremiah, the apostle Paul, and many others.” Ryan Denton, Into the Highways and Hedges: A Primer for Open-Air Preaching, Desiring God, Feb. 4, 2024, available at www.desiringgod.org/articles/into-the-highways-and-hedges. “They went to the people and preached God’s word.” Id. After all, “Jesus went into the boat or up on the mountainside to preach the good news.” Id. My dissenting colleagues question why I invoke “thousands of years” of religious tradition. Post, at 24. But they invoked millennia of legal tradition in Environment Texas Citizen Lobby, Inc. v. ExxonMobil Corp., 123 F.4th 309 (5th Cir. 2024)—even where it didn’t seem relevant. 5 So if we’re going to consider millennia of legal tradition since at least the time of Roman rule, then we ought to consider millennia of religious tradition since Roman times as well—particularly where it’s actually relevant to the issues presented.

The Harvard Religious Freedom Clinic certainly thinks it relevant.

Just this year, they briefed the U.S. Supreme Court that street preaching is an “ancient tradition” with “deep historical roots.” Brief of Stephen Nylen et al. as Amici Curiae Supporting Petitioner, Olivier v. City of Brandon, No. _____________________ In that case, I questioned the relevance of Roman law to the established practice of American courts on dismissing appeals as improvidently granted. I’m grateful that my colleagues stepped away from the point shortly thereafter. See, e.g., Silverthorne Seismic v. Sterling Seismic Srvs., 125 F.4th 593, 598 & n.5 (5th Cir. 2025) (discussing practice of dismissing discretionary appeals as improvidently granted).

No. 21-30754

24-993 (U.S.), 2025 WL 2676057, *4. I agree that it’s relevant in cases like this that the devout have sought to spread the gospel to others as an essential part of their faith for millennia.

The most effective way to spread the word to others, of course, is to go where lots of people congregate. “The First Amendment protects the right of every citizen to reach the minds of willing listeners and to do so there must be opportunity to win their attention.” Heffron v. Int’l Soc. for Krishna Consciousness, Inc., 452 U.S. 640, 655 (1981) (cleaned up, emphasis added).

Concert venues are an ideal forum for the street preacher. As one of the most famous street preachers has observed, “music venues” are a “good open-air setting” for street preaching because it’s a “public location where people gather where they’re not in a hurry and can take time to listen.” Ray Comfort, Fifty Years Of Open-Air Preaching 50 (2024). See also Denton, Into the Highways (“open-air preaching is especially useful . . . outside a sporting event”).

Municipalities are not only well aware of this—a number of them don’t like it, and endeavor to stop it.

In fact, municipalities are so actively opposed to street preaching that our court has had to adjudicate constitutional objections to their restrictive policies in case after case. See, e.g., Siders v. City of Brandon, 123 F.4th 293 (5th Cir. 2024) (Christian advocate prevented from evangelizing outside public amphitheater); Herridge v. Montgomery Cnty., No. 21-20264, 2022 WL 989421 (5th Cir. Apr. 1, 2022) (Christian advocate prevented from distributing pamphlets near outdoor pavilion in Houston); Olivier v. City of Brandon, 2023 WL 5500223 (5th Cir.) (street preacher prevented from evangelizing outside public amphitheater), cert. granted, 145 S. Ct. 2871 (2025); Denton v. City of El Paso, 861 F. App’x 836 (5th Cir. 2021) (Christian advocate prevented from evangelizing at an outdoor farmer’s market); Roy v.

No. 21-30754

City of Monroe, 950 F.3d 245 (5th Cir. 2020) (street preacher prevented from evangelizing on the sidewalk outside of a local bar); Allen v. Cisneros, 815 F.3d 239 (5th Cir. 2016) (street preacher prevented from evangelizing on the streets of Houston); Grisham v. City of Fort Worth, 837 F.3d 564 (5th Cir. 2016) (Christian advocate prevented from distributing literature on public sidewalks).

Our dissenting colleagues know this. Take Siders, for example, about a “Christian who seeks to evangelize and share the gospel with others.” 123 F.4th at 298. She “particularly likes to share the gospel on public ways near and outside of public events . . . because such occasions offer her an opportunity to reach a meaningful number of people with her message.” Id. She likes preaching at music venues because “she can find meaningful pedestrian traffic flow on days of amphitheater events as attendees walk toward the amphitheater.” Id. Our colleagues have acknowledged that this is no obscure exercise of faith—to the contrary, they’ve suggested that there’s so much of it that they worry about obstructing traffic. See Siders v. Mississippi, 130 F.4th 188, 190 (5th Cir. 2025) (noting concerns about “congestion,” “interference with ingress or egress,” “ensuring public safety and order,” and “promoting the free flow of traffic on streets and sidewalks” to justify shutting down street preachers).

In sum: In Hershey, the activity is rare, so there’s no liability. In Siders, the activity occurs far too much, so there’s no liability.

No. 21-30754

I would not treat these conflicts as infinitesimally rare. I would suggest that we’re all just making a judgment call as to whether we think this is a First Amendment right worth protecting—or not. 6 III.

That brings us to the main event—religious liberty. As I noted at the outset, the First Amendment violation presented here should be obvious. Of course people have the right to spread the gospel in public spaces. Yet our colleagues deny that this case presents a legitimate religious liberty issue.

Richard Hershey claims the right to share religious materials in public spaces. But our colleagues deny that his claim has anything to do with religious liberty.

To begin with, their challenge to Hershey’s sincerity is ultimately a red herring. Because the rule of liability they propose would govern the most faithful as well as the most imperfect (which is to say, all of us).

_____________________ Our dissenting colleagues highlight the amicus brief filed by the states in our circuit. It’s not surprising, of course, that the states would oppose litigants who seek to hold government officials liable. Counsel for government always argues against liability for government. That’s their job. It’s our job to enforce the Constitution.

The states are concerned about being flooded with frivolous prisoner complaints.

So am I. Of course, they’re already flooded with such suits—since well before Hershey.

We can—and we must—protect the First Amendment rights of law-abiding citizens, even as we work to shield prison officials from frivolous prisoner complaints. We can walk and chew gum at the same time. As federal judges, we’re paid to do so. And that leads me to the question we should really be asking. Why acknowledge only the government amici? Why not also acknowledge the numerous amici who have filed brief after brief after brief expressing grave concerns about the state of freedom of speech and religious liberty in our circuit? See, e.g., McMurry, 142 F.4th at 305–06 (Ho, J., concurring) (noting broad range of amici and religious liberty experts who have criticized our decisions in Morgan and Villarreal); Gonzalez v. Trevino, 60 F.4th 906, 913 & n.4 (5th Cir. 2023) (Ho, J., dissenting from the denial of rehearing en banc) (same).

No. 21-30754

It’s also an issue to be resolved by a jury—not appellate judges. The amended complaint makes clear that “Hershey was distributing free, educational, noncommercial, religious booklets on behalf of a nonprofit organization named the Christian Vegetarian Association.” ROA.75. It’s a message that Hershey has made clear he “feels compelled to share with others.” ROA.192. My colleagues question Hershey’s bona fides as a Christian. But that strikes me as (among other things) beyond our province at this stage of the proceedings.

Hershey’s sincerity was not doubted by the district court. And even at the panel stage, one of our dissenting colleagues doubted only whether Defendants had ever “previously” targeted “someone engaging in free exercise of religion or free speech.” Hershey, 156 F.4th at 577–78 (Richman, J., concurring in part and dissenting in part). Defendants themselves have acknowledged on appeal that Hershey “travels around the country spreading his Christian vegetarian message on college campuses, at event centers, and other forums.” Brief for Defendants, Hershey v. City of Bossier, 156 F.4th 555 (2025) (No. 21-30754), 2022 WL 1136998, *3 (quotations omitted).

But our dissenting colleagues now question “what [the Christian Vegetarian Association] has to do with Christianity.” Post, at 24.

Paul’s epistle to the Romans acknowledges that some people of faith will only eat vegetables. Paul warns believers to respect such views. “The one who eats everything must not treat with contempt the one who does not, and the one who does not eat everything must not judge the one who does, for God has accepted them. Who are you to judge someone else’s servant?

To their own master, servants stand or fall. And they will stand, for the Lord is able to make them stand.” Romans 14:3-4.

Our colleagues also highlight the fact that “Hershey gets paid” for spreading these religious views. Post, at 24. But recall Paul’s first epistle to

No. 21-30754

the church in Corinth. See 1 Corinthians 9:3–14. Paul observes that “those who preach the gospel should receive their living from the gospel.” 1 Corinthians 9:14 (emphasis added). See also Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 761 (1976) (“Speech . . . is protected even though it is . . . ‘sold’ for profit.”); Villarreal, 94 F.4th at 420 (Ho, J., dissenting) (“There is no pro bono requirement to the freedom of speech.”).

So I would not question the sincerity of Hershey’s faith (and especially not at this early stage of the proceedings, on review of a motion to dismiss). Nor is this the first time we’ve seen unwarranted skepticism when it comes to religious liberty. In Siders, our colleagues argued that the ordinance challenged there did not prevent citizens from evangelizing on public grounds. See, e.g., 130 F.4th at 189 (Oldham, J., concurring in the denial of rehearing en banc) (“[T]he ordinance does not purport to regulate prayer, conversation, t-shirts, evangelism, or tracts.”). But in Olivier, by contrast, they condemned the same ordinance at the same amphitheater because it prevents “an evangelical Christian who feels called to share the good news with his fellow citizens . . . from doing so outside the city’s public amphitheater.” 121 F.4th 511, 512 (5th Cir. 2024) (Ho, J., dissenting from denial of rehearing en banc, joined by five members of the court). 7 _____________________ Finally, the dissent suggests that I am violating party presentation by treating this as a religious liberty case, rather than as a free speech case. Post, at 33–34. My colleagues made the same point in Siders. And my response is the same: “religious expression is, of course, protected under the Freedom of Speech Clause.” Siders, 130 F.4th at 192 (Ho, J., dissenting from the denial of rehearing en banc). The Supreme Court’s recent decision in Creative LLC v. Elenis, 600 U.S. 570 (2023), was based on the Free Speech Clause.

Would the dissenters be horrified if anyone called that decision a win for religious liberty?

Surely not. As the Supreme Court recently reminded us, “the Free Speech Clause provides overlapping protection for expressive religious activities.” Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 523 (2022). “That the First Amendment doubly protects religious speech is no accident. It is a natural outgrowth of the framers’ distrust of government

No. 21-30754

*** Judicial opinions should be read charitably. But I struggle to see how our dissenting colleagues haven’t reversed themselves on qualified immunity under Villarreal and Morgan—or on whether Monell applies when there’s been no training whatsoever—or on what government actions threaten religious liberty. If I was one of the religious liberty groups who devotes their God-given talents, time, and treasure to advocating for folks like Spring Siders, Gabriel Olivier, Richard Hershey, and Priscilla Villarreal, I’d wonder why they’re being accused of misrepresenting this court’s precedents and fabricating religious liberty threats out of thin air. I’d also wonder about the unexplained changes of position that they’ve seen from this court. I’d “wonder what’s driving all of these gymnastics.” Post, at 35.

_____________________ attempts to regulate religion and suppress dissent. In Anglo–American history, government suppression of speech has so commonly been directed precisely at religious speech that a free-speech clause without religion would be Hamlet without the prince.” Id. (cleaned up).

No. 21-30754

Andrew S. Oldham, Circuit Judge, joined by Jones, Smith, Richman, Duncan, Engelhardt, and Wilson, Circuit Judges, dissenting from the denial of rehearing en banc: Richard Hershey is a “vegetarian advocate whose ethical beliefs compel him to share his message with others.” ROA.75. When security officers told Hershey to stop distributing his leaflets, he sued for “deprivation of his rights of speech.” ROA.85, 86–87. Hershey does not allege that the officers even knew of the content of his vegetarianism leaflets—let alone targeted him for his vegetarian views. See ROA.75–79.

Nor does Hershey allege anything about his religion. You’ll look in vain for any mention in Hershey’s complaint about faith, religiosity, the First Amendment’s Religion Clauses, or evangelism. The closest he gets is to allege that he distributed leaflets from “the Christian Vegetarian Association”—without any allegation about what that organization is, what it has to do with Christianity, what it has to do with Hershey’s faith, or whether Hershey is even a man of any faith. ROA.75. To the contrary, the complaint alleges that Hershey gets paid to distribute leaflets “by various nonprofit organizations”—without regard to religion of any kind. Ibid.

But you would not know that from the opinion concurring in the denial of rehearing en banc. In that opinion, this case about vegetarian ethics somehow transforms into a battle over street preaching, the Great Commission, hatred of Christians, and religious persecution dating back “thousands of years.” Ante, at 17 (Ho, J., concurring); see also Siders v. Mississippi, 130 F.4th 188, 191 (Ho, J., dissenting from the denial of

No. 21-30754

rehearing en banc) (5th Cir. 2025) (doing the same thing). This quixotic effort does nothing to justify the panel’s badly splintered, three-judge-four- opinion approach to this case. And while it tilts at windmills that appear nowhere in this case, it does nothing to justify our court’s refusal to reconsider the matter en banc.

At the panel stage, the three-judge panel issued four opinions. The per curiam opinion recognized that the panel was (to put it mildly) “splintered.”

Hershey v. City of Bossier City, 156 F.4th 555, 556 (5th Cir. 2025) (per curiam).

Then all three judges issued their own opinions. Judges Richman and Ho wrote the controlling rule on qualified immunity. Judge Ho declared that he would have sided with Judge Dennis on qualified immunity had circuit precedent not prevented him. Then Judges Dennis and Ho wrote the controlling rule on the City’s liability under Monell v. Department of Social Services of City of New York, 436 U.S. 658 (1978). Given the far-reaching implications of the qualified immunity rule embraced by Judge Ho, given the Monell holding embraced by Judges Dennis and Ho, and given that future parties will need Venn diagrams to understand what this deeply fractured panel held, we should have taken the case en banc.

I begin with (I) the panel’s Monell holding, because it’s the most egregious. Then I address (II) the panel’s qualified-immunity holding.

Finally, I address the hyperbolic assertion that (III) rehearing this case would have been comparable to throwing Christians to the lions.

No. 21-30754

I First, Monell. I begin with (A) the Monell rule announced by Judges Dennis and Ho and then turn to (B) its anomalies.

A Judges Dennis and Ho turn the world upside down. See Hershey, 156 F.4th at 561 (Ho, J., concurring); id. at 563–64 (Dennis, J., concurring in relevant part). Before the panel’s decision, plaintiffs could allege a Monell claim based on a single incident only if a municipality “fail[ed] to train its employees concerning a clear constitutional duty implicated in recurrent situations that a particular employee is certain to face.” Littell v. Houston Indep. Sch. Dist., 894 F.3d 616, 624–25 (5th Cir. 2018) (quoting City of Canton v. Harris, 489 U.S. 378, 396 (1989) (O’Connor, J., concurring in part and dissenting in part)). Such circumstances were “very narrow.” Id. at 624. The classic example: if a city gave its officers guns but no training on when to use them. See Canton, 498 U.S. at 390 n.10.

Under the rule embraced by Judges Dennis and Ho, however, plaintiffs need only allege that defendants lacked training to handle plaintiff’s case—even if the ex ante risk of a constitutional violation was infinitesimal.

ROA.86; see also Hershey, 156 F.4th at 561 (Ho, J., concurring); id. at 563– (Dennis, J., concurring in relevant part). This doctrinal Calvinball will have predictable effect: As Judge Richman ably explained, plaintiffs can always “granulate allegations so finely that they arrive at a ‘complete failure to train,’” so they will be able to allege municipal liability in “virtually every

No. 21-30754

instance.” Id. at 578 (Richman, J., dissenting in relevant part) (quotation omitted).

That might be a good rule; it might be a bad one. But either way, it’s a sea change from our court’s prior approach. See Pineda v. City of Houston, 291 F.3d 325, 334 (5th Cir. 2002) (emphasizing that this court has “necessarily [] been wary of finding municipal liability” based on a single incident). And it’s the sort of thing that our en banc court should announce with clarity— not the sort of thing we should sort-of say in a four-way fractured decision that has all the precision of a cubist painting.

B The controlling Monell rule creates a host of anomalies. Three bear emphasis.

First, Monell liability is generally derivative of an underlying constitutional tort committed by an officer. “As is well established, every Monell claim requires an underlying constitutional violation.” Hicks-Fields v. Harris County, 860 F.3d 803, 808 (5th Cir. 2017) (quotation omitted).

Consequently, we routinely dismiss Monell claims against political subdivisions where the plaintiff fails to establish a § 1983 claim against an individual officer. But one of the panel opinions does the opposite: It says no individual officer ever can be held liable—even for the most obvious First Amendment violation—but every political subdivision can always be held liable for any failure to train—even for potentialities that could not reasonably be foreseen. See Hershey, 156 F.4th at 561 (Ho, J., concurring).

No. 21-30754

Second, the panel’s Monell rule proves both too much and too little.

The rule provides that a political subdivision is liable under Monell when it provides “no training whatsoever” on the First Amendment. Hershey, 156 F.4th at 561 (Ho, J., concurring); id. at 563–64 (Dennis, J., concurring in relevant part). That proves far too much because if Bossier City had provided training on say, the intricacies of commercial speech, then presumably this case would have come out differently? It also proves far too little because Hershey provided no evidence whatsoever of a failure to train. Nor did Hershey do anything to show that officers ever faced another vegetarian- ethicist pamphleteer (or any other pamphleteer, or any other speech problem for that matter). See ROA.85–86. So even if Bossier City did not provide any training, that fact does not save Hershey’s deficient pleadings.

Third, the Monell rule is equally applicable to failure-to-train cases under the Fourth or Eighth Amendments. Plaintiffs need only allege that officers were not trained the specific situation at hand and, voila, the plaintiff has a Monell claim. That is why every State in our circuit urged us to rehear this case—and warned us that lawsuits will proliferate, discovery costs will soar, and municipalities will be forced to re-allocate precious time and resources. Br. for the States of Louisiana, Mississippi, and Texas as Amici Curiae, at 6–10. This burden will weigh most heavily on prisons and jails, which face constant failure-to-train allegations. See id. at 8. And it will embolden judges to find Monell violations in such contexts. See, e.g., Alvarez v. City of Brownsville, 904 F.3d 382, 402 (5th Cir. 2018) (en banc) (Dennis, J., dissenting) (dissenting from court’s rejection of Monell claim that city

No. 21-30754

failed to disclose exculpatory evidence); see also Bond v. Nueces County, No. 20-40050, 2022 WL 4595000, at *9 (5th Cir. Sept. 30, 2022) (opinion of Dennis., J.) (vacating district court’s rejection of plaintiff’s Monell claim based on alleged constitutional violations by a prison).

The predictable consequences of the panel’s flawed Monell holding further underscore why we should have taken this case en banc.

II Next, qualified immunity. According to Judge Ho, our court’s approach to qualified immunity in First Amendment cases is deeply flawed.

So flawed that it has garnered scathing criticism from the Alliance Defending Freedom, the First Liberty Institute, and other religious-liberty organizations. If our precedent is that bad, however, we should obviously go en banc to overturn it. It’s surpassing strange to say, “our precedent requires persecution of Christians,” and then say, “we should not go en banc to fix it!” See ante, at 3 (Ho, J., concurring) (saying that).

I explain (A) Fifth Circuit precedent and (B) the puzzling decision to recognize an error, refuse to overturn it, and then complain about it anyway.

A At the panel stage in this case, two judges would have denied qualified immunity. See Hershey, 156 F.4th at 558–59 (Ho, J., concurring); id. at 564 (Dennis, J., dissenting in relevant part). But one of these panel members opted to grant qualified immunity because, apparently, our court has held that the obviousness exception does not apply in First Amendment cases. See

No. 21-30754

id. at 560 (Ho, J., concurring) (citing Villarreal v. City of Laredo, 94 F.4th 374, 395 (5th Cir. 2024) (en banc) (Villarreal II); see also ante, at 3 (same).

I do not read Villarreal II that way. In that case, the panel itself held the state statute at issue was “not one of those ‘obviously unconstitutional’ statutes” on its face. Villarreal v. City of Laredo, 44 F.4th 363, 372 (5th Cir. 2022) (opinion of Ho, J.) (Villarreal I). So the question for the en banc court in Villarreal II was whether officers unconstitutionally applied that valid statute to a “citizen journalist”—even after an independent intermediary found probable cause for the citizen’s arrest and issued a valid warrant. 94 F.4th at 393–94; see also id. at 394 (noting Villarreal alleged nothing to say the independent intermediary was somehow tainted). “[I]n this context,” the court held, cases discussing “obvious” constitutional violations like Hope v. Pelzer, 536 U.S. 730 (2002), and Taylor v. Riojas, 592 U.S. 7 (2020), were “inappropriate templates” for the qualified-immunity inquiry. Villarreal II, 94 F.4th at 395. That does not forever and for all reasons reject obviousness as a ground for denying qualified immunity. It simply holds that, in the context of the particular circumstances of the concededly untainted intermediary and the valid arrest warrant in Villarreal’s case, the officers did not obviously violate the Constitution by arresting her. See id. at 397 (holding Villarreal failed to show an “obvious” constitutional violation—not that “obviousness” is no longer a cognizable claim in our circuit (quotation omitted)).

Judge Ho’s contrary reading of Villarreal II is the opposite of “charitabl[e].” Ante, at 23. Both because it requires reading lines in Villarreal

No. 21-30754

II out of context, and because it pits Villarreal II against binding Supreme Court precedent. As the Supreme Court has said, “[t]here can be no doubt that the First Amendment protects the right to pray”—even in the absence of a case so holding. Sause v. Bauer, 585 U.S. 957, 959 (2018) (per curiam).

Indeed, the entire point of § 1983 liability in First Amendment cases is that plaintiffs can prevail without pointing to earlier, on-point cases. See, e.g., Gonzalez v. Trevino, 602 U.S. 653, 658 (2024) (per curiam). Our sister circuits agree. Some have denied qualified immunity in the First Amendment context on obviousness grounds. See MacIntosh v. Clous, 69 F.4th 309, 320 (6th Cir. 2023); Tobey v. Jones, 706 F.3d 379, 392 (4th Cir. 2013). Many have granted qualified immunity in the First Amendment context while acknowledging that alleged facts did not fall within the obviousness exception. See Diaz-Bigio v. Santini, 652 F.3d 45, 50 (1st Cir. 2011); Nagle v. Marron, 663 F.3d 100, 115–16 (2d. Cir. 2011); Kristofek v. Vill. of Orland Hills, 832 F.3d 785, 798 (7th Cir. 2016); Galvin v. Hay, 374 F.3d 739, 746–47 (9th Cir. 2004); Frasier v. Evans, 992 F.3d 1003, 1021–22 (10th Cir. 2021); Leslie v. Hancock Cnty. Bd. of Educ., 720 F.3d 1338, 1346 (11th Cir. 2013). We should not lightly presume that Villarreal II meant to break with all these authorities—especially when a more plausible and narrower reading of the precedent is readily available.

That is not a “surprise switcheroo.” Ante, at 13 (Ho, J., concurring).

That is simply reading an opinion in the context in which it was decided.

No. 21-30754

B But insofar as Villarreal II held what Judge Ho says it held, that is all the more reason to go en banc and to do so immediately. “This is terrible, so please do not fix it” is an incoherent basis for voting against rehearing. It is even more incoherent to deny rehearing, end the case in this court, and then invite the parties to continue litigating. See ante, at 14 (Ho, J., concurring).

Why say “We are bound by this terrible opinion,” oppose rehearing to fix the purportedly terrible opinion, celebrate the order preserving the purportedly terrible opinion, and then invite additional litigation to, what, remain bound by the purportedly terrible opinion? Wouldn’t it be easier to do the right thing and vote for rehearing now? Unless the point of all of this is to imagine a horrible rule for the purpose of inveighing against it? Whatever the point, a vote to deny en banc rehearing here protects a precedent that Judge Ho reads to “treat[] claims from incarcerated criminals more favorably than law- abiding citizens.” Ante, at 6. An “oddity” for sure. Ibid.

III Finally, a word about the religious-liberty implications of this case.

Judge Ho says the court was correct to deny rehearing in this case because doing so protects “the right to spread the gospel in public spaces.” Ante, at 20. Again, this assertion is bizarre.

Perhaps it’s true that the rules announced in the panel’s four-way, deeply fractured opinions implicate religious liberty. That is all the more reason to rehear this case en banc rather than rely on future parties and panels

No. 21-30754

to decipher the relevant holding in the short per curiam opinion—which is the only opinion that garnered a majority vote. That is especially true if the premise of this entire mess was a reading of Villarreal II that conflicts with Supreme Court precedent, splits from every other circuit to consider the question, and portends disastrous consequences for people of faith. See Part II, supra. But even if it’s true that this case has implications for Christian evangelism, that does not mean that Richard Hershey is himself a Christian evangelist. If Hershey alleged something—anything—about his faith, we’d obviously be bound to accept it. Courts have “no license to declare . . . whether an adherent has ‘correctly perceived’ the commands of his religion.” Masterpiece Cakeshop, Ltd. v. Colo. C.R. Comm’n, 584 U.S. 617, (2018) (Gorsuch, J., concurring). We take a plaintiff’s religious claims as they come, since it is “not for us to say that [a plaintiff’s] religious beliefs are mistaken or insubstantial.” Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 725 (2014). The problem is that Hershey himself alleges nothing about Rome, persecuted Christians, street preaching, the Great Commission, religious liberty, or anything that implicates the First Amendment’s Religion Clauses in any way. But see ante, at 22 (Ho, J., concurring) (invoking 1 Corinthians 9:3–14 for the proposition that Richard Hershey is a Christian pastor?). Observing that fact says nothing about the sincerity of Hershey’s belief. In fact, it does the opposite: it recognizes and honors the belief that Hershey chose to identify in his court filings—namely, the belief

No. 21-30754

that he has a free speech right to hand out vegetarianism pamphlets for money.

In other circumstances, members of our court care about party presentation: In our adversarial system of adjudication, we follow the principle of party presentation. We rely on the parties to frame the issues for decision and assign to courts the role of neutral arbiter of matters the parties present. Our system is designed around the premise that parties represented by competent counsel know what is best for them, and are responsible for advancing the facts and argument entitling them to relief. In short: Courts are essentially passive instruments of government. They do not sally forth each day looking for wrongs to right. They wait for cases to come to them, and when cases arise, courts normally decide only questions presented by the parties.

Gonzalez v. CoreCivic, Inc., 986 F.3d 536, 540 (5th Cir. 2021) (Ho, J., concurring) (quotations omitted). But in this case, the court apparently is an active instrument of government that can sally forth and right wrongs that Richard Hershey and his able counsel never imagined.

* * * At the end of the day, my concurring colleague presents an imaginary case that implicates persecuted Christians in Rome, “religious liberty,” and the “ancient tradition” of street preaching, ante, at 17, 4 (Ho, J., concurring)—rather than the facts of this case. He reads our precedent to

No. 21-30754

create catastrophic consequences for religious liberty—and then says we should retain that horrible precedent and embrace the horrible consequences.

And then he turns Monell upside down and celebrates a new failure-to-train theory that will benefit hardened criminals and saddle political subdivisions in every § 1983 case. You have to wonder what’s driving all of these gymnastics. I would have granted en banc rehearing to figure it out.

I respectfully dissent.