NIEVES ET AL. v. BARTLETT (2019)

Respondent Russell Bartlett was arrested by police officers Luis Nieves and Bryce Weight for disorderly conduct and resisting arrest during "Arctic Man," a raucous winter sports festival held in a remote part of Alaska. According to Sergeant Nieves, he was speaking with a group of attendees when a seemingly intoxicated Bartlett started shouting at them not to talk to the police. When Nieves approached him, Bartlett began yelling at the officer to leave. Rather than escalate the situation, Nieves left. Bartlett disputes that account, claiming that he was not drunk at that time and did not yell at Nieves. Minutes later, Trooper Weight says, Bartlett approached him in an aggressive manner while he was questioning a minor, stood between Weight and the teenager, and yelled with slurred speech that Weight should not speak with the minor. When Bartlett stepped toward Weight, the officer pushed him back. Nieves saw the confrontation and initiated an arrest. When Bartlett was slow to comply, the officers forced him to the ground. Bartlett denies being aggressive and claims that he was slow to comply because of a back injury. After he was handcuffed, Bartlett claims that Nieves said "bet you wish you would have talked to me now."

          Bartlett sued under 42 U. S. C. §1983, claiming that the officers violated his First Amendment rights by arresting him in retaliation for his speech--i.e., his initial refusal to speak with Nieves and his intervention in Weight's discussion with the minor. The District Court granted summary judgment for the officers, holding that the existence of probable cause to arrest Bartlett precluded his claim. The Ninth Circuit reversed. It held that probable cause does not defeat a retaliatory arrest claim and concluded that Bartlett's affidavit about what Nieves allegedly said after the arrest could enable Bartlett to prove that the officers' desire to chill his speech was a but-for cause of the arrest.

Held: Because there was probable cause to arrest Bartlett, his retaliatory arrest claim fails as a matter of law. Pp. 4-16.

     (a) To prevail on a claim such as Bartlett's, the plaintiff must show not only that the official acted with a retaliatory motive and that the plaintiff was injured, but also that the motive was a "but-for" cause of the injury. Hartman v. Moore, 547 U. S. 250, 259-260. Establishing that causal connection may be straightforward in some cases, see, e.g., Mt. Healthy City Bd. of Ed. v. Doyle, 429 U. S. 274, but other times it is not so simple. In retaliatory prosecution cases, for example, the causal inquiry is particularly complex because the official alleged to have the retaliatory motive does not carry out the retaliatory action himself. Instead, the decision to bring charges is made by a prosecutor--who is generally immune from suit and whose decisions receive a presumption of regularity. To account for that "problem of causation," plaintiffs in retaliatory prosecution cases must prove as a threshold matter that the decision to press charges was objectively unreasonable because it was not supported by probable cause. Hartman, 547 U. S., at 263. Pp. 5-7.

     (b) Because First Amendment retaliatory arrest claims involve causal complexities akin to those identified in Hartman--see, e.g., Reichle v. Howards, 566 U. S. 658; Lozman v. Riviera Beach, 585 U. S. ___--the same no-probable-cause requirement generally should apply. The causal inquiry is complex because protected speech is often a "wholly legitimate consideration" for officers when deciding whether to make an arrest. Reichle, 566 U. S., at 668. In addition, "evidence of the presence or absence of probable cause for the arrest will be available in virtually every retaliatory arrest case." Ibid. Its absence will generally provide weighty evidence that the officers' animus caused the arrest, whereas its presence will suggest the opposite. While retaliatory arrest cases do not implicate the presumption of prosecutorial regularity or necessarily involve multiple government actors, the ultimate problem remains the same: For both claims, it is particularly difficult to determine whether the adverse government action was caused by the officers' malice or by the plaintiff's potentially criminal conduct.

     Bartlett's proposed approach disregards the causal complexity involved in these cases and dismisses the need for any threshold objective showing, moving directly to consideration of the officers' subjective intent. In the Fourth Amendment context, however, this Court has "almost uniformly rejected invitations to probe [officers'] subjective intent," Ashcroft v. al-Kidd, 563 U. S. 731, 737. A purely subjective approach would undermine that precedent, would "dampen the ardor of all but the most resolute, or the most irresponsible, in the unflinching discharge of their duties," Gregoire v. Biddle, 177 F. 2d 579, 581, would compromise evenhanded application of the law by making the constitutionality of an arrest "vary from place to place and from time to time" depending on the personal motives of individual officers, Devenpeck v. Alford, 543 U. S. 146, 154, and would encourage officers to minimize communication during arrests to avoid having their words scrutinized for hints of improper motive. Pp. 8-11.

     (c) When defining the contours of a §1983 claim, this Court looks to "common-law principles that were well settled at the time of its enactment." Kalina v. Fletcher, 522 U. S. 118, 123. In 1871, when §1983 was enacted, there was no common law tort for retaliatory arrest based on protected speech. Turning to the "closest analog[s]," Heck v. Humphrey, 512 U. S. 477, 484, both false imprisonment and malicious prosecution suggest the same result: The presence of probable cause should generally defeat a First Amendment retaliatory arrest claim. Pp. 12-13.

     (d) Because States today permit warrantless misdemeanor arrests for minor criminal offenses in a wide range of situations--whereas such arrests were privileged only in limited circumstances when §1983 was adopted--a narrow qualification is warranted for circumstances where officers have probable cause to make arrests, but typically exercise their discretion not to do so. An unyielding requirement to show the absence of probable cause in such cases could pose "a risk that some police officers may exploit the arrest power as a means of suppressing speech." Lozman, 585pp U. S., at ___. Thus, the no-probable-cause requirement should not apply when a plaintiff presents objective evidence that he was arrested when otherwise similarly situated individuals not engaged in the same sort of protected speech had not been. Cf. United States v. Armstrong, 517 U. S. 456, 465. Because this inquiry is objective, the statements and motivations of the particular arresting officer are irrelevant at this stage. After making the required showing, the plaintiff's claim may proceed in the same manner as claims where the plaintiff has met the threshold showing of the absence of probable cause. Pp. 13-15.

712 Fed. Appx. 613, reversed and remanded.

     Roberts, C. J., delivered the opinion of the Court, in which Breyer, Alito, Kagan, and Kavanaugh, JJ., joined, and in which Thomas, J., joined except as to Part II-D. Thomas, J., filed an opinion concurring in part and concurring in the judgment. Gorsuch, J., filed an opinion concurring in part and dissenting in part. Ginsburg, J., filed an opinion concurring in the judgment in part and dissenting in part. Sotomayor, J., filed a dissenting opinion.

Opinion of the Court

 587 U. S. ____ (2019)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

No. 17-1174

     Chief Justice Roberts delivered the opinion of the Court.

     Respondent Russell Bartlett sued petitioners--two police officers--alleging that they retaliated against him for his protected First Amendment speech by arresting him for disorderly conduct and resisting arrest. The officers had probable cause to arrest Bartlett, and we now decide whether that fact defeats Bartlett's First Amendment claim as a matter of law.

I

A

     Bartlett was arrested during "Arctic Man," a weeklong winter sports festival held in the remote Hoodoo Mountains near Paxson, Alaska. Paxson is a small community that normally consists of a few dozen residents. But once a year, upwards of 10,000 people descend on the area for Arctic Man, an event known for both extreme sports and extreme alcohol consumption. The mainstays are high-speed ski and snowmobile races, bonfires, and parties. During that week, the Arctic Man campground briefly becomes one of the largest and most raucous cities in Alaska.

     The event poses special challenges for law enforcement. Snowmobiles, alcohol, and freezing temperatures do not always mix well, and officers spend much of the week responding to snowmobile crashes, breaking up fights, and policing underage drinking. Given the remote location of the event, Alaska flies in additional officers from around the State to provide support. Still, the number of police remains limited. Even during the busiest periods of the event, only six to eight officers are on patrol at a time.

     On the last night of Arctic Man 2014, Sergeant Luis Nieves and Trooper Bryce Weight arrested Bartlett. The parties dispute certain details about the arrest but agree on the general course of events, some of which were captured on video by a local news reporter.

     At around 1:30 a.m., Sergeant Nieves and Bartlett first crossed paths. Nieves was asking some partygoers to move their beer keg inside their RV because minors had been making off with alcohol. According to Nieves, Bartlett began belligerently yelling to the RV owners that they should not speak with the police. Nieves approached Bartlett to explain the situation, but Bartlett was highly intoxicated and yelled at him to leave. Rather than escalate the situation, Nieves left. Bartlett disputes that account. According to Bartlett, he was not drunk at that time and never yelled at Nieves. He claims it was Nieves who became aggressive when Bartlett refused to speak with him.

     Several minutes later, Bartlett saw Trooper Weight asking a minor whether he and his underage friends had been drinking. According to Weight, Bartlett approached in an aggressive manner, stood between Weight and the teenager, and yelled with slurred speech that Weight should not speak with the minor. Weight claims that Bartlett then stepped very close to him in a combative way, so Weight pushed him back. Sergeant Nieves saw the confrontation and rushed over, arriving right after Weight pushed Bartlett. Nieves immediately initiated an arrest, and when Bartlett was slow to comply with his orders, the officers forced him to the ground and threatened to tase him.

     Again, Bartlett tells a different story. He denies being aggressive, and claims that he stood close to Weight only in an effort to speak over the loud background music. And he was slow to comply with Nieves's orders, not because he was resisting arrest, but because he did not want to aggravate a back injury. After Bartlett was handcuffed, he claims that Nieves said: "[B]et you wish you would have talked to me now." 712 Fed. Appx. 613, 616 (CA9 2017).

     The officers took Bartlett to a holding tent, where he was charged with disorderly conduct and resisting arrest. He had sustained no injuries during the episode and was released a few hours later.

B

     The State ultimately dismissed the criminal charges against Bartlett, and Bartlett then sued the officers under 42 U. S. C. §1983, which provides a cause of action for state deprivations of federal rights. As relevant here, he claimed that the officers violated his First Amendment rights by arresting him in retaliation for his speech. The protected speech, according to Bartlett, was his refusal to speak with Nieves earlier in the evening and his intervention in Weight's discussion with the underage partygoer. The officers responded that they arrested Bartlett because he interfered with an investigation and initiated a physical confrontation with Weight. The District Court granted summary judgment for the officers. The court determined that the officers had probable cause to arrest Bartlett and held that the existence of probable cause precluded Bartlett's First Amendment retaliatory arrest claim.

     The Ninth Circuit disagreed. 712 Fed. Appx. 613. Relying on its prior decision in Ford v. Yakima, 706 F. 3d 1188 (2013), the court held that a plaintiff can prevail on a First Amendment retaliatory arrest claim even in the face of probable cause for the arrest. According to the Ninth Circuit, Bartlett needed to show only (1) that the officers' conduct would "chill a person of ordinary firmness from future First Amendment activity," and (2) that he had advanced evidence that would "enable him ultimately to prove that the officers' desire to chill his speech was a but-for cause" of the arrest. 712 Fed. Appx., at 616 (internal quotation marks omitted). The court concluded that Bartlett had satisfied both requirements: A retaliatory arrest is sufficiently chilling, and Bartlett had presented enough evidence that his speech was a but-for cause of the arrest. The only causal evidence relied on by the court was Bartlett's affidavit alleging that Sergeant Nieves said "bet you wish you would have talked to me now." If that allegation were true, the court reasoned, a jury might conclude that the officers arrested Bartlett in retaliation for his statements earlier that night.

     The officers petitioned for review in this Court, and we granted certiorari. 585 U. S. ___ (2018).

II

     We are asked to resolve whether probable cause to make an arrest defeats a claim that the arrest was in retaliation for speech protected by the First Amendment. We have considered this issue twice in recent years. On the first occasion, we ultimately left the question unanswered because we decided the case on the alternative ground of qualified immunity. See Reichle v. Howards, 566 U. S. 658 (2012). We took up the question again last Term in Lozman v. Riviera Beach, 585 U. S. ___ (2018). Lozman involved unusual circumstances in which the plaintiff was arrested pursuant to an alleged "official municipal policy" of retaliation. Id., at ___ (slip op., at 11). Because those facts were "far afield from the typical retaliatory arrest claim," we reserved judgment on the broader question presented and limited our holding to arrests that result from official policies of retaliation. Id., at ___ (slip op., at 10). In such cases, we held, probable cause does not categorically bar a plaintiff from suing the municipality. Id., at ___-___ (slip op., at 11-12). We now take up the question once again, this time in a more representative case.

A

     "[A]s a general matter the First Amendment prohibits government officials from subjecting an individual to retaliatory actions" for engaging in protected speech. Hartman v. Moore, 547 U. S. 250, 256 (2006). If an official takes adverse action against someone based on that forbidden motive, and "non-retaliatory grounds are in fact insufficient to provoke the adverse consequences," the injured person may generally seek relief by bringing a First Amendment claim. Ibid. (citing Crawford-El v. Britton, 523 U. S. 574, 593 (1998); Mt. Healthy City Bd. of Ed. v. Doyle, 429 U. S. 274, 283-284 (1977)).

     To prevail on such a claim, a plaintiff must establish a "causal connection" between the government defendant's "retaliatory animus" and the plaintiff's "subsequent injury." Hartman, 547 U. S., at 259. It is not enough to show that an official acted with a retaliatory motive and that the plaintiff was injured--the motive must cause the injury. Specifically, it must be a "but-for" cause, meaning that the adverse action against the plaintiff would not have been taken absent the retaliatory motive. Id., at 260 (recognizing that although it "may be dishonorable to act with an unconstitutional motive," an official's "action colored by some degree of bad motive does not amount to a constitutional tort if that action would have been taken anyway").

     For example, in Mt. Healthy, a teacher claimed that a school district refused to rehire him in retaliation for his protected speech. We held that even if the teacher's "protected conduct played a part, substantial or otherwise, in [the] decision not to rehire," he was not entitled to reinstatement "if the same decision would have been reached" absent his protected speech. 429 U. S., at 285. Regardless of the motives of the school district, we concluded that the First Amendment "principle at stake is sufficiently vindicated if such an employee is placed in no worse a position than if he had not engaged in the [protected speech]." Id., at 285-286.

     For a number of retaliation claims, establishing the causal connection between a defendant's animus and a plaintiff's injury is straightforward. Indeed, some of our cases in the public employment context "have simply taken the evidence of the motive and the discharge as sufficient for a circumstantial demonstration that the one caused the other," shifting the burden to the defendant to show he would have taken the challenged action even without the impermissible motive. Hartman, 547 U. S., at 260 (citing Mt. Healthy, 429 U. S., at 287; Arlington Heights v. Metropolitan Housing Development Corp., 429 U. S. 252, 270, n. 21 (1977)). But the consideration of causation is not so straightforward in other types of retaliation cases.

     In Hartman, for example, we addressed retaliatory prosecution cases, where "proving the link between the defendant's retaliatory animus and the plaintiff's injury . . . 'is usually more complex than it is in other retaliation cases.' " Lozman, 585 U. S., at ___ (slip op., at 8) (quoting Hartman, 547 U. S., at 261). Unlike most retaliation cases, in retaliatory prosecution cases the official with the malicious motive does not carry out the retaliatory action himself--the decision to bring charges is instead made by a prosecutor, who is generally immune from suit and whose decisions receive a presumption of regularity. Lozman, 585 U. S., at ___-___ (slip op., at 8-9). Thus, even when an officer's animus is clear, it does not necessarily show that the officer "induced the action of a prosecutor who would not have pressed charges otherwise." Hartman, 547 U. S., at 263.

     To account for this "problem of causation" in retaliatory prosecution claims, Hartman adopted the requirement that plaintiffs plead and prove the absence of probable cause for the underlying criminal charge. Ibid.; see id., at 265-266. As Hartman explained, that showing provides a "distinct body of highly valuable circumstantial evidence" that is "apt to prove or disprove" whether retaliatory animus actually caused the injury: "Demonstrating that there was no probable cause for the underlying criminal charge will tend to reinforce the retaliation evidence and show that retaliation was the but-for basis for instigating the prosecution, while establishing the existence of probable cause will suggest that prosecution would have occurred even without a retaliatory motive." Id., at 261. Requiring plaintiffs to plead and prove the absence of probable cause made sense, we reasoned, because the existence of probable cause will be at issue in "practically all" retaliatory prosecution cases, has "high probative force," and thus "can be made mandatory with little or no added cost." Id., at 265. Moreover, imposing that burden on plaintiffs was necessary to suspend the presumption of regularity underlying the prosecutor's charging decision--a presumption we "do not lightly discard." Id., at 263; see also id., at 265. Thus, Hartman requires plaintiffs in retaliatory prosecution cases to show more than the subjective animus of an officer and a subsequent injury; plaintiffs must also prove as a threshold matter that the decision to press charges was objectively unreasonable because it was not supported by probable cause.

B

     Officers Nieves and Weight argue that the same no-probable-cause requirement should apply to First Amendment retaliatory arrest claims. Their primary contention is that retaliatory arrest claims involve causal complexities akin to those we identified in Hartman, and thus warrant the same requirement that plaintiffs plead and prove the absence of probable cause. Brief for Petitioners 20-30.

     As a general matter, we agree. As we recognized in Reichle and reaffirmed in Lozman, retaliatory arrest claims face some of the same challenges we identified in Hartman: Like retaliatory prosecution cases, "retaliatory arrest cases also present a tenuous causal connection between the defendant's alleged animus and the plaintiff's injury." Reichle, 566 U. S., at 668. The causal inquiry is complex because protected speech is often a "wholly legitimate consideration" for officers when deciding whether to make an arrest. Ibid.; Lozman, 585 U. S., at ___ (slip op., at 9). Officers frequently must make "split-second judgments" when deciding whether to arrest, and the content and manner of a suspect's speech may convey vital information--for example, if he is "ready to cooperate" or rather "present[s] a continuing threat." Id., at ___ (slip op., at 9) (citing District of Columbia v. Wesby, 583 U. S. ___, ___ (2018) (slip op., at 10) ("suspect's untruthful and evasive answers to police questioning could support probable cause")). Indeed, that kind of assessment happened in this case. The officers testified that they perceived Bartlett to be a threat based on a combination of the content and tone of his speech, his combative posture, and his apparent intoxication.

     In addition, "[l]ike retaliatory prosecution cases, evidence of the presence or absence of probable cause for the arrest will be available in virtually every retaliatory arrest case." Reichle, 566 U. S., at 668. And because probable cause speaks to the objective reasonableness of an arrest, see Ashcroft v. al-Kidd, 563 U. S. 731, 736 (2011), its absence will--as in retaliatory prosecution cases--generally provide weighty evidence that the officer's animus caused the arrest, whereas the presence of probable cause will suggest the opposite.

     To be sure, Reichle and Lozman also recognized that the two claims give rise to complex causal inquiries for somewhat different reasons. Unlike retaliatory prosecution cases, retaliatory arrest cases do not implicate the presumption of prosecutorial regularity or necessarily involve multiple government actors (although this case did). Reichle, 566 U. S., at 668-669; Lozman, 585 U. S., at ___ (slip op., at 10). But regardless of the source of the causal complexity, the ultimate problem remains the same. For both claims, it is particularly difficult to determine whether the adverse government action was caused by the officer's malice or the plaintiff's potentially criminal conduct. See id., at ___ (slip op., at 9) (referring to "the complexity of proving (or disproving) causation" in retaliatory arrest cases). Because of the "close relationship" between the two claims, Reichle, 566 U. S., at 667, their related causal challenge should lead to the same solution: The plaintiff pressing a retaliatory arrest claim must plead and prove the absence of probable cause for the arrest.

     Bartlett, in defending the decision below, argues that the "causation in retaliatory-arrest cases is not inherently complex" because the "factfinder simply must determine whether the officer intended to punish the plaintiff for the plaintiff's protected speech." Brief for Respondent 36-37; see also post, at 5 (Sotomayor, J., dissenting). That approach fails to account for the fact that protected speech is often a legitimate consideration when deciding whether to make an arrest, and disregards the resulting causal complexity previously recognized by this Court. See Reichle, 566 U. S., at 668; Lozman, 585 U. S., at ___ (slip op., at 9).

     Bartlett's approach dismisses the need for any threshold showing, moving directly to consideration of the subjective intent of the officers. In the Fourth Amendment context, however, "we have almost uniformly rejected invitations to probe subjective intent." al-Kidd, 563 U. S., at 737; see also Kentucky v. King, 563 U. S. 452, 464 (2011) ("Legal tests based on reasonableness are generally objective, and this Court has long taken the view that evenhanded law enforcement is best achieved by the application of objective standards of conduct, rather than standards that depend upon the subjective state of mind of the officer." (internal quotation marks omitted)). Police officers conduct approximately 29,000 arrests every day--a dangerous task that requires making quick decisions in "circumstances that are tense, uncertain, and rapidly evolving." Graham v. Connor, 490 U. S. 386, 397 (1989). To ensure that officers may go about their work without undue apprehension of being sued, we generally review their conduct under objective standards of reasonableness. See Atwater v. Lago Vista, 532 U. S. 318, 351, and n. 22 (2001); Harlow v. Fitzgerald, 457 U. S. 800, 814-819 (1982). Thus, when reviewing an arrest, we ask "whether the circumstances, viewed objectively, justify [the challenged] action," and if so, conclude "that action was reasonable whatever the subjective intent motivating the relevant officials." al-Kidd, 563 U. S., at 736 (internal quotation marks omitted). A particular officer's state of mind is simply "irrelevant," and it provides "no basis for invalidating an arrest." Devenpeck v. Alford, 543 U. S. 146, 153, 155 (2004).

     Bartlett's purely subjective approach would undermine that precedent by allowing even doubtful retaliatory arrest suits to proceed based solely on allegations about an arresting officer's mental state. See Lozman, 585 U. S., at ___ (slip op., at 9). Because a state of mind is "easy to allege and hard to disprove," Crawford-El, 523 U. S., at 585, a subjective inquiry would threaten to set off "broad-ranging discovery" in which "there often is no clear end to the relevant evidence," Harlow, 457 U. S., at 817. As a result, policing certain events like an unruly protest would pose overwhelming litigation risks. Any inartful turn of phrase or perceived slight during a legitimate arrest could land an officer in years of litigation. Bartlett's standard would thus "dampen the ardor of all but the most resolute, or the most irresponsible, in the unflinching discharge of their duties." Gregoire v. Biddle, 177 F. 2d 579, 581 (CA2 1949) (Learned Hand, C. J.). It would also compromise evenhanded application of the law by making the constitutionality of an arrest "vary from place to place and from time to time" depending on the personal motives of individual officers. Devenpeck, 543 U. S., at 154. Yet another "predictable consequence" of such a rule is that officers would simply minimize their communication during arrests to avoid having their words scrutinized for hints of improper motive--a result that would leave everyone worse off. Id., at 155.

     Adopting Hartman's no-probable-cause rule in this closely related context addresses those familiar concerns. Absent such a showing, a retaliatory arrest claim fails. But if the plaintiff establishes the absence of probable cause, "then the Mt. Healthy test governs: The plaintiff must show that the retaliation was a substantial or motivating factor behind the [arrest], and, if that showing is made, the defendant can prevail only by showing that the [arrest] would have been initiated without respect to retaliation." Lozman, 585 U. S., at ___ (slip op., at 8) (citing Hartman, 547 U. S., at 265-266).¹

C

     Our conclusion is confirmed by the common law approach to similar tort claims. When defining the contours of a claim under §1983, we look to "common-law principles that were well settled at the time of its enactment." Ka- lina v. Fletcher, 522 U. S. 118, 123 (1997); Manuel v. Joliet, 580 U. S. ___, ___ (2017) (slip op., at 12) (common law principles "guide" the definition of claims under §1983).

     As the parties acknowledge, when §1983 was enacted in 1871, there was no common law tort for retaliatory arrest based on protected speech. See Brief for Petitioners 43; Brief for Respondent 20. We therefore turn to the common law torts that provide the "closest analogy" to retaliatory arrest claims. Heck v. Humphrey, 512 U. S. 477, 484 (1994). The parties dispute whether the better analog is false imprisonment or malicious prosecution. At common law, false imprisonment arose from a "detention without legal process," whereas malicious prosecution was marked "by wrongful institution of legal process." Wallace v. Kato, 549 U. S. 384, 389-390 (2007).² Here, both claims suggest the same result: The presence of probable cause should generally defeat a First Amendment retaliatory arrest claim. See generally Lozman, 585 U. S., at ___-___ (slip op., at 4-6) (Thomas, J., dissenting).

     Malicious prosecution required the plaintiff to show that the criminal charge against him "was unfounded, and that it was made without reasonable or probable cause, and that the defendant in making or instigating it was actuated by malice." Wheeler v. Nesbitt, 24 How. 544, 549-550 (1861); see also Restatement of Torts §653 (1938). It has long been "settled law" that malicious prosecution requires proving "the want of probable cause," and Bartlett does not argue otherwise. Brown v. Selfridge, 224 U. S. 189, 191 (1912); see also Wheeler, 24 How., at 550 (noting that "[w]ant of reasonable and probable cause" is an "element in the action for a malicious criminal prosecution").

     For claims of false imprisonment, the presence of probable cause was generally a complete defense for peace officers. See T. Cooley, Law of Torts 175 (1880); 1 F. Hilliard, The Law of Torts or Private Wrongs 207-208, and n. (a) (1859). In such cases, arresting officers were protected from liability if the arrest was "privileged." At common law, peace officers were privileged to make warrantless arrests based on probable cause of the commission of a felony or certain misdemeanors. See Restatement of Torts §§118, 119, 121 (1934); see also Cooley, Law of Torts, at 175-176 (stating that peace officers who make arrests based on probable cause "will be excused, even though it appear afterwards that in fact no felony had been committed"); see generally Atwater, 532 U. S., at 340-345 (reviewing the history of warrantless arrests for misdemeanors). Although the exact scope of the privilege varied somewhat depending on the jurisdiction, the consistent rule was that officers were not liable for arrests they were privileged to make based on probable cause.

D

     Although probable cause should generally defeat a retaliatory arrest claim, a narrow qualification is warranted for circumstances where officers have probable cause to make arrests, but typically exercise their discretion not to do so. In such cases, an unyielding requirement to show the absence of probable cause could pose "a risk that some police officers may exploit the arrest power as a means of suppressing speech." Lozman, 585 U. S., at ___ (slip op., at 10).

     When §1983 was adopted, officers were generally privileged to make warrantless arrests for misdemeanors only in limited circumstances. See Restatement of Torts §121, Comments e, h, at 262-263. Today, however, "statutes in all 50 States and the District of Columbia permit warrantless misdemeanor arrests" in a much wider range of situations--often whenever officers have probable cause for "even a very minor criminal offense." Atwater, 532 U. S., at 344-345, 354; see id., at 355-360 (listing state statutes).

     For example, at many intersections, jaywalking is endemic but rarely results in arrest. If an individual who has been vocally complaining about police conduct is arrested for jaywalking at such an intersection, it would seem insufficiently protective of First Amendment rights to dismiss the individual's retaliatory arrest claim on the ground that there was undoubted probable cause for the arrest. In such a case, because probable cause does little to prove or disprove the causal connection between animus and injury, applying Hartman's rule would come at the expense of Hartman's logic.

     For those reasons, we conclude that the no-probable-cause requirement should not apply when a plaintiff presents objective evidence that he was arrested when otherwise similarly situated individuals not engaged in the same sort of protected speech had not been. Cf. United States v. Armstrong, 517 U. S. 456, 465 (1996). That showing addresses Hartman's causal concern by helping to establish that "non-retaliatory grounds [we]re in fact insufficient to provoke the adverse consequences." 547 U. S., at 256. And like a probable cause analysis, it provides an objective inquiry that avoids the significant problems that would arise from reviewing police conduct under a purely subjective standard. Because this inquiry is objective, the statements and motivations of the particular arresting officer are "irrelevant" at this stage. Devenpeck, 543 U. S., at 153. After making the required showing, the plaintiff's claim may proceed in the same manner as claims where the plaintiff has met the threshold showing of the absence of probable cause. See Lozman, 585 U. S., at ___ (slip op., at 8).

*  *  *

     In light of the foregoing, Bartlett's retaliation claim cannot survive summary judgment. As an initial matter, the record contains insufficient evidence of retaliation on the part of Trooper Weight. The only evidence of retaliatory animus identified by the Ninth Circuit was Bartlett's affidavit stating that Sergeant Nieves said "bet you wish you would have talked to me now." 712 Fed. Appx., at 616. But that allegation about Nieves says nothing about what motivated Weight, who had no knowledge of Bartlett's prior run-in with Nieves. Cf. Lozman, 585 U. S., at ___ (slip op., at 10) (plaintiff "likely could not have maintained a retaliation claim against the arresting officer" when there was "no showing that the officer had any knowledge of [the plaintiff's] prior speech").

     In any event, Bartlett's claim against both officers cannot succeed because they had probable cause to arrest him. As the Court of Appeals explained:

"When Sergeant Nieves initiated Bartlett's arrest, he knew that Bartlett had been drinking, and he observed Bartlett speaking in a loud voice and standing close to Trooper Weight. He also saw Trooper Weight push Bartlett back. . . . [T]he test is whether the information the officer had at the time of making the arrest gave rise to probable cause. We agree with the district court that it did; a reasonable officer in Sergeant Nieves's position could have concluded that Bartlett stood close to Trooper Weight and spoke loudly in order to challenge him, provoking Trooper Weight to push him back." 712 Fed. Appx., at 615 (citations and internal quotation marks omitted).

     Because there was probable cause to arrest Bartlett, his retaliatory arrest claim fails as a matter of law. Accordingly, the judgment of the United States Court of Appeals for the Ninth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Opinion of Thomas, J.

 587 U. S. ____ (2019)

No. 17-1174

     Justice Thomas, concurring in part and concurring in the judgment.

     When 42 U. S. C. §1983 was enacted, "the common law recognized probable cause as an important element for ensuring that arrest-based torts did not unduly interfere with the objectives of law enforcement." Lozman v. Riviera Beach, 585 U. S. ___, ___ (2018) (Thomas, J., dissenting) (slip op., at 6). Applying that principle resolves this case: "[P]laintiffs bringing a First Amendment retaliatory-arrest claim under §1983 should have to plead and prove a lack of probable cause." Ibid. The Court acknowledges as much, ante, at 12-13, and I join the portions of the Court's opinion adopting that rule.¹ I do not join Part II-D, however, because I do not agree that "a narrow qualification is warranted for circumstances where officers have probable cause to make arrests, but typically exercise their discretion not to do so." Ante, at 13-14. That qualification has no basis in either the common law or our First Amendment precedents.

     As the Court explains, "[w]hen defining the contours of a claim under §1983, we look to 'common-law principles that were well settled at the time of its enactment.' " Ante, at 12. Because no common-law tort for retaliatory arrest in violation of the freedom of speech existed when §1983 was enacted, we "look to the common-law torts that 'provid[e] the closest analogy' to this claim." Lozman, 585 U. S., at ___ (opinion of Thomas, J.) (slip op., at 4). Here, those torts are false imprisonment, malicious arrest, and malicious prosecution. Ibid.

     The existence of probable cause generally excused an officer from liability for these three torts, without regard to the treatment of similarly situated individuals. For instance, a constable who made an arrest "on reasonable grounds of belief" that a felony had been committed was "excused" from liability for false imprisonment. T. Cooley, Law of Torts 175 (1879) (Cooley); Lozman, supra, at ___-___ (opinion of Thomas, J.) (slip op., at 4-5). And the absence of probable cause was central to both malicious arrest and malicious prosecution. Cooley 180-181; Lozman, supra, at ___ (opinion of Thomas, J.) (slip op., at 5). As the Court puts it, "the consistent rule was that officers were not liable for arrests they were privileged to make based on probable cause." Ante, at 13.

     Rather than adhere to this rule, the majority carves out an exception to the no-probable-cause requirement for plaintiffs who "presen[t] objective evidence" that they were "arrested when otherwise similarly situated individuals not engaged in the same sort of protected speech had not been." Ante, at 14. The common law provides no support for this exception. Indeed, the majority cites not a single common-law case that supports imposing liability based on an officer's treatment of similarly situated individuals. The majority instead suggests that its exception responds to the fact that States today " 'permit warrantless misdemeanor arrests' " for many " 'minor criminal offense[s],' " whereas "[w]hen §1983 was adopted, officers were generally privileged to make warrantless arrests for misdemeanors only in limited circumstances." Ibid. But discomfort with the number of warrantless arrests that are privileged today is an issue for state legislatures, not a license for this Court to fashion an exception to a previously "consistent rule." Ante, at 13.

     The majority's exception is also untethered from our First Amendment precedents. In Hartman v. Moore, 547 U. S. 250 (2006), we expressly declined to create any exceptions to the rule that a plaintiff alleging retaliatory prosecution in violation of the First Amendment must plead and prove the absence of probable cause. See id., at 264-266, and n. 10. The majority today imports its "qualification" from our jurisprudence on selective-prosecution claims. Ante, at 14, 15 (citing United States v. Armstrong, 517 U. S. 456, 465 (1996)). But "[t]he requirements for a selective-prosecution claim draw on 'ordinary equal protection standards,' " not the First Amendment. Id., at 465. That jurisprudence therefore is not relevant here. Cf. Whren v. United States, 517 U. S. 806, 813 (1996) ("[T]he constitutional basis for objecting to intentionally discriminatory application of laws is the Equal Protection Clause . . . ").

     With no guidance from the common law or relevant precedents, the majority crafts its exception as a matter of policy. But this "narrow" qualification threatens to derail our retaliation jurisprudence in several ways. For one, although the majority's stated concern is with " 'warrantless misdemeanor arrests' " for " 'very minor' " offenses like "jaywalking," ante, at 14, its exception apparently applies to all offenses, including serious felonies. This overbroad exception thus is likely to encourage protracted litigation about which individuals are "similarly situated," ibid., while doing little to vindicate First Amendment rights. Moreover, the majority's rule risks chilling law enforcement officers from making arrests for fear of liability, thus flouting the reasoning behind the emphasis on probable cause in arrest-based torts at common law. Lozman, supra, at ___-___ (opinion of Thomas, J.) (slip op., at 4-5). In short, the majority's exception lacks the support of history, precedent, and sound policy.

*  *  *

     The requirement that plaintiffs bringing First Amendment retaliatory-arrest claims plead and prove the absence of probable cause is supported by the common law and our First Amendment precedents. The majority's new exception has no basis in either. Accordingly, I join all but Part II-D of the majority opinion.

Opinion of Gorsuch, J.

 587 U. S. ____ (2019)

No. 17-1174

     Justice Gorsuch, concurring in part and dissenting in part.

     No one agrees who started the trouble at the 2014 Alaska Arctic Man Festival, but everyone agrees two troopers ended it by arresting Russell Bartlett for harassing one of them. And it's that arrest that led to this lawsuit. Mr. Bartlett alleges that the real reason the troopers arrested him had nothing to do with harassment and everything to do with his decision to exercise his First Amendment free speech rights by voicing his opinions to the troopers. Mr. Bartlett contends that the officers' retaliation against his First Amendment protected speech entitles him to damages under 42 U. S. C. §1983. For their part, the troopers insist that the fact they had probable cause to arrest Mr. Bartlett for some crime, or really any crime, should be enough to shield them from liability.

     The parties approach their dispute from some common ground. Both sides accept that an officer violates the First Amendment when he arrests an individual in retaliation for his protected speech. They seem to agree, too, that the presence of probable cause does not undo that violation or erase its significance. And for good reason. History shows that governments sometimes seek to regulate our lives finely, acutely, thoroughly, and exhaustively. In our own time and place, criminal laws have grown so exuberantly and come to cover so much previously innocent conduct that almost anyone can be arrested for something. If the state could use these laws not for their intended purposes but to silence those who voice unpopular ideas, little would be left of our First Amendment liberties, and little would separate us from the tyrannies of the past or the malignant fiefdoms of our own age. The freedom to speak without risking arrest is "one of the principal characteristics by which we distinguish a free nation." Houston v. Hill, 482 U. S. 451, 463 (1987).

     So if probable cause can't erase a First Amendment violation, the question becomes whether its presence at least forecloses a civil claim for damages as a statutory matter under §1983. But look at that statute as long as you like and you will find no reference to the presence or absence of probable cause as a precondition or defense to any suit. Instead, the statute imposes liability on anyone who, under color of state law, subjects another person "to the deprivation of any rights, privileges, or immunities secured by the Constitution." Maybe it would be good policy to graft a no-probable-cause requirement onto the statute, as the officers insist; or maybe not. Either way, that's an appeal better directed to Congress than to this Court. Our job isn't to write or revise legislative policy but to apply it faithfully.

     Admittedly, though, that's not quite the end of the statutory story. Courts often assume that Congress adopts statutes against the backdrop of the common law. And, for this reason, we generally read §1983's terms "in harmony with general principles of tort immunities and defenses" that existed at the time of the statute's adoption. Imbler v. Pachtman, 424 U. S. 409, 418 (1976). As the officers before us are quick to point out, too, law enforcement agents who made a lawful arrest at the time of §1983's adoption couldn't be held liable at common law for the tort of false arrest or false imprisonment. Of course, at common law a police officer often needed a warrant to execute a lawful arrest. But today warrantless arrests are often both authorized by state law and permitted by the Constitution (as this Court has interpreted it), so long as the officer possesses probable cause to believe a crime has been committed. And, given this development, you might wonder if the presence of probable cause should be enough to foreclose any First Amendment claim arising out of an arrest.

     But that much doesn't follow. As the officers' own reasoning exposes, the point of the common law tort of false arrest or false imprisonment was to remedy arrests and imprisonments effected without lawful authority. See Director General of Railroads v. Kastenbaum, 263 U. S. 25, 27 (1923); accord, Brief for United States as Amicus Cu- riae 9. So maybe probable cause should be enough today to defeat claims for false arrest or false imprisonment, given that arrests today are usually legally authorized if supported by probable cause. But that doesn't mean probable cause is also enough to defeat a First Amendment retaliatory arrest claim. The point of this kind of claim isn't to guard against officers who lack lawful authority to make an arrest. Rather, it's to guard against officers who abuse their authority by making an otherwise lawful arrest for an unconstitutional reason.

     Here's another way to look at it. The common law tort of false arrest translates more or less into a Fourth Amendment claim. That's because our precedent considers a warrantless arrest unsupported by probable cause--the sort that gave rise to a false arrest claim at common law--to be an unreasonable seizure in violation of the Fourth Amendment. See Manuel v. Joliet, 580 U. S. ___, ___ (2017) (slip op., at 10). But the First Amendment operates independently of the Fourth and provides different protections. It seeks not to ensure lawful authority to arrest but to protect the freedom of speech.

     Here's a way to test the point, too. Everyone accepts that a detention based on race, even one otherwise authorized by law, violates the Fourteenth Amendment's Equal Protection Clause. In Yick Wo v. Hopkins, 118 U. S. 356 (1886), for example, San Francisco jailed many Chinese immigrants for operating laundries without permits but took no action against white persons guilty of the same infraction. Even if probable cause existed to believe the Chinese immigrants had broken a valid law--even if they had in fact violated the law--this Court held that the city's discriminatory enforcement violated the Fourteenth Amendment. Id., at 373-374; see also Whren v. United States, 517 U. S. 806, 813 (1996). Following our lead, the courts of appeals have recognized that §1983 plaintiffs alleging racially selective arrests in violation of the Fourteenth Amendment don't have to show a lack of probable cause, even though they might have to show a lack of probable cause to establish a violation of the Fourth Amendment: "[S]imply because a practice passes muster under the Fourth Amendment (arrest based on probable cause) does not mean that unequal treatment with respect to that practice is consistent with equal protection." Hedgepeth v. Washington Metropolitan Area Transit Auth., 386 F. 3d 1148, 1156 (CADC 2004); see also Gibson v. Superintendent of N. J. Dept. of Law and Public Safety-Div. of State Police, 411 F. 3d 427, 440 (CA3 2005); Marshall v. Columbia Lea Regional Hospital, 345 F. 3d 1157, 1166 (CA10 2003); Vakilian v. Shaw, 335 F. 3d 509, 521 (CA6 2003); Johnson v. Crooks, 326 F. 3d 995, 999-1000 (CA8 2003); Holland v. Portland, 102 F. 3d 6, 11 (CA1 1996).

     I can think of no sound reason why the same shouldn't hold true here. Like a Fourteenth Amendment selective arrest claim, a First Amendment retaliatory arrest claim serves a different purpose than a Fourth Amendment unreasonable arrest claim, and that purpose does not depend on the presence or absence of probable cause. We thus have no legitimate basis for engrafting a no-probable-cause requirement onto a First Amendment retaliatory arrest claim.

     But while it would be a mistake to think the absence of probable cause is an essential element of a First Amendment retaliatory arrest claim under §1983--or that the presence of probable cause is an absolute defense to such a claim--I acknowledge that it may also be a mistake to assume probable cause is entirely irrelevant to the analysis. It seems to me that probable cause to arrest could still bear on the claim's viability in at least two ways that warrant further exploration in future cases.

     First, consider causation. To show an arrest violated the First Amendment, everyone agrees a plaintiff must prove the officer would not have arrested him but for his protected speech. And if the only offense for which probable cause to arrest existed was a minor infraction of the sort that wouldn't normally trigger an arrest in the circumstances--or if the officer couldn't identify a crime for which probable cause existed until well after the arrest--then causation might be a question for the jury. By contrast, if the officer had probable cause at the time of the arrest to think the plaintiff committed a serious crime of the sort that would nearly always trigger an arrest regardless of speech, then (absent extraordinary circumstances) it's hard to see how a reasonable jury might find that the plaintiff's speech caused the arrest. In cases like that, it would seem that officers often will be entitled to dismissal on the pleadings or summary judgment.

     In the name of causation concerns, the officers ask us to go further still and hold that a plaintiff can never prove protected speech caused his arrest without first showing that the officers lacked probable cause to make an arrest. But that absolute rule doesn't wash with common experience. No one doubts that officers regularly choose against making arrests, especially for minor crimes, even when they possess probable cause. So the presence of probable cause does not necessarily negate the possibility that an arrest was caused by unlawful First Amendment retaliation.

     If logic doesn't support their argument, the officers reply, at least precedent does. They point to Hartman v. Moore, 547 U. S. 250 (2006), where the plaintiff alleged that a law enforcement officer retaliated against him for his speech by convincing a prosecutor to bring criminal charges against him. This Court held the claim presented a "distinct problem of causation" that could be overcome only with proof that the prosecutor lacked probable cause for the charges. Id., at 263. And as a matter of precedent, the officers argue, what was true there should hold true here.

     But Hartman does not demand nearly as much of us as the officers suggest. Hartman justified its rule by explaining that the "causal connection required" in that case wasn't just between the officer's retaliatory intent and his own injurious action. Instead, it was between the officer's intent and the injurious action of the prosecutor, whose charging decision was subject to a "presumption of regularity." Id., at 262-263. In that particular context, Hartman said, proof that the prosecutor lacked probable cause was necessary to "bridge the gap between the [officer's] motive and the prosecutor's action." Id., at 263. But Hartman made equally clear that its reasoning did not extend to "ordinary retaliation claims, where the government agent allegedly harboring the animus is also the individual allegedly taking the adverse action." Id., at 259-260. And that describes a retaliatory arrest claim like Mr. Bartlett's to a tee. So while probable cause may wind up defeating causation in some retaliatory arrest cases, nothing in our precedent (let alone logic) suggests that causation is always unprovable just because the officer had probable cause to arrest.

     Second, our precedent suggests the possibility that probable cause may play a role in light of the separation of powers and federalism. In United States v. Armstrong, 517 U. S. 456 (1996), a plaintiff alleged that he was prosecuted because of his race. This Court responded by acknowledging that racially selective prosecutions can violate the Equal Protection Clause even when the prosecutor possesses probable cause to believe a crime has been committed. But because the decision whether to institute criminal charges is one our Constitution vests in state and federal executive officials, not judges, this Court also held that, to respect the separation of powers and federalism, a plaintiff must present " 'clear evidence' " of discrimination when a federal or state official possesses probable cause to support his prosecution. Id., at 465. Explaining the content of the clear evidence requirement, this Court indicated that a plaintiff generally must produce evidence that the prosecutor failed to charge other similarly situated persons. At the same time, however, the Court also suggested that equally clear evidence in the form of " 'direct admissions by prosecutors of discriminatory purpose' " might be enough to allow a claim to proceed. Id., at 469, n. 3 (alterations omitted).

     Though this case involves a retaliatory arrest claim rather than a selective prosecution claim, it's at least an open question whether the concerns that drove this Court's decision in Armstrong may be in play here. No one before us argues that Armstrong was wrongly decided. And the Court today seems to indicate that something like Armstrong's standard might govern a retaliatory arrest claim when probable cause exists to support an arrest. See ante, at 14 (citing Armstrong). Some courts of appeals, too, have already applied Armstrong to claims alleging selective arrest under the Fourteenth Amendment. See, e.g., United States v. Mason, 774 F. 3d 824, 829-830 (CA4 2014); United States v. Alcaraz-Arellano, 441 F. 3d 1252, 1264 (CA10 2006); Johnson, 326 F. 3d, at 1000.

     At the same time, enough questions remain about Armstrong's potential application that I hesitate to speak definitively about it today. Some courts of appeals have argued that Armstrong should not extend, at least without qualification, beyond prosecutorial decisions to arrests by police. These courts have suggested that the presumptions of regularity and immunity that usually attach to official prosecutorial decisions do not apply equally in the less formal setting of police arrests. They've reasoned, too, that comparative data about similarly situated individuals may be less readily available for arrests than for prosecutorial decisions, and that other kinds of evidence--such as an officer's questions and comments to the defendant--may be equally if not more probative in the arrest context. See, e.g., United States v. Sellers, 906 F. 3d 848, 856 (CA9 2018); United States v. Washington, 869 F. 3d 193, 219 (CA3 2017); United States v. Davis, 793 F. 3d 712, 720-721 (CA7 2015); Marshall, 345 F. 3d, at 1168. Importantly, we did not grant certiorari to resolve exactly how Armstrong might apply to retaliatory arrest claims. Nor did the briefing before us explore the competing arguments in this circuit split. And given all this, I believe it would be rash for us to do more at this point than acknowledge the possibility of Armstrong's application.

     Dissenting, Justice Sotomayor reads the majority opinion as adopting a rigid rule (more rigid, in fact, than Armstrong's) that First Amendment retaliatory arrest plaintiffs who can't prove the absence of probable cause must produce "comparison-based evidence" in every case. Post, at 4. But I do not understand the majority as going that far. The only citation the majority offers in support of its new standard is Armstrong, which expressly left open the possibility that other kinds of evidence, such as admissions, might be enough to allow a claim to proceed. Given that, I retain hope that lower courts will apply today's decision "commonsensically," post, at 12, and with sensitivity to the competing arguments about whether and how Armstrong might apply in the arrest setting.

     For today, I believe it is enough to resolve the question on which we did grant certiorari--whether "probable cause defeats . . . a First Amendment retaliatory-arrest claim under §1983." Pet. for Cert. i. I would hold, as the majority does, that the absence of probable cause is not an absolute requirement of such a claim and its presence is not an absolute defense. At the same time, I would also acknowledge that this does not mean the presence of probable cause is categorically irrelevant: It may bear on causation, and it may play a role under Armstrong. But rather than attempt to sort out precisely when and how probable cause plays a role in First Amendment claims, I would reserve decision on those questions until they are properly presented to this Court and we can address them with the benefit of full adversarial testing.

Opinion of Ginsburg, J.

 587 U. S. ____ (2019)

No. 17-1174

     Justice Ginsburg, concurring in the judgment in part and dissenting in part.

     Arrest authority, as several decisions indicate, can be abused to disrupt the exercise of First Amendment speech and press rights. See, e.g., Lacey v. Maricopa County, 693 F. 3d 896, 907-910 (CA9 2012) (en banc) (newspaper publishers alleged they were arrested in nighttime raid after publishing story on law enforcement's investigation of the newspaper); Roper v. New York, 2017 WL 2483813, *1 (SDNY, June 7, 2017) (photographers documenting Black Lives Matter protest alleged they were arrested for standing in street and failing to use crosswalk; sidewalks and crosswalks were blocked by police); Morse v. San Francisco Bay Area Rapid Transit Dist. (BART), 2014 WL 572352, *1-*7 (ND Cal., Feb. 11, 2014) (only journalist arrested at protest was journalist critical of Bay Area Rapid Transit Police). Given the array of laws proscribing, e.g., breach of the peace, disorderly conduct, obstructing public ways, failure to comply with a peace officer's instruction, and loitering, police may justify an arrest as based on probable cause when the arrest was in fact prompted by a retaliatory motive. If failure to show lack of probable cause defeats an action under 42 U. S. C. §1983, only entirely baseless arrests will be checked. I remain of the view that the Court's decision in Mt. Healthy City Bd. of Ed. v. Doyle, 429 U. S. 274 (1977), strikes the right balance: The plaintiff bears the burden of demonstrating that unconstitutional animus was a motivating factor for an adverse action; the burden then shifts to the defendant to demonstrate that, even without any impetus to retaliate, the defendant would have taken the action complained of. See Hartman v. Moore, 547 U. S. 250, 266-267 (2006) (Ginsburg, J., dissenting).

     In this case, I would reverse the Ninth Circuit's judgment as to Trooper Weight. As the Court points out, the record is bereft of evidence of retaliation on Weight's part. See ante, at 15. As to Sergeant Nieves, there is some evidence of animus in Nieves' statement, "bet you wish you would have talked to me now," App. 376, but perhaps not enough to survive summary judgment. Cf. Higginbotham v. Sylvester, 741 Fed. Appx. 28, 31 (CA2 2018) ("[A] reasonable factfinder could not find that [the plaintiff's] exercise of his First Amendment right . . . was the 'but-for' cause of his arrest."). In any event, I would not use this thin case to state a rule that will leave press members and others exercising First Amendment rights with little protection against police suppression of their speech.

Sotomayor, J., dissenting

 587 U. S. ____ (2019)

No. 17-1174

     Justice Sotomayor, dissenting.

     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 14. 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 ante, at 1-2 (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 2-4. I agree that Hartman v. Moore, 547 U. S. 250 (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 6. And I agree that--while probable 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 5-6; see also ante, at 1 (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 1-2 (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 (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; see also Lozman v. Riviera Beach, 585 U. S. ___, ___ (2018) (slip op., at 7).

     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, 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; see ante, at 1-2 (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 14. 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 ante, at 1 (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 13, 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 8-9, 12-13. For the reasons Justice Gorsuch explains, that reasoning is not sound. See ante, at 2-4, 6. 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 8; some plaintiffs who raise dubious challenges to lawful arrests may evade early dismissal under Mt. Healthy, see ante, at 10-11. 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 (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 (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.

     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 Even 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. 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 ___ (slip op., at 12); 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 (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 11, 14.

     This analogy is misguided, and the Court has rightly disavowed it before. In Whren v. United States, 517 U. S. 806 (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; accord, ante, at 4 (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. 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, when it comes to First Amendment freedom of speech, "the government's reason" is often "what counts," see Heffernan v. City of Paterson, 578 U. S. ___, ___ (2016) (slip op., at 6). 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 motives 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 15, 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. ___, ___-___ (2018) (slip op., at 13-14); Cooper v. Harris, 581 U. S. ___, ___-___ (2017) (slip op., at 10-12); Foster v. Chatman, 578 U. S. ___, ___-___ (2016) (slip op., at 10-25); Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U. S. 520, 540-542 (1993); Arlington Heights v. Metropolitan Housing Development Corp., 429 U. S. 252, 270 (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 11 (quoting Crawford-El, 523 U. S., at 585). Such concerns, whatever their merits, are insufficient reason "to change the burden of proof for an entire category of claims." Id., at 594. 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 majority 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 (CD 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?

     Worse, because the majority disclaims reliance on "statements and motivations" for its threshold inquiry, ante, at 15, 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 (WD 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 (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 14.⁷ 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 borrowing a requirement to adduce "clear evidence" of prohibited purpose from our cases concerning equal-protection-based selective-prosecution claims. See ante, at 7-8 (citing United States v. Armstrong, 517 U. S. 456 (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, 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 (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 (1927) (Brandeis, J., concurring). Because the majority shortchanges that hard-earned wisdom in the name of marginal convenience, I respectfully dissent.

FOOTNOTES

 Justice Sotomayor would have us extend Mt. Healthy and rely on that "tried and true" approach as the exclusive standard in the retaliatory arrest context. See post, at 1-5, 14 (dissenting opinion). But not even respondent Bartlett argues for such a rule. And since our decisions in Hartman and Reichle, no court of appeals has applied that approach in retaliatory arrest cases of this sort. Justice Sotomayor criticizes the Court for spending "[m]uch of its opinion . . . analogizing to Hartman," post, at 4, but of course Hartman is our precedent most directly on point. To the extent retaliatory arrest cases raise concerns distinct from that precedent, we have departed from Hartman to afford greater First Amendment protection. See infra, at 13-15.

 For our purposes, we need not distinguish between the torts of false imprisonment and false arrest, which are "virtually synonymous." 35 C. J. S., False Imprisonment §2, p. 522 (2009); see also Wallace, 549 U. S., at 388-389.

FOOTNOTES

 Because no party questions whether §1983 claims for retaliatory arrests under the First Amendment are actionable, I assume that §1983 permits such claims. See Lozman v. Riviera Beach, 585 U. S. ___, ___, n. 2 (2018) (Thomas, J., dissenting) (slip op., at 3, n. 2).

FOOTNOTES

 The majority implies that the Ninth Circuit does not apply Mt. Healthy. Ante, at 12, n. 1 ("since . . . Hartman and Reichle, no court of appeals has applied [the Mt. Healthy] approach"). That is not readily apparent. Because Hartman's no-probable-cause requirement does not apply to retaliatory police action in the Ninth Circuit, such claims are handled as " 'ordinary' retaliation claim[s]," Skoog v. County of Clackamas, 469 F. 3d 1221, 1234 (2006), which in the Ninth Circuit (as elsewhere) means that retaliatory motive must be the "but-for cause of the defendant's action," id., at 1232. That but-for causation requirement for retaliation claims derives from Mt. Healthy. See Hartman v. Moore, 547 U. S. 250, 260 (2006); Crawford-El v. Britton, 523 U. S. 574, 593 (1998); see also Lacey v. Maricopa County, 693 F. 3d 896, 916-917 (CA9 2012) (en banc) (retaliatory arrest plaintiff must show that deterrence of speech "was a substantial or motivating factor" and also "ultimately" be able to show " 'but-for causation' " (quoting Hartman's discussion of Mt. Healthy)).

In any event, the majority's criticism is a red herring. There is nothing novel about applying Mt. Healthy in the retaliatory arrest context. E.g., Lozman v. Riviera Beach, 585 U. S. ___, ___-___ (2018) (slip op., at 12-13). The same cannot be said of the test concocted by the majority.

 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 United States 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 15. 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 15, 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 14, 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. 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 7. 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 (1961). It is an emphatic "extension of federal power" into state affairs, id., at 182, 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; see also id., at 174, and n. 10.