We granted certiorari to decide whether probable cause alone always suffices to defeat a First Amendment retaliatory arrest claim under 42 U.S.C. § 1983. The Court answers that question "no"-a correct and sensible bottom line on which eight Justices agree. There is no basis in § 1983 or in the Constitution to withhold a remedy for an arrest that violated the First Amendment solely because the officer could point to probable cause that some offense, no matter how trivial or obviously pretextual, has occurred.
Unfortunately, a slimmer majority of the Court chooses not to stop there. The majority instead announces a different rule: that a showing of probable cause will defeat a § 1983 First Amendment retaliatory arrest claim unless the person arrested happens to be able to show that "otherwise similarly situated individuals" whose speech differed were not arrested. Ante, at 1733. The Court barely attempts to explain where in the First Amendment or § 1983 it finds any grounding for that rule, which risks letting flagrant violations go unremedied. Because the correct approach would be simply to apply the well-established, carefully calibrated standards that govern First Amendment retaliation claims in other contexts, I respectfully dissent.
I
As Justice GORSUCH explains, the issue here is not whether an arrest motivated by protected speech may violate the First Amendment despite probable cause for the arrest; the question is under what circumstances § 1983 permits a remedy for such a violation. See 1734 - 1737 (opinion concurring in part and dissenting in part). From that common starting point, Justice GORSUCH and I travel far down the same path. I agree that neither the text nor the common-law backdrop of § 1983 supports imposing on First Amendment retaliatory arrest claims a probable-cause requirement that we would not impose in other contexts. See ante, at 1734 - 1737. I agree that Hartman v. Moore , 547 U.S. 250, 126 S.Ct. 1695, 164 L.Ed.2d 441 (2006), turned on concerns specific to malicious prosecution, and that its automatic probable-cause bar therefore does not extend to claims like this one. See ante, at 1732 - 1733. And I agree that-while *1736probable cause has undeniable evidentiary significance to the underlying question of what motivated an arrest-some arrests are demonstrably retaliation for protected speech, notwithstanding probable cause of some coincidental infraction. See ante, at 1732 - 1733; see also ante, at 1730 -1731 (GINSBURG, J., concurring in judgment in part and dissenting in part). Plaintiffs should have a meaningful opportunity to prove such claims when they arise.
I follow this logic to its natural conclusion: Courts should evaluate retaliatory arrest claims in the same manner as they would other First Amendment retaliation claims. Accord, ante, at 1730 - 1731 (opinion of GINSBURG, J.). The standard framework for distinguishing legitimate exercises of governmental authority from those intended to chill protected speech is well established. See Mt. Healthy City Bd. of Ed. v. Doyle , 429 U.S. 274, 287, 97 S.Ct. 568, 50 L.Ed.2d 471 (1977). The plaintiff must first establish that constitutionally protected conduct was a " 'substantial' " or " 'motivating' " factor in the challenged governmental action (here, an arrest). Ibid. If the plaintiff can make that threshold showing, the question becomes whether the governmental actor (here, the arresting officer) can show that the same decision would have been made regardless of the protected conduct. Ibid . If not, the governmental actor is liable. Ibid . In other words, if retaliatory animus was not a "but-for cause" of an arrest, a suit seeking to hold the arresting officer liable will fail "for lack of causal connection between unconstitutional motive and resulting harm." Hartman , 547 U.S. at 260, 126 S.Ct. 1695 ; see also Lozmanv.Riviera Beach , 585 U.S. ----, ----, 138 S.Ct. 1945, 1951-1952, 201 L.Ed.2d 342 (2018)
This timeworn standard is by no means easily satisfied. Even in cases where there is "proof of some retaliatory animus," Hartman , 547 U.S. at 260, 126 S.Ct. 1695, if evidence of retaliatory motive is weak, or evidence of nonretaliatory motive is strong, but-for causation will generally be lacking. That is why probable cause to believe that someone was a serial killer would defeat any First Amendment retaliatory arrest claim-even if, say, there were evidence that the officers also detested the suspect's political beliefs.
With sufficient evidence of retaliatory motive and sufficiently weak evidence of probable cause, however, Mt. Healthy is surmountable. Its orderly framework thus "protects against the invasion of constitutional rights" while burdening legitimate exercises of governmental authority only so far as is "necessary to the assurance of those rights." Mt. Healthy , 429 U.S. at 287, 97 S.Ct. 568 ; see ante, at 1734 - 1735 (opinion of GINSBURG, J.) (Mt. Healthy "strikes the right balance").
II
Regrettably, the Court casts aside the Mt. Healthy standard for many arrests. It instead announces that courts should look beyond the presence of probable cause only when (in its view) the evidence of a constitutional violation is "objective" enough to warrant further inquiry-namely, when a plaintiff can muster evidence "that he was arrested when otherwise similarly situated individuals not engaged in the same sort of protected speech had not been." Ante, at 1727. Plaintiffs who would rely on other evidence to prove a First Amendment retaliatory arrest claim appear to be out of luck, even if they could offer other, unassailable proof of an officer's unconstitutional "statements and motivations." Ibid .
To give partial credit where due: The Court sensibly rejects the absolute probable-cause bar urged by petitioners and embraced by Justice THOMAS, see *1737ante, at 1730 - 1731 (opinion concurring in part and concurring in judgment), and its contrary rule will at least allow the First Amendment to operate in some cases where it is sorely needed. The majority's reasons for imposing a probable-cause bar in some cases but not others, however, do not withstand scrutiny. And by arbitrarily insisting upon comparison-based evidence, the majority's rule fences out First Amendment violations for which redress is equally if not more "warranted," ante, at 1726 - 1727, leaving the public exposed potentially to flagrant abuses.
A
The Court's rationale for the rule it ultimately adopts is hard to discern and, once unearthed, not persuasive. Much of its opinion is spent analogizing to Hartman and to common-law privileges. See ante, at 1723 - 1724, 1726 - 1727. For the reasons Justice GORSUCH explains, that reasoning is not sound. See ante, at 1730 - 1732, 1732 - 1733. Those authorities, in any event, do not support the novel rule the Court imposes. What remains is the majority's practical concern about applying normal First Amendment standards in this context, as well as a handful of inapposite cases involving different constitutional rights.
On the practical side, the majority worries that because discerning the connection between an arrest and a retaliatory motive may involve "causal complexities," ante, at 1723 - 1724; some plaintiffs who raise dubious challenges to lawful arrests may evade early dismissal under Mt. Healthy , see ante, at 1724 - 1725. Our precedents do not permit an interpretation of § 1983 to rest on such "a freewheeling policy choice," Malley v. Briggs , 475 U.S. 335, 342, 106 S.Ct. 1092, 89 L.Ed.2d 271 (1986), and in any event the majority's concerns do not withstand scrutiny.
With regard to the majority's concern that establishing a causal link to retaliatory animus will sometimes be complex: That is true of most unconstitutional motive claims, yet we generally trust that courts are up to the task of managing them. See Crawford-El v. Britton , 523 U.S. 574, 597-601, 118 S.Ct. 1584, 140 L.Ed.2d 759 (1998). And the Mt. Healthy standard accounts for the delicacy of such inquiries with its but-for causation requirement, calibrated to balance governmental interests with individual rights. See 429 U.S. at 287, 97 S.Ct. 568.
As for the risk of litigating dubious claims, the Court pays too high a price to avoid what may well be a marginal inconvenience. Prevailing First Amendment standards have long governed retaliatory arrest cases in the Ninth Circuit, and experience there suggests that trials in these cases are rare-the parties point to only a handful of cases that have reached trial in more than a decade. See Brief for Petitioners 36-39 (identifying four examples); Brief for Respondent 42-49 (discussing those and other Ninth Circuit cases).1
*1738Even accepting that, every so often, a police officer who made a legitimate arrest might have to explain that arrest to a jury, that is insufficient reason to curtail the First Amendment. No legal standard bats a thousand, and district courts already possess helpful tools to minimize the burdens of litigation in cases alleging constitutionally improper motives. See Crawford-El , 523 U.S. at 597-601, 118 S.Ct. 1584. In addition, the burden of a (presumably indemnified2 ) officer facing trial pales in comparison to the importance of guarding core First Amendment activity against the clear potential for abuse that accompanies the arrest power. See Lozman , 585 U.S., at ----, 138 S.Ct., at 1954-1955 ; Part II-B-2, infra .
Finally, and more fundamentally, even if the majority's practical concerns were valid, they would not justify the Court's mix-and-match approach to constitutional law. The Court relies heavily on Devenpeck v. Alford , 543 U.S. 146, 153-154, 125 S.Ct. 588, 160 L.Ed.2d 537 (2004), which held that an arresting officer's state of mind does not matter for purposes of determining whether a Fourth Amendment seizure was supported by probable cause. From this Fourth Amendment holding, the Court extrapolates that First Amendment plaintiffs must show a lack of probable cause (or satisfy its new comparison-based workaround) before their retaliation claims can proceed. See ante, at 1725 - 1726, 1727 - 1728.
This analogy is misguided, and the Court has rightly disavowed it before. In Whren v. United States , 517 U.S. 806, 116 S.Ct. 1769, 135 L.Ed.2d 89 (1996), the Court explained that while "[s]ubjective intentions play no role in ordinary, probable-cause Fourth Amendment analysis," that does not make evidence of an officer's "actual motivations" any less relevant to claims of "selective enforcement" under the Equal Protection Clause. Id., at 813, 116 S.Ct. 1769 ; accord, ante, at 1731 - 1732 (opinion of GORSUCH, J.). First Amendment retaliation claims and equal protection claims are indistinguishable for these purposes; both inherently require inquiry into "an official's motive." Crawford-El , 523 U.S. at 585, 118 S.Ct. 1584. Thus, even if the "[s]ubjective intent of the arresting officer ... is simply no basis for invalidating an arrest" under the Fourth Amendment, Devenpeck , 543 U.S. at 154-155, 125 S.Ct. 588, when it comes to First Amendment freedom of speech, "the government's reason" is often "what counts," see Heffernanv.City of Paterson , 578 U.S. ----, ----, 136 S.Ct. 1412, 1418, 194 L.Ed.2d 508 (2016). Far from supporting the novel burden the Court imposes on First Amendment retaliatory arrest plaintiffs, then, the analogy on which the majority's analysis depends is an unfounded exercise in hybridizing two different constitutional protections. The result is a Frankenstein-like constitutional tort that may do more harm than good.
B
Were it simply an unorthodox solution to an illusory problem, the standard announced today would be benign. But by rejecting direct evidence of unconstitutional *1739motives in favor of more convoluted comparative proof, the majority's standard proposes to ration First Amendment protection in an illogical manner. And those arbitrary legal results in turn will breed opportunities for the rare ill-intentioned officer to violate the First Amendment without consequence-and, in some cases, openly and unabashedly. These are costs the Court should not tolerate.
1
The basic error of the Court's new rule is that it arbitrarily fetishizes one specific type of motive evidence-treatment of comparators-at the expense of other modes of proof.3 In particular, the majority goes out of its way to forswear reliance on an officer's own "statements," ante, at 1727 - 1728, even though such direct admissions may often be the best available evidence of unconstitutional motive. As a result, the Court's standard in some cases will have the strange effect of requiring courts to blind themselves to smoking-gun evidence while simultaneously insisting upon an inferential sort of proof that, though potentially powerful, can be prohibitively difficult to obtain.
The Court's decision to cast aside evidence of the arresting officer's own statements is puzzling. See ibid . In other contexts, when the ultimate question is why a decisionmaker took a particular action, the Court considers the decisionmaker's own statements (favorable or not) to be highly relevant evidence. See, e.g., Masterpiece Cakeshop, Ltd.v.Colorado Civil Rights Comm'n , 584 U.S. ----, ---- - ----, 138 S.Ct. 1719, 1729-1730, 201 L.Ed.2d 35 (2018) ; Cooper v.Harris , 581 U.S. ----, ---- - ----, 137 S.Ct. 1455, 1468-1470, 197 L.Ed.2d 837 (2017) ; Fosterv.Chatman , 578 U.S. ----, ---- - ----, 136 S.Ct. 1737, 1747-1755, 195 L.Ed.2d 1 (2016) ); Church of Lukumi Babalu Aye, Inc. v. Hialeah , 508 U.S. 520, 540-542, 113 S.Ct. 2217, 124 L.Ed.2d 472 (1993) ; Arlington Heights v. Metropolitan Housing Development Corp. , 429 U.S. 252, 270, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977). There is no reason to treat the same sort of evidence so differently here.
Perhaps the majority is worried that statements, like the mental states they evince, can be " 'easy to allege and hard to disprove.' " Ante , at 1725 (quoting Crawford-El , 523 U.S. at 585, 118 S.Ct. 1584 ). Such concerns, whatever their merits, are insufficient reason "to change the burden of proof for an entire category of claims." Id ., at 594, 118 S.Ct. 1584. Besides, more than ever before, an audiovisual record of key events is now often obtainable. Many police departments, for example, equip their officers with body-worn cameras.4 Smartphones that become video cameras with the flick of a thumb are ubiquitous, creating still more potential records.5 In this very case, a local news crew captured some of the relevant events on video, and the officers were wearing audio recorders (though neither had turned his on). The *1740majority appears ready to forsake this body of probative evidence, even though it has the potential to narrow factual disputes and avert trials.
Instead, the majority suggests that comparison-based evidence is the sole gateway through the probable-cause barrier that it otherwise erects. Such evidence can be prohibitively difficult to come by in other selective-enforcement contexts, and it may be even harder for retaliatory arrest plaintiffs to muster.6 After all, while records of arrests and prosecutions can be hard to obtain, it will be harder still to identify arrests that never happened. And unlike race, gender, or other protected characteristics, speech is not typically sorted into statistical buckets that are susceptible of ready categorization and comparison.
The threshold exercise prescribed today-comparing and contrasting a plaintiff's protected speech and allegedly illegal actions with the speech and behavior of others who could have been arrested but were not-is likely to prove vexing in most cases. I suspect that those who can navigate this requirement predominantly will be arrestees singled out at protests or other large public gatherings, where a robust pool of potential comparators happens to be within earshot, eyeshot, or camera-shot. See, e.g., Mam v. Fullerton , 2013 WL 951401, *5 (C.D. Cal., Mar. 12, 2013) (denying summary judgment to an arresting officer where "the only difference between [the plaintiff] and those near him [in a crowd] was the cell phone being used to record"). While some who fit that bill undoubtedly need the protection, see, e.g. , Brief for National Press Photographers Association et al. as Amici Curiae 9-15 (collecting examples of journalists arrested during public protests or gatherings), it is hard to see why those plaintiffs are the only ones deserving of a § 1983 remedy.
2
Put into practice, the majority's approach will yield arbitrary results and shield willful misconduct from accountability. As one example, suppose police respond to reports of a man prowling a front porch. The man says that he is a locked-out homeowner; the police want ID. The man alleges profiling; the officers insist they are just doing their jobs. Tempers flare. A passerby, stepping into a next-door neighbor's yard for a clearer view of the confrontation, pulls out a cell phone camera and begins streaming video of the encounter to her social media followers. One of the officers notices and orders the passerby to stop recording. When the passerby persists, the officer places the passerby under arrest for trespassing.
Will this citizen journalist have an opportunity to prove that the arrest violated her First Amendment rights? Under the majority's test, the answer seems to turn on how many other curious bystanders she can identify who were not arrested in a situation like hers. If she was one of a crowd to enter the neighbor's yard that night, she can sue using her readily available comparator neighbors. But if she was keeping a lonely vigil, she is out of luck (unless she can find some other pool of comparable individuals). And the video of the officer demanding she stop recording moments before the arrest? Irrelevant, apparently. What sense does that make?
*1741Worse, because the majority disclaims reliance on "statements and motivations" for its threshold inquiry, ante , at 1727 - 1728, it risks licensing even clear-cut abuses. Imagine that a reporter is investigating corruption in a police unit. An officer from that unit follows the reporter until the reporter exceeds the speed limit by five miles per hour, then delivers a steep ticket and an explicit message: "Until you find something else to write about, there will be many more where this came from." Cf. Torries v. Hebert , 111 F. Supp. 2d 806, 812 (W.D. La. 2000) (describing allegations that a sheriff arrested proprietors of a local business and "threatened to arrest them again if they continued to play [rap] music"). If even such objectively probative evidence is irrelevant, § 1983 will provide no redress for such flagrant conduct. Meanwhile, the majority's embrace of the Devenpeck rule suggests that a particularly brazen officer could arrest on transparently speech-based grounds and check the statute books later for a potential justification. See 543 U.S. at 153, 125 S.Ct. 588 (holding that probable cause need not be for an "offense actually invoked at the time of arrest").
I do not mean to overstate the clarity of today's holding. What exactly the Court means by "objective evidence," "otherwise similarly situated," and "the same sort of protected speech" is far from clear. See ante, at 1727 -1728.7 I hope that courts approach this new standard commonsensically. It is hard to see what point is served by requiring a journalist arrested for jaywalking to point to specific other jaywalkers who got a free pass, for example, if statistics or common sense confirm that jaywalking arrests are extremely rare. Otherwise, there will be little daylight between the comparison-based standard the Court adopts and the absolute bar it ostensibly rejects.8
C
Justice GORSUCH, alert to the illogic of the majority's position, instead contemplates *1742borrowing a requirement to adduce "clear evidence" of prohibited purpose from our cases concerning equal-protection-based selective-prosecution claims. See ante, at 1733 - 1734 (citing United States v. Armstrong , 517 U.S. 456, 116 S.Ct. 1480, 134 L.Ed.2d 687 (1996) ).9 This suggestion, though perhaps an improvement over the majority's approach, would nevertheless take a doctrine applying (1) equal protection principles (2) in a criminal proceeding to (3) charging decisions by prosecutors, see id. , at 458-459, 464-465, 116 S.Ct. 1480, and ask it also to govern the application of (1) First Amendment principles (2) in a suit for civil damages challenging (3) arrests by police officers. Justice GORSUCH commendably reserves judgment on a proposal not yet subjected to adversarial testing, so I too refrain from speaking too definitively. But I do note that we rejected a very similar rule in Crawford-El . See 523 U.S. at 594, 118 S.Ct. 1584 (rejecting a "clear and convincing" standard for "constitutional claims that require proof of improper intent"). And whatever the merits of the Armstrong rule as currently applied in other contexts, there are good reasons-unexplored by the parties here-to hesitate before extending it.10
III
For the foregoing reasons, I agree with Justice GINSBURG that the tried-and-true Mt. Healthy approach remains the correct one. And because petitioners have not asked us to revisit the Court of Appeals' application of the governing standard, I would affirm.
* * *
The power to constrain a person's liberty is delegated to law enforcement officers by the public in a sacred trust. The First Amendment stands as a bulwark of that trust, erected by people who knew from personal experience the dangers of abuse that follow from investing anyone with such awesome power. Cf. Whitney v. California , 274 U.S. 357, 375-376, 47 S.Ct. 641, 71 L.Ed. 1095 (1927) (Brandeis, J., concurring). Because the majority shortchanges that hard-earned wisdom in the name of marginal convenience, I respectfully dissent.
Because no party questions whether § 1983 claims for retaliatory arrests under the First Amendment are actionable, I assume that § 1983 permits such claims. See Lozmanv.Riviera Beach , 585 U.S. ----, ----, n. 2, 138 S.Ct. 1945, 1956-1957, n. 2, 201 L.Ed.2d 342 (2018) (THOMAS, J., dissenting)
See generally Schwartz, Police Indemnification, 89 N. Y. U. L. Rev. 885, 890 (2014) (empirical study finding that "[p]olice officers are virtually always indemnified").
See Koerner, Note, Between Healthy and Hartman : Probable Cause in Retaliatory Arrest Cases, 109 Colum. L. Rev. 755, 790-794 (2009) (identifying different varieties of evidence potentially available to prove retaliatory motive).
See Fan, Justice Visualized: Courts and the Body Camera Revolution, 50 U. C. D. L. Rev. 897, 930-931 (2017) (noting that police departments in 88 of the 100 largest cities in the United States had "piloted or used police body cameras or ha[d] plans to do so" as of December 2015).
See Simonson, Beyond Body Cameras: Defending a Robust Right To Record the Police, 104 Geo. L. J. 1559, 1564-1565 (2016) ; Kreimer, Pervasive Image Capture and the First Amendment: Memory, Discourse, and the Right To Record, 159 U. Pa. L. Rev. 335, 340-341, 344-351 (2011).
See, e.g., McAdams, Race and Selective Prosecution: Discovering the Pitfalls of Armstrong , 73 Chi.-Kent L. Rev. 605, 618-623 (1998) (describing barriers to obtaining comparator evidence in selective-prosecution cases); Poulin, Prosecutorial Discretion and Selective Prosecution: Enforcing Protection After UnitedStates v. Armstrong , 34 Am. Crim. L. Rev. 1071, 1102-1106 (1997) (discussing examples).
It is also unclear what the majority means when it says that because its threshold "inquiry is objective, the statements and motivations of the particular arresting officer are 'irrelevant.' " Ante , at 1727. That could conceivably be read to mean that all statements are irrelevant, even objectively probative statements describing events in the world-e.g., "I am arresting the libertarians, but not the nonlibertarian protesters who were also trespassing." The facts asserted therein-that libertarians were arrested, nonlibertarians were not, and all were similarly trespassing-are precisely the kind of objective evidence the Court seeks. Similarly, routine police reports-on which the majority surely must intend for plaintiffs to rely-are generally authored by, and thus "statements of," arresting officers. More likely, then, the majority means only that statements describing the officer's internal thought processes are irrelevant (e.g., "I hate libertarians"). But many statements will fall somewhere in between (e.g., "I'm only arresting you because I hate libertarians"). It is hard to see how workable lines can be drawn here.
That could be the unintended result if courts interpret their new task too rigidly. Given the significant evidentiary challenges plaintiffs may face, the best assumption is that the Court intends courts to afford some latitude, especially at the outset of a case. That could mean relying on common experience to assess the most self-evidently minor infractions (such as the Court's jaywalking example); allowing plaintiffs to rely on rough comparisons or inexact statistical evidence where laboratory-like controls cannot realistically be expected; and permitting discovery into potential comparator evidence where a complaint raises a strong inference of unconstitutional motive. For similar reasons, I assume the Court intends courts to permit plaintiffs to draw from a broad universe of potential comparators. And because the test is "objective," ante, at 1727 - 1728, plaintiffs presumably can look beyond the practices of the specific officer or officers who arrested them, to see how other officers handle comparable infractions.
To whatever extent the Court's opinion also seeks kinship with Armstrong , see ante, at 1727 - 1728, I note that Armstrong expressly reserved the question whether comparator evidence alone can provide sufficiently clear evidence of discrimination or whether " 'direct admissions ... of discriminatory purpose' " might also suffice. 517 U.S. at 469, n. 3, 116 S.Ct. 1480. The Court should have followed that example here.
For example, Justice GORSUCH suggests that a potential Armstrong -like rule might be supported by a concern for "separation of powers and federalism." Ante, at 1723 - 1724. While those values are undoubtedly important, they have no apparent interpretive role to play here. Section 1983 exercises Congress' Fourteenth Amendment power to enforce the Constitution against those who wield state authority, "whether they act in accordance with their authority or misuse it." Monroe v. Pape , 365 U.S. 167, 172, 81 S.Ct. 473, 5 L.Ed.2d 492 (1961). It is an emphatic "extension of federal power" into state affairs, id., at 182, 81 S.Ct. 473, one that by its plain terms covers all traditional state prerogatives-including the power to arrest-when wielded to violate federal constitutional rights. Abuses of the arrest power thus are unquestionably among the unconstitutional acts "under color of" state law against which § 1983 operates. See id., at 184, 81 S.Ct. 473 ; see also id., at 174, and n. 10, 81 S.Ct. 473.