United States Court of Appeals for the Fifth Circuit
_____________ United States Court of Appeals Fifth Circuit
FILED
No. 24-30494 July 30, 2026
_____________
Lyle W. Cayce
Clerk
John Ford, former Officer John Doe Police Officer,
Plaintiff—Appellant,
versus
DeRay Mckesson; Black Lives Matter; Black Lives Matter Network, Incorporated,
Defendants—Appellees.
________________________________
Appeal from the United States District Court for the Middle District of Louisiana USDC No. 3:16-CV-742
________________________________
ON PETITION FOR REHEARING EN BANC
Before King, Jones, and Oldham, Circuit Judges. Per Curiam:
Treating the petition for rehearing en banc as a petition for panel rehearing (5th Cir. R.40 I.O.P.), the petition for panel rehearing is DENIED. The petition for rehearing en banc is DENIED because, 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. 24-30494
In the en banc poll, six judges voted in favor of rehearing, Stewart, Graves, Higginson, Willett, Douglas, and Ramirez, Circuit Judges, and ten voted against rehearing Elrod, Chief Judge; Jones, Smith, Richman, Southwick, Haynes, Ho, Duncan, Oldham, and Wilson, Circuit Judges. *
*
Judge Kurt D. Engelhardt, did not participate in the consideration of the rehearing en banc.
3
No. 24-30494
James C. Ho, Circuit Judge, concurring in the denial of rehearing en banc:
The dissent calls for principled enforcement of the First Amendment:
“Constitutional rules cannot turn on the cause, the crowd, or the speaker.” Post, at _ (Willett, J., dissenting from the denial of rehearing en banc). “The First Amendment must protect advocacy without regard to whether the cause is popular or unpopular, favored or disfavored.” Id.
I certainly agree: “Popular speech doesn’t need protection. It’s only when speech is unpopular that you need the First Amendment.” Hershey v. Bossier City, 165 F.4th 292, 292 (5th Cir. 2025) (Ho, J., concurring in the denial of rehearing en banc). “The First Amendment doesn’t mean much if you’re only allowed to express views favored by the government.” Mayfield v. Butler Snow, 78 F.4th 796, 798 (5th Cir. 2023) (Ho, J., dissenting from the denial of rehearing en banc). See also Villarreal v. City of Laredo, 94 F.4th 374, 409 (5th Cir. 2024) (Ho, J., dissenting) (same).
So our rulings must be scrupulously viewpoint neutral. We must apply the same standards, regardless of whether the views at issue are favored or disfavored by judges, cultural elites, or society at large. See, e.g., Oliver v. Arnold, 19 F.4th 843, 844–45 (5th Cir. 2021) (Ho, J., concurring in the denial of rehearing en banc) (urging consistent enforcement of the First Amendment regardless of viewpoint); Gonzalez v. Trevino, 60 F.4th 906, 911–12 (5th Cir. 2023) (Ho, J., dissenting from the denial of rehearing en banc) (expressing concerns about inconsistent enforcement of the First Amendment); Perez v. City of San Antonio, 168 F.4th 345, 346 (5th Cir. 2026) (Ho, J., dissenting from the denial of rehearing en banc) (noting that physical location can play a critical role in the exercise of religion, whether it’s the Lipan-Apache or evangelical Christianity) (citing Siders v. City of Brandon, 130 F.4th 188 (5th Cir. 2025)); Hershey, 165 F.4th at 301 (Ho, J., concurring in the denial of rehearing en banc) (same).
4
No. 24-30494
Where I ultimately part company with my distinguished dissenting colleague is that, unlike him, I see no violation of viewpoint neutrality in this case—only the dutiful application of a simple principle: “The First Amendment protects protest, not trespass.” Doe v. Mckesson, 947 F.3d 874, 878 (5th Cir. 2020) (Ho, J., concurring in the denial of rehearing en banc). This principle “applies with equal force to pro-police protestors” (or to, say, “pro-life and pro-choice protestors alike”) “who unlawfully obstruct a public highway.” Id. at 877. And Plaintiff has presented ample evidence for a jury to put this case in the category of trespass, not protest—as our court has already detailed on more than one occasion. See, e.g., Doe v. Mckesson, 71 F.4th 278, 281–82, 283, 288–89 (5th Cir. 2023); Ford v. Mckesson, 171 F.4th 332, 336, 339–41 (5th Cir. 2026).
I concur in the denial of rehearing en banc.
5
No. 24-30494
Don R. Willett, Circuit Judge, joined by Stewart, Graves *, Higginson, Douglas, and Ramirez, Circuit Judges, dissenting from the denial of rehearing en banc:
The Supreme Court has already called the constitutional issue presented here “undeniably important”: 1 whether the First Amendment tolerates a negligent-protest theory that substitutes the mere foreseeability of a stranger’s violence for intent. After a decade of litigation, full discovery, and another appeal, every threshold obstacle is gone. At last, the question is cleanly presented.
Baton Rouge Police Officer John Ford was grievously injured when someone hurled a rock or piece of concrete into his face during a political protest. The assailant has never been identified. Neither has the assailant’s affiliation, motive, or connection—if any—to DeRay Mckesson. Yet today the court sends Mckesson to trial—not because evidence connects him to the assault, but because he was a prominent activist, allegedly helped lead demonstrators into a roadway, and could have foreseen that disorder might follow.
The First Amendment does not permit prominence to substitute for proof. Under NAACP v. Claiborne Hardware Co., liability for another person’s violence demands proof that the defendant authorized, directed, or ratified the specific tortious activity 2—or that his speech was directed to inciting or producing imminent lawless action and likely to do so under the demanding rule of Brandenburg v. Ohio. 3 And Counterman v. Colorado
*
Judge Graves joins all but Part V.C.
1
Mckesson v. Doe, 592 U.S. 1, 4 (2020) (per curiam).
2
458 U.S. 886, 927–29 (1982).
3
395 U.S. 444, 447 (1969) (per curiam).
6
No. 24-30494
confirms why ordinary negligence cannot do that work when legal sanctions burden political advocacy at the First Amendment’s core. 4 Negligence supplies neither personal responsibility nor the culpable mental state the Constitution demands.
The panel reaches the opposite result through four substitutions. An alleged direction to obstruct traffic becomes a direction to assault an officer; prominence in a national movement becomes command of a particular crowd; proximity becomes causation; and foreseeability becomes intent. Those substitutions evade three constitutional requirements: Claiborne demands personal responsibility and direct, proximate causation; Counterman rejects objective negligence and requires a culpable mental state tied to the conduct for which liability is imposed; and Brandenburg requires intent to produce imminent lawless action and a likelihood that the action will occur. The panel supplies none. Together, its substitutions erase the constitutional boundary between advocacy and violence.
Strip away the euphemisms, and the rule is stark: a freestanding tort of negligent protest—liability imposed on Mckesson not for causing Ford’s injury, but for failing to foresee that someone else might cause it.
Nor will the rule remain confined to Mckesson, Black Lives Matter, or protests against police conduct. It will govern every cause and every viewpoint. The most visible speaker may become the ready-made defendant whenever an unidentified person in the crowd—even an interloper or provocateur—turns violent. The predictable consequence is self-censorship. Citizens will think twice before organizing, promoting, attending, or lending
4 600 U.S. 66, 81–82 (2023).
7
No. 24-30494
their names to public causes. Civil damages can chill political expression as surely as criminal punishment—and often more effectively. 5 Ford’s injuries are abhorrent. He deserves a remedy from whoever is legally responsible for them. But grievous injury does not relax the First Amendment’s demand for proof. The Constitution does not let courts solve the mystery of an unknown assailant by turning the most recognizable speaker into a substitute defendant.
Those inclined to applaud today’s result should consider how the same rule will operate in tomorrow’s case. Today’s defendant is associated with Black Lives Matter. Tomorrow’s may be a gun-rights advocate, a pro-life leader, a pastor protesting a government mandate, or parents addressing a school-board meeting. Constitutional rules cannot turn on the cause, the crowd, or the speaker. The First Amendment must protect advocacy without regard to whether the cause is popular or unpopular, favored or disfavored. Otherwise, constitutional guarantees become fair-weather guarantees.
The en banc court should have stopped this doctrinal erasure. It did not. Now the Supreme Court should. I respectfully dissent.
I. Discovery Disproved the Pleading-Stage Theory The case first reached our court at the pleading stage. We then accepted the complaint’s allegations that Mckesson had organized the Baton Rouge demonstration, “regularly gave orders” to protesters, “personally assumed control” of their movements, and deliberately led them onto a highway. 6 The Supreme Court vacated that decision because our
5 See N.Y. Times Co. v. Sullivan, 376 U.S. 254, 277 (1964) (“The fear of damage awards . . . may be markedly more inhibiting than the fear of prosecution under a criminal statute.”).
6 Doe v. Mckesson, 71 F.4th 278, 288–92 (5th Cir. 2023) (Mckesson I).
8
No. 24-30494
interpretation of Louisiana tort law was “too uncertain a premise” for deciding the First Amendment question. 7 After the Louisiana Supreme Court answered certified questions based on Ford’s allegations, a divided panel reinstated the negligence claim and again rejected Mckesson’s constitutional defense. 8 The case is different now in the way that matters most: discovery tested the allegations, and the resulting record does not show that Mckesson planned the protest, controlled its participants, or had any relationship with the person who struck Ford, much less that he directed the assault on Ford.
Local community members organized the July 2016 demonstration and invited Mckesson to attend. He arrived the day before and retweeted details of an event already circulating online. On the day of the protest, he walked with the crowd and recorded events on his phone. The demonstration began peacefully. Later, some participants looted a convenience store and threw water bottles at police. There is no evidence that Mckesson participated in or encouraged those acts.
As evening approached, word spread that demonstrators should move toward the highway. Some entered the roadway, and officers began clearing the street and making arrests. Mckesson continued filming until his own arrest. While Ford was arresting another person, someone in the crowd threw the object that struck him. The assailant was never identified, and the record never established whether that person was a protester, a bystander, an interloper, a provocateur, or someone else entirely.
7 Mckesson, 592 U.S. at 4–5.
8 Doe v. McKesson, 339 So. 3d 524, 530–36 (La. 2022); Mckesson I, 71 F.4th at 286– 300.
9
No. 24-30494
The district court granted summary judgment. It found no competent evidence that Mckesson organized or led the protest in the manner previously alleged, no evidence that his conduct caused Ford’s injury, and an independent First Amendment bar under Counterman. 9 The panel reversed over Judge King’s searing dissent. It opened by declaring that “[e]ight years of pretrial litigation are enough” and that it was “time for Officer Ford to have a jury assess his claim.”10 But Rule 56 contains no litigation-fatigue exception. The age of a case cannot manufacture a genuine dispute of material fact.
More fundamentally, the panel carried pleading-stage assumptions into the evidentiary record. It declared that our court and the Louisiana Supreme Court had “already held” the facts sufficient for a jury to find causation. 11 Neither court had done so. The Louisiana Supreme Court repeatedly emphasized that it was answering the certified question on “the pleadings alone,” and the prior Fifth Circuit panel likewise relied on “wellpleaded allegations” that Ford still had to prove. 12 The complaint supplied a theory; discovery had to supply proof. It did not. 13
9 Ford v. Mckesson, 739 F. Supp. 3d 344, 350–53 (M.D. La. 2024).
10 Ford v. Mckesson, 171 F.4th 332, 335–36 (5th Cir. 2026) (Mckesson II).
11 Id. at 341.
12 McKesson, 339 So. 3d at 531, 544 n.8 (Crain, J., concurring in part); Mckesson I, 71 F.4th at 292.
13 Ford invokes law of the case, but the earlier decisions did not decide whether the record developed in discovery would establish leadership and causation. The doctrine applies only to issues actually decided, not to questions that had not—and could not yet have—been resolved. See Carpa, Inc. v. Ward Foods, Inc., 567 F.2d 1316, 1320 (5th Cir. 1978), overruled on other grounds by Copper Liquor, Inc. v. Adolph Coors Co., 701 F.2d 542 (5th Cir. 1983).
10
No. 24-30494
A second development matters too. Eleven days after our 2023 decision, the Supreme Court decided Counterman. When the Court later denied certiorari, Justice Sotomayor specifically observed that our court had lacked the benefit of Counterman and expected the lower courts to give that decision “full and fair consideration” in later proceedings. 14 The district court did so. The panel answered that Counterman “changes nothing.” 15 Respectfully, that answer cannot be squared with Claiborne, Counterman, or Brandenburg. The case therefore arrives in a materially different posture from the one before us in 2023: the allegations have been tested, the evidentiary record is closed, and Counterman is now on the books. Two independent constitutional defects bar trial. II. Claiborne Requires Personal Responsibility for the Specific Violence
Foreseeability Is Not Authorization, Direction, or Ratification Claiborne arose from a civil-damages judgment imposed after a political boycott that included both protected advocacy and violence. The Supreme Court reversed and identified three possible bases for liability: (1) unlawful conduct the defendant personally committed; (2) another person’s unlawful conduct, but only upon proof that the defendant “authorized, directed, or ratified specific tortious activity”; or (3) speech satisfying the constitutional test for incitement. 16
14 Mckesson v. Doe, 144 S. Ct. 913, 914 (2024) (Sotomayor, J., statement respecting the denial of certiorari).
15 Mckesson II, 171 F.4th at 343.
16 458 U.S. at 916–18, 927–29.
11
No. 24-30494
Those categories enforce a foundational First Amendment principle:
association may be collective, but culpability is personal. Membership in a movement, participation in a demonstration, even leadership of a political campaign or rally does not make a person answerable for every unlawful act committed by someone associated with the cause. “Civil liability may not be imposed merely because an individual belonged to a group, some members of which committed acts of violence.” 17 Ford does not contend that Mckesson threw the rock. No evidence shows that he asked anyone to throw objects, approved the assault, ratified it afterward, or even knew the assailant. The negligence theory instead treats Mckesson’s alleged direction of demonstrators into a roadway as creating a foreseeable risk of violence.
That theory changes the subject—and with it, the constitutional inquiry. The “specific tortious activity” for which Ford seeks damages is the assault that shattered his jaw—not the obstruction of traffic. Claiborne’s three verbs—authorized, directed, ratified—require a personal nexus between the defendant and the violence of others. 18 They are not satisfied merely because Mckesson allegedly directed demonstrators into the road. Intent to commit one unlawful act does not establish authorization of another person’s violent felony. I previously called this maneuver by its true name: a “negligent protest” theory that lets a stranger’s foreseeable violence stand in for Mckesson’s own wrongdoing. 19
17
Id. at 918–20.
18
Mckesson I, 71 F.4th at 308–10 (Willett, J., concurring in part and dissenting in part).
19
See id. at 302, 313.
12
No. 24-30494
Our 2023 decision purported to require a “sufficiently close relationship” between the leader’s conduct and the third party’s tort, yet deemed that requirement satisfied whenever a leader negligently created conditions in which injury was likely. 20 That formulation supplies no meaningful boundary. Large demonstrations, marches, labor actions, sporting celebrations, and political rallies all carry some foreseeable risk that an attendee will act unlawfully. If furnishing the occasion for possible violence is enough, Claiborne’s personal-responsibility rule disappears precisely when it is needed most.
Claiborne itself exposes the gulf. Charles Evers unquestionably led the boycott. His speeches included threatening language, and violence did occur. Even so, the Court refused to hold him liable without proof that he had authorized, directed, or ratified the particular violence at issue. 21 Mckesson’s connection to the Baton Rouge assault is far weaker: the assailant is unknown, and no evidence ties that person’s act to anything Mckesson said or did.
Proximity Is Not Causation Claiborne also requires “precision of regulation” when liability burdens protected political activity. 22 Damages must be confined to injuries “directly and proximately caused by wrongful conduct chargeable to the defendants.” 23 That requirement is constitutional, not ornamental. Without it, liability migrates from unlawful conduct to protected association and advocacy.
20 Id. at 290, 298 (majority opinion).
21 Claiborne, 458 U.S. at 902, 926–29.
22 Id. at 916 (quoting NAACP v. Button, 371 U.S. 415, 438 (1963)).
23 Id. at 916, 918.
13
No. 24-30494
The panel’s causation analysis leans heavily on decisions made before discovery. The Louisiana Supreme Court could not have found the evidence sufficient for a jury; it had no evidentiary record before it. Our 2023 panel likewise warned that Ford still had to prove causation and described that showing as a “tall task” likely to succeed only in exceptional cases. 24 This appeal is where that proof came due.
It never arrived. No one knows who threw the rock. The record does not show that the assailant attended the demonstration for its political purpose, heard Mckesson, followed him into the roadway, or acted in response to any instruction. Identity, affiliation, and motive all remain unknown.
The timing further weakens the asserted chain. Ford testified that objects had been thrown at police before Mckesson allegedly directed anyone into the road. He also described independent agitators urging people toward the roadway and testified that he was struck while arresting one of them— someone other than Mckesson. 25 On this record, the crowd’s movement and violence had causes independent of Mckesson.
The asserted chain runs from an alleged direction to enter the street, through an already unruly crowd, a police response, the arrest of someone else, and finally an unknown assailant’s independent decision to throw a rock. That is chronology, not causation—a series of increasingly remote contingencies. Claiborne requires direct and proximate causation. 26 The sequence supplies temporal and physical proximity, not the indispensable link between Mckesson and the assault.
24 Mckesson I, 71 F.4th at 292, 299.
25 See id. at 292 n.8; Mckesson II, 171 F.4th at 350–52 (King, J., dissenting).
26 458 U.S. at 916, 918.
14
No. 24-30494
The missing assailant is not an evidentiary gap at the margin. It is the void at the center of the case. The panel’s assurance that its rule will rarely reach interlopers and counterprotesters assumes the very fact the record does not establish—that the person who injured Ford was a protester whose conduct Mckesson incited, organized, or directed. 27 The record supplies no basis for that assumption. That failure of proof is independently dispositive.
III. Counterman Forecloses Liability Based on Objective Negligence
Counterman Requires a Culpable Mental State Tied to the Assault Counterman identifies a second constitutional defect: Ford’s theory rests on objective negligence untethered to the assault. The panel dismisses Counterman as an inapposite criminal true-threat case. The Supreme Court’s holding concerned true threats, but its reasoning speaks directly to the constitutional problem here.
Counterman rejected objective negligence for threatening speech and required at least subjective recklessness. 28 More important here, the Court distinguished incitement. Incitement decisions “demand more,” the Court explained, because political advocacy lies “a hair’s-breadth away” from punishable incitement. 29 Those cases require specific intent—purpose or knowledge—to prevent legal sanctions from bleeding into “dissenting political speech at the First Amendment’s core.” 30 The Court cited Claiborne as part of that line.
27 Mckesson II, 171 F.4th at 341–42.
28 600 U.S. at 72, 79 & n.5.
29 Id. at 81.
30 Id. at 81–82 (discussing Brandenburg, Hess, and Claiborne).
15
No. 24-30494
A negligence action arising from a protest against government conduct sits squarely within that concern. The panel’s rule imposes liability without proof that Mckesson intended violence, knew that his own words or conduct would cause it, or consciously disregarded a substantial risk that they would do so. It asks only whether a reasonable person should have foreseen some confrontation at some point.
That is the very objective-negligence standard Counterman rejected, applied in a setting where the Court said the Constitution demands more protection, not less. 31 Counterman does not mechanically convert every First Amendment case into a true-threat prosecution. It does, however, foreclose the premise underlying our 2023 decision—that protest-related speech may support liability for another person’s violence with no “intent condition.” 32 The panel’s rule-of-orderliness analysis underscores why rehearing en banc was necessary. 33 A later panel may have considered itself bound. The en banc court was not. We had both the authority and the responsibility to reconcile circuit law with the Supreme Court’s treatment of Counterman and incitement as intent-based doctrines.
Intent to Obstruct Traffic Is Not Intent to Cause the Assault The panel says Counterman is satisfied because a jury could find that Mckesson intended “disorder” when he allegedly directed people into a public street. 34 The panel invokes the right mens rea but attaches it to the wrong act.
31 Id. at 81.
32 Mckesson I, 71 F.4th at 297.
33 See Mckesson II, 171 F.4th at 342–43.
34 Id. at 343 (first quoting Counterman, 600 U.S. at 76; then quoting Cantwell v.
Connecticut, 310 U.S. 296, 308 (1940)).
16
No. 24-30494
The required mental state must attach to the violence for which Ford seeks damages—a rock thrown into his face, not delayed traffic. Intent to obstruct traffic is not intent to shatter an officer’s jaw. Proof that Mckesson intended civil disobedience does not show that he intended, knew of, or consciously disregarded a substantial risk of this particular assault by an unknown person. Otherwise, any intent to violate a minor law during a protest would expose the speaker to liability for every foreseeable crime that followed.
That inversion is impossible to reconcile with the Court’s decisions in Brandenburg and Hess v. Indiana. Even express advocacy of force or lawbreaking remains protected unless it is directed to producing imminent lawless action and likely to do so. 35 Under the panel’s foreseeability approach, nonviolent advocacy of a road obstruction receives less protection than explicit advocacy of violence: the former may support damages for a stranger’s assault on negligence alone, while the latter cannot be punished without specific intent, imminence, and likelihood. Brandenburg buried that bad-tendency rationale. 36 The panel exhumes it, rechristens it “foreseeable violence,” and turns every speaker into the insurer of every stranger’s crime. Tort law cannot accomplish what Brandenburg forbids.
Civil Damages Do Not Evade the First Amendment The panel also emphasizes that Counterman involved a criminal conviction. But the First Amendment does not disappear when punishment
35 Brandenburg, 395 U.S. at 447; Hess v. Indiana, 414 U.S. 105, 108–09 (1973) (per curiam).
36 395 U.S. at 447 (overruling Whitney v. California, which permitted punishment of advocacy based on its perceived tendency to produce violence); id. at 447 n.2 (explaining that Yates v. United States overturned convictions resting on “mere advocacy, unrelated to its tendency to produce forcible action”).
17
No. 24-30494
arrives through a verdict form rather than an indictment. The constitutional mens rea principles on which Counterman relied cross the civil-criminal line. The Court drew from defamation, incitement, and obscenity—not to mention Claiborne itself, a civil-damages case. 37 Garrison v. Louisiana likewise explained that freedom of expression requires the same constitutional standard for civil and criminal remedies. 38 The distinction is especially hollow when the constitutional concern is chilling protected speech. Civil liability can impose crushing financial consequences without the criminal law’s heightened burden of proof and related safeguards. The Supreme Court has long recognized that “[t]he fear of damage awards” may inhibit speech even more than criminal prosecution. 39 Justice Barrett’s Counterman dissent made the civil reach explicit: because the case concerned the scope of the First Amendment, not merely the construction of a criminal statute, its reasoning affects “the civil consequences” as well. 40 And when Justice Sotomayor addressed this very litigation after Counterman, she explained that the Court’s decision bars objective negligence as a basis for punishing speech and treats Claiborne and the incitement cases as demanding intent. 41 The panel devoted one paragraph to that intervening guidance—and gave it no effect. It then returned to a negligence regime that allows a jury to impose damages without finding any culpable mental state tied to the assault.
37 Counterman, 600 U.S. at 73, 76, 80–82.
38 379 U.S. 64, 74 (1964).
39 N.Y. Times, 376 U.S. at 277.
40 Counterman, 600 U.S. at 119 (Barrett, J., dissenting).
41 Mckesson, 144 S. Ct. at 914 (Sotomayor, J., statement respecting the denial of certiorari).
18
No. 24-30494
That is not “full and fair consideration” of Counterman. 42 It is acknowledgment without application.
IV. Prominence Is Not Proof of Control The panel compounds the constitutional error with the evidence it uses to send the case to trial. The record may show that Mckesson was a nationally prominent activist. It does not show that he organized this protest, commanded this crowd, or caused this injury. The assailant is unknown. The activist is famous. The panel lets fame fill the evidentiary void.
Promotion Is Not Organization The panel quotes Mckesson as saying he was “part of” a group of “incredible leaders who were planning things . . . associated . . . with the protest.” 43 But the omitted testimony is the point. Mckesson explained that there was no formal “planning group”; he was describing a decentralized, community-led effort involving hundreds of people. The ellipses do not merely shorten the answer. They excise the qualification that defeats the inference of centralized command.
The rest of the record points in the same direction. Baton Rouge community leaders invited Mckesson to attend. Being invited suggests attendance, not control. Mckesson was a guest, not a host. His retweet merely amplified an event already announced; it did not organize the event or give him authority over it. Social media enables speakers to publicize events without planning or directing them. Treating that amplification as operational
42 See id.
43 Mckesson II, 171 F.4th at 339.
19
No. 24-30494
command converts ordinary participation in “the modern public square” into proof of control. 44 The panel then reaches further afield—to Mckesson’s national profile, his involvement in other protests, and his meetings with federal officials.45 Those facts may show influence within a movement. They do not establish who planned this demonstration, commanded this crowd, or caused this injury. The panel converts a political résumé into a chain of command— the very associative shortcut Claiborne forbids. 46 Presence Is Not Command
Ford testified that the crowd followed Mckesson and that Mckesson appeared to be getting people ready to enter the highway. But Ford repeatedly acknowledged that the crowd was too loud for him to hear what Mckesson said and that he did not recall hand signals or gestures. Only later, while being questioned by his own counsel, did he recall an instruction to enter the road.
Even crediting that recollection, it proves at most a single instruction to enter the roadway during a moving demonstration. It does not establish the sweeping allegations that carried the case past dismissal: that Mckesson organized the event, regularly gave orders, personally controlled the protest’s movements, and directed it at all times. 47 The contemporaneous video is more telling. The panel reads it to show police cutting off the march, the crowd turning, and Mckesson at the
44 See Packingham v. North Carolina, 582 U.S. 98, 107 (2017).
45 Mckesson II, 171 F.4th at 340.
46 458 U.S. at 918–19.
47 Mckesson I, 71 F.4th at 288–92.
20
Case: 24-30494 Document: 120-1 Page: 20 Date Filed: 07/30/2026
No. 24-30494
front. 48 Judge King reviewed the same footage and explained that it shows no such sequence. 49 The crowd does not change direction on camera; police remain behind it; and Mckesson appears within the crowd, sometimes stopping or moving backward to film arrests. The video shows no order, no signal, no command.
When video evidence blatantly contradicts a witness’s account, courts must view the facts as the recording depicts them. 50 The panel sees a commander. The video shows a man documenting arrests. Visibility in a crowd is not command of it.
Movement Leadership Is Not Control of This Crowd The panel cites a television interview from more than a year before the Baton Rouge protest, characterizing Mckesson as supporting violent protests and refusing to condemn violence. 51 But as the record shows, the interview also contains statements favoring peaceful protest, and it concerned neither this demonstration nor the assault on Ford. Even general advocacy of violence would not support liability for specific violence that is remote in time and unconnected to the speech. 52 A year-old interview about other events does still less to prove that Mckesson commanded this crowd or caused this injury.
The panel next assigns similar weight to the word “leadership” in a class-action settlement filing. 53 The term appeared among qualities offered to
48
Mckesson II, 171 F.4th at 341.
49
Id. at 346–47 (King, J., dissenting).
50
Scott v. Harris, 550 U.S. 372, 380–81 (2007).
51
Mckesson II, 171 F.4th at 336, 340.
52
See Claiborne, 458 U.S. at 928; see also Mckesson I, 71 F.4th at 291.
53
Mckesson II, 171 F.4th at 340.
21
No. 24-30494
show that several named plaintiffs would adequately represent a class under Rule 23. In context, it referred to Mckesson’s national advocacy and commitment to the litigation—not operational control of the Baton Rouge march. 54 Leadership of a movement, adequacy as a class representative, and command of a particular crowd are distinct propositions. The panel treats them as interchangeable.
Most troubling, the panel treats Mckesson’s political activity—his public profile, associations, prior protests, media appearances, and meetings with government officials—as evidence that he led this event and negligently caused a stranger’s violence. 55 That method turns civic effectiveness into evidentiary exposure: the more prominent the advocate, the more readily prominence can be recast as control when disorder erupts nearby. Visibility becomes liability.
The First Amendment does not convert visibility into vicarious liability. Claiborne protected even the acknowledged leader of a campaign because the Constitution required evidence connecting him personally to the specific violence at issue. The panel sends Mckesson to trial on far less.
Once prominence may substitute for control, and foreseeability for intent, the rule cannot be confined to this case.
V. The Panel’s Rule Has No Limiting Principle The panel’s rule is anything but narrow. The record confirms it.
Liability may attach where the tortfeasor is unidentified; where no evidence connects that person to the defendant; where the alleged direction concerned
54 Memorandum in Support of Motion for Final Approval of Class Action Settlement at 15, McKesson v. City of Baton Rouge, No. 16-520-JWD-RLB, 2017 WL 11675692 (M.D. La. Nov. 16, 2017); Mckesson II, 171 F.4th at 351 (King, J., dissenting).
55 Mckesson II, 171 F.4th at 340.
22
No. 24-30494
a different unlawful act; and where leadership is inferred largely from prominence, association, and public advocacy. A rule is not narrow merely because its limitless premises are recited one case at a time.
A. The Rule Creates a Civil Heckler’s Veto That rule creates a civil heckler’s veto. If the assailant may have been a counterprotester, outsider, or unaffiliated agitator, then violence by an opponent can increase the organizer’s exposure. The panel calls that concern unlikely because liability supposedly requires direction of the protester’s actions. 56 But without proof that the assailant was a protester, the asserted safeguard is no safeguard at all. A provocateur need not silence the speaker; he need only make speaking ruinously expensive.
B. Police Reaction Becomes Evidence Against Police Critics The panel also ties foreseeability to the prospect that police would respond to demonstrators in the roadway. 57 That creates a constitutional inversion: the more aggressively police respond to a protest against police, the greater the potential liability of the protest’s most visible participant. Official reaction thus becomes evidence against official critics. 58 History supplies a familiar illustration. In 1968, exactly one week before his assassination, Dr. Martin Luther King Jr. led a Memphis march supporting striking sanitation workers. Some participants broke windows; police intervened; and violence followed. Dr. King unquestionably led that march, and disorder at a large demonstration was foreseeable. Yet Claiborne would not have allowed an injured officer or shopkeeper to hold him
56 Id. at 342.
57 Id. at 341.
58 Id. at 352 (King, J., dissenting) (calling the rule a “‘heckler’s veto’ on steroids” because law enforcement’s response determines liability exposure).
23
No. 24-30494
personally liable without proof connecting him to the specific wrongful act. 59 The panel’s negligence theory offers no principled reason why Dr. King would have fared differently—and thus no principled stopping point.
C. Constitutional Rules Do Not Keep Partisan Company The next defendant may be a gun-rights advocate whose rally is disrupted by an armed provocateur, a pro-life leader whose march attracts a vandal, a pastor whose protest against a government mandate is infiltrated by an agitator, or parents whose school-board testimony is thrown into disorder by someone else. Today’s result may seem compelling because the plaintiff is an injured officer. But constitutional protections cannot turn on the speaker, the cause, or the sympathies a particular case evokes. The First Amendment protects advocacy across the spectrum—popular and despised, orderly and unruly, left and right. The breadth of the panel’s rule makes the constitutional danger substantial—and Supreme Court review unusually appropriate.
VI. This Case Is Now a Clean Vehicle for Review Every obstacle that complicated Supreme Court review in 2020 is now gone. The state-law uncertainty that prompted the Court’s vacatur has been resolved. Discovery is complete. The pleading-stage assumptions have been tested. The district court entered final summary judgment. Counterman intervened, and the panel expressly held that it changes nothing. The en banc court has now declined to intervene.
The constitutional question is therefore squarely presented and unavoidable: may a political speaker be held liable under a negligence theory for an unidentified stranger’s violence without proof that the speaker authorized, directed, ratified, or intentionally incited that violence?
59 Mckesson I, 71 F.4th at 313 (Willett, J., concurring in part and dissenting in part).
24
No. 24-30494
Claiborne, Counterman, and Brandenburg answer no. Our court now answers yes.
VII. The Supreme Court Should Restore the Line Between Advocacy and Violence Officer Ford suffered a devastating injury in the line of duty. The person who struck him should answer. So should anyone who authorized, directed, ratified, or intentionally incited the attack. But the First Amendment does not permit liability to migrate to the most visible activist simply because the actual assailant cannot be found.
The panel replaces every constitutional safeguard with an evidentiary shortcut: negligence for personal culpability, generalized disorder for intent to produce imminent violence, proximity for causation, and political prominence for event-specific proof. Claiborne demands personal responsibility. Counterman demands a culpable mental state. Brandenburg demands intent, imminence, and likelihood. The panel supplies none. Political advocacy may not be punished merely because a reasonable person should have foreseen someone else’s crime.
The constitutional damage begins long before any jury deliberates.
Citizens deciding whether to organize, publicize, attend, or lend their names to a demonstration must now weigh the risk of financial ruin for an unknown stranger’s violence before exercising a First Amendment right. Mckesson has lived under that threat for more than a decade. When the price of advocacy may be financial ruin for a stranger’s violence, silence is the predictable result—and the constitutional injury.
The Supreme Court called this issue “undeniably important” more than half a decade ago, when unsettled state law stood in the way. 60 That
60 Mckesson, 592 U.S. at 4.
25
No. 24-30494
obstacle is gone. Discovery is complete. The record is closed. Counterman has intervened. And the panel has squarely held that it changes nothing. The conflict is direct, acknowledged, and outcome-determinative—not latent or academic. The panel has also resurrected the bad-tendency rationale that Brandenburg buried.
The Court should grant review and restore the constitutional line between advocacy and violence. Political speakers may be held to account for violence they commit, authorize, direct, ratify, or intentionally incite. But they may not be conscripted as insurers of a crowd—or as substitute defendants for an unknown assailant—simply because they were visible, prominent, and nearest the microphone.
I respectfully dissent.