Michael Rodgers v. Bill Bryant

942 F.3d 451

• Court: Court of Appeals for the Eighth Circuit
• Decided: November 6, 2019
• Docket: 17-3219
• Status: Published

Opinion

United States Court of Appeals For the Eighth Circuit ___________________________

No. 17-3219 ___________________________

Michael Andrew Rodgers; Glynn Dilbeck

Plaintiffs - Appellees

v.

Bill Bryant, Colonel, in His Official Capacity as Director of the Arkansas State Police

Defendant - Appellant ____________

Appeal from United States District Court for the Eastern District of Arkansas - Little Rock ____________

Submitted: September 27, 2018 Filed: November 6, 2019 ____________

Before SMITH, Chief Judge, MELLOY and STRAS, Circuit Judges. ____________

MELLOY, Circuit Judge.

Michael Andrew Rodgers and Glynn Dilbeck challenge an Arkansas anti- loitering law that bans begging in a manner that is harassing, causes alarm, or impedes traffic. The district court1 granted a statewide preliminary injunction preventing Arkansas from enforcing the ban while Rodgers and Dilbeck pursue their claim that the law violates the First Amendment. Having jurisdiction under 28 U.S.C. § 1292(a)(1), we affirm.

I.

Rodgers and Dilbeck, who have been begging in Arkansas for a long time, claim that Arkansas’s anti-loitering law violates their free-speech rights. According to the law:

(a) A person commits the offense of loitering if he or she:

....

(3) Lingers or remains on a sidewalk, roadway, or public right-of- way, in a public parking lot or public transportation vehicle or facility, or on private property, for the purpose of asking for anything as charity or a gift:

(A) In a harassing or threatening manner;

(B) In a way likely to cause alarm to the other person; or

(C) Under circumstances that create a traffic hazard or impediment[.]

Ark. Code Ann. § 5-71-213 (2017) (amended 2019). Violations are punishable by up to 30 days in jail and a fine of up to $500. See id. §§ 5-4-201(b)(3), -4-401(b)(3), -71-213(e).

1 The Honorable Billy Roy Wilson, United States District Judge for the Eastern District of Arkansas. -2- Rodgers and Dilbeck, who were arrested or cited under a prior version of Arkansas’s anti-loitering law, see id. § 5-71-213(a)(3) (1995) (making it a misdemeanor to “[l]inger[] or remain[] in a public place or on the premises of another for the purpose of begging”), allege that they have changed their behavior because they fear arrest under the new law. Among other things, they changed the locations where they beg and stopped using signs in some areas; Dilbeck even claims to have moved to Tennessee in an alleged effort to avoid application of the law. Absent the anti-loitering law, Rodgers and Dilbeck say they would beg “openly and without fear” in Arkansas.

Rodgers and Dilbeck brought a First Amendment challenge to the anti- loitering law and requested a preliminary injunction. 2 The district court granted a statewide preliminary injunction based on its conclusion that the law, which it classified as a content-based restriction on speech, does not serve a compelling state interest and is “plainly unconstitutional.” Arkansas filed this interlocutory appeal challenging the preliminary injunction. Arkansas argues that Rodgers and Dilbeck lack standing to bring their constitutional challenge, that the anti-loitering law is constitutional, and, in the alternative, that the district court should have entered an injunction preventing enforcement of the law only against Rodgers and Dilbeck.

II.

To have standing to challenge the Arkansas law, Rodgers and Dilbeck must establish (1) an injury in fact; (2) a causal connection between the injury and the law; and (3) that a favorable decision will “likely” redress the injury. See Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61 (1992). We review questions of standing de novo. In re SuperValu, Inc., 870 F.3d 763, 768 (8th Cir. 2017). At this stage of the litigation, we assume that the allegations in the complaint are true and view the record in the light most favorable to Rodgers and Dilbeck. See id.

2 Rodgers and Dilbeck also claimed that the law is void for vagueness under the Fourteenth Amendment. The district court did not reach this issue, and neither do we. -3- The parties disagree about whether Rodgers and Dilbeck have suffered an injury in fact. After all, neither specifically claims to have violated the law, let alone to have been prosecuted under it. In the First Amendment context, however, “[s]elf- censorship can itself constitute injury in fact” if the “plaintiff[s] show[] an intention to engage in . . . conduct arguably affected with a constitutional interest” and “there exists a credible threat of prosecution.” 281 Care Comm. v. Arneson, 638 F.3d 621, 627 (8th Cir. 2011) (internal quotation marks and citation omitted). In other words, a law’s “chilling effect” can create standing. Id. at 627–28; see also Steffel v. Thompson, 415 U.S. 452, 459 (1974) (“[I]t is not necessary that petitioner first expose himself to actual arrest or prosecution to be entitled to challenge a statute that he claims deters the exercise of his constitutional rights.”).

Rodgers and Dilbeck have adequately alleged that their speech is, and will continue to be, chilled by Arkansas’s anti-loitering law. They claim to have changed when, where, and how they beg due to fear of being criminally prosecuted. See 281 Care Comm., 638 F.3d at 627. They also allege that their begging has caused people to react in fear or alarm and has even occasionally slowed traffic. Viewing the record in the light most favorable to them, SuperValu, 870 F.3d at 768, Rodgers and Dilbeck have established a credible threat of prosecution that gives them standing to challenge the law.

Arkansas claims that it would never enforce its anti-loitering law against “polite” and “courteous” beggars like Rodgers and Dilbeck. Even if true now, however, Arkansas’s in-court assurances do not rule out the possibility that it will change its mind and enforce the law more aggressively in the future. See United Food & Commercial Workers Int’l Union v. IBP, Inc., 857 F.2d 422, 429 (8th Cir. 1988) (rejecting the argument that the plaintiffs lacked standing because the defendants did not plan to enforce an anti-picketing law against them). Nor do these assurances make Rodgers and Dilbeck’s fear of prosecution objectively unreasonable because the law’s plain language covers their intended activities, and

-4- they have already been arrested or cited under a prior version of the law.3 See Saint Paul Area Chamber of Commerce v. Gaertner, 439 F.3d 481, 485 (8th Cir. 2006) (stating that “fear of prosecution is not imaginary or speculative” when the law, “on [its] face,” prohibits the plaintiffs’ conduct).

Having established that Rodgers and Dilbeck’s chilled speech amounts to a constitutional injury, we have no trouble concluding that the injury is fairly traceable to the potential enforcement of the anti-loitering law and would be redressed by an injunction prohibiting its enforcement. Lujan, 504 U.S. at 560–61. Accordingly, they have standing to seek a preliminary injunction. We now turn to the injunction itself.

III.

The district court granted a preliminary injunction after weighing four factors: “(1) the threat of irreparable harm to [Rodgers and Dilbeck]; (2) the state of the balance between this harm and the injury that granting the injunction [would] inflict on [Arkansas]; (3) the probability that [Rodgers and Dilbeck would] succeed on the merits; and (4) the public interest.” Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d 109, 113 (8th Cir. 1981) (en banc). Under the third Dataphase factor, parties seeking to preliminarily enjoin the “implementation of a state statute” must demonstrate that they are “likely to prevail on the merits.” Planned Parenthood Minn., N.D., S.D. v. Rounds, 530 F.3d 724, 731–32 (8th Cir. 2008) (en banc) (citation omitted). This is in contrast to the “fair chance” of success that is typically required. Id. at 732. The higher bar “reflects the idea that governmental policies implemented through

3 In defending the constitutionality of its anti-loitering law, Arkansas argues that one of the law’s chief virtues is the absence of an intent requirement. This so- called “virtue,” however, undermines its standing argument. If the law does not require potential violators to intend to alarm or harass others or create a traffic hazard, the risk “of inadvertently or negligently” violating the law can itself establish “a reasonable fear of prosecution” under the right circumstances. 281 Care Comm., 638 F.3d at 629. -5- legislation . . . [and] developed through presumptively reasoned democratic processes are entitled to a higher degree of deference and should not be enjoined lightly.” Id. (citation omitted). Generally, if a party shows a “likely violation of his or her First Amendment rights, the other requirements for obtaining a preliminary injunction are . . . deemed to have been satisfied.” Minn. Citizens Concerned for Life, Inc. v. Swanson, 692 F.3d 864, 870 (8th Cir. 2012) (en banc) (citation omitted). We review a district court’s balancing of the Dataphase factors for an abuse of discretion. Rounds, 530 F.3d at 733.

A.

The district court ruled that Rodgers and Dilbeck were likely to prevail on their claim that Arkansas’s anti-loitering law violates the First Amendment. Because Arkansas has so far failed to establish that the law is narrowly tailored to achieve a compelling interest, we agree.

1.

Rodgers and Dilbeck want to go to public areas, hold signs, and speak. The fact that they intend to ask for money does not mean that their speech is unprotected. To the contrary, asking for charity or gifts, whether “on the street or door to door,” is protected First Amendment speech. Vill. of Schaumburg v. Citizens for a Better Env’t, 444 U.S. 620, 632 (1980); see also Reynolds v. Middleton, 779 F.3d 222, 225 (4th Cir. 2015) (“There is no question that panhandling and solicitation of charitable contributions are protected speech.”); Speet v. Schuette, 726 F.3d 867, 874–78 (6th Cir. 2013) (collecting cases and concluding that begging is protected by the First Amendment).

But just because speech is protected does not mean that it must go unregulated. See, e.g., United States v. Stevens, 559 U.S. 460, 468 (2010) (listing traditional limitations upon the freedom of speech); Burson v. Freeman, 504 U.S. 191, 211 (1992) (plurality opinion) (upholding a law restricting vote solicitation within 100

-6- feet of a polling place). The Supreme Court has long held that a state can regulate speech if it can establish a good enough reason and a fit between the claimed reason and the chosen means of regulation. See NAACP v. Button, 371 U.S. 415, 438–39 (1963). The state’s burden depends on the features of the regulation itself. See, e.g., Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 641–42 (1994) (explaining that the level of scrutiny varies depending upon whether a regulation restricts speech based on its content). If a law limits speech based on subject matter—otherwise referred to as a content-based restriction—then it is subject to strict scrutiny. Citizens United v. FEC, 558 U.S. 310, 340 (2010). Strict scrutiny requires a state to show that its law is narrowly tailored to serve a compelling interest. Id.

Arkansas’s anti-loitering law is a content-based restriction because it regulates speech based on “the topic discussed or the idea or message expressed.” Reed v. Town of Gilbert, 135 S. Ct. 2218, 2227 (2015). It applies only to those asking for charity or gifts, not those who are, for example, soliciting votes, seeking signatures for a petition, or selling something. In other words, its application depends on the “communicative content” of the speech. Id. at 2226.

To be sure, the law has other elements that limit its scope. As relevant here, it covers only speakers asking for charity or gifts who “[l]inger[] or remain[]” in certain areas and, even then, only when they engage in harassing speech or speech that is “likely to cause alarm” or a “traffic hazard.” Ark. Code Ann. § 5-71- 213(a)(3). But these limitations do not transform the law into a content-neutral restriction. Rather, because the law targets only a subset of the speech that meets these criteria—specifically, asking for charity or a gift—it is a content-based restriction. See Reed, 135 S. Ct. at 2227; see also R.A.V. v. City of St. Paul, 505 U.S. 377, 391 (1992) (explaining that, although the government can limit certain categories of speech, like fighting words, it cannot single out speech within a category for “special prohibitions” based on its content). Strict scrutiny therefore applies.

-7- 2.

Under strict scrutiny, Arkansas’s anti-loitering law will survive only if it is narrowly tailored to further a compelling interest. Reed, 135 S. Ct. at 2231. Arkansas argues that the law furthers the state’s compelling interest in public and motor vehicle safety through the prevention of aggressive conduct and traffic hazards. But even if this interest is “compelling,” the state has not satisfied the other half of strict scrutiny: showing that the law is narrowly tailored to achieve its public and motor-vehicle safety interest.

As written, Arkansas’s anti-loitering the law is underinclusive. 281 Care Comm. v. Arneson, 766 F.3d 774, 787 (8th Cir. 2014) (noting that a law is not narrowly tailored if, among other things, it “leave[s] significant influence bearing on the interest unregulated” (citation omitted)). Other types of solicitation (e.g., political or commercial) done (1) “[i]n a harassing or threatening manner,” (2) “[i]n a way likely to cause alarm,” or (3) “[u]nder circumstances that create a traffic hazard or impediment” are equally dangerous. Ark. Code Ann. § 5-71-213(a)(3). While the Supreme Court has held that “a legislature may deal with one part of a problem without addressing all of it,” that rule “has less force when a classification turns on the subject matter of expression.” Erznoznik v. City of Jacksonville, 422 U.S. 205, 215 (1975). Thus, a “regulation cannot discriminate on the basis of content unless there are clear reasons for the distinctions.” Id. Because Arkansas has offered no justification for its decision to single out charitable solicitation from other types of solicitation, the anti-loitering law is underinclusive and, consequently, not narrowly tailored. See Reed, 135 S. Ct. at 2231–32; Brown v. Entm’t Merchs. Ass’n, 564 U.S. 786, 801–02 (2011) (stating that a law restricting the sale of violent video games to minors was “wildly underinclusive” because the state “declined to restrict” other forms of media that could be equally dangerous to children); Johnson v. Minneapolis Park & Recreation Bd., 729 F.3d 1094, 1100–01 (8th Cir. 2013) (holding that a regulation designed to reduce crowd congestion at a festival was underinclusive because it prohibited the distribution of literature without a booth but

-8- allowed “[b]uskers like mimes, musicians, and living statues [who] aim to attract an audience”).

We conclude that Rodgers and Dilbeck will likely succeed in proving that Arkansas’s anti-loitering law violates the First Amendment. Moreover, because they have established that the law likely violates the First Amendment, we find that Rodgers and Dilbeck have satisfied the remaining three Dataphase factors as well. Swanson, 692 F.3d at 870. Therefore, injunctive relief is proper.

B.

The lone remaining issue for decision is whether the district court abused its discretion in applying the preliminary injunction statewide rather than limiting its application to Rodgers and Dilbeck. Arkansas, citing Califano v. Yamasaki, argues that the injunctive relief should be limited to Rodgers and Dilbeck because “injunctive relief should be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.”4 442 U.S. 682, 702 (1979). But the Supreme Court also wrote in Califano that one of the “principles of equity jurisprudence” is that “the scope of injunctive relief is dictated by the extent of the violation established, not by the geographical extent of the plaintiff class.” Id.; see also Trump v. Int’l Refugee Assist. Project, 137 S. Ct. 2080, 2087 (2017) (per curiam) (“Crafting a preliminary injunction is an exercise of discretion and judgment, often dependent as much on the equities of a given case as the substance of the legal issues it presents.”); Lemon v. Kurtzman, 411 U.S. 192, 200 (1973) (plurality opinion) (“[E]quitable remedies are a special blend of what is necessary,

4 To the extent that Arkansas relies on authorities discussing the impropriety of nationwide injunctions, we note that statewide injunctions do not raise all the same concerns as nationwide injunctions. See Amanda Frost, In Defense of Nationwide Injunctions, 93 N.Y.U. L. Rev. 1065, 1104–15 (2018) (identifying forum shopping, the risk of conflicting injunctions, and doctrinal inconsistencies as potential “costs” of nationwide injunctions).

-9- what is fair, and what is workable.” (footnote omitted)); De Beers Consol. Mines Ltd. v. United States, 325 U.S. 212, 220 (1945) (“A preliminary injunction is always appropriate to grant intermediate relief of the same character as that which may be granted finally.”). Thus, Califano supports the entirely opposite conclusion: that injunctive relief should extend statewide because the violation established—the plain unconstitutionality of Arkansas’s anti-loitering law—impacts the entire state of Arkansas. Moreover, Arkansas’s reading of Califano would, in effect, require every plaintiff seeking statewide relief from legislative overreach to file for class certification. That cannot be the law.

Because Califano itself does not foreclose the imposition of statewide injunctive relief, we consider more generally whether the district court abused its discretion in imposing such relief. Arkansas argues that the district court “gave no rationale for enjoining enforcement as to all beggars in [Arkansas].” However, the district court specifically found that: (1) Arkansas’s anti-loitering law is “plainly unconstitutional”; (2) Arkansas’s public interest “is best served by preventing governmental intrusions into the rights protected under the Federal Constitution”; and (3) “preventing [Arkansas] from enforcing a law that is plainly unconstitutional” would cause “no injury.” These findings were sufficient to justify the district court’s imposition of a statewide preliminary injunction, particularly because they in no way depended on facts unique to Rodgers and Dilbeck.

Arkansas urges us to consider the fact that the district court did not make a final determination regarding the constitutionality of its anti-loitering law. But the remote chance that Arkansas may, at a later stage, prove its anti-loitering law constitutional cannot be held sufficient to overcome the public’s interest in protecting freedom of expression under the First Amendment. Phelps-Roper v. Nixon, 545 F.3d 685, 690 (8th Cir. 2008) (finding that “it is always in the public interest to protect constitutional rights” and “[t]he balance of equities . . . generally favors the constitutionally-protected freedom of expression”), overruled on other grounds by Phelps-Roper v. City of Manchester, 697 F.3d 678, 692 (8th Cir. 2012) (en banc).

-10- We recognize that, under Rounds, “legislation or regulations developed through presumptively reasoned democratic processes are entitled to a higher degree of deference and should not be enjoined lightly.” 530 F.3d at 732. But we account for this higher degree of deference by requiring district courts to “make a threshold finding that a party is likely to prevail on the merits” when “a preliminary injunction is sought to enjoin the implementation of a duly enacted state statute.” Id. at 732– 33 (emphasis added). Rounds in no way imposes additional burdens on plaintiffs seeking statewide injunctive relief. Here, the district court found that Rodgers and Dilbeck met Rounds’s substantial burden, determining Arkansas’s anti-loitering law is not only “likely” unconstitutional, but “plainly unconstitutional.” Considering that finding, as well as the other findings discussed above, it was well within the broad discretion of the district court to grant Rodgers and Dilbeck a statewide preliminary injunction.

The dissent raises concerns for the limits of judicial power in deciding the scope of preliminary injunctive relief and in granting broad relief by default. However, broad preliminary relief is often appropriate under current law where, as here, a plaintiff brings a facial challenge to a statute under the First Amendment. See Ashcroft v. Am. Civil Liberties Union, 542 U.S. 656, 671 (2004). In Ashcroft, the Supreme Court upheld a nationwide preliminary injunction against enforcement of the Child Online Protection Act. Id. at 666–72. The Court explained that, when balancing the equities, there were “important practical reasons to let the injunction stand pending a full trial on the merits.” Id. at 670. Specifically, “[w]here a prosecution is a likely possibility, yet only an affirmative defense is available, speakers may self-censor rather than risk the perils of trial.” Id. at 670–71. In such a situation, as in the present case, the potential for harm extends beyond the parties. “There is a potential for extraordinary harm and a serious chill upon protected speech,” and such extraordinary harm outweighs any harm of leaving the preliminary injunction in place where “[n]o prosecutions have yet been undertaken under the law” and the Government will be able to enforce related laws already in existence. Id. at 671.

-11- This court and other courts have followed this guidance, finding that, “in cases such as this, where we determine the appellants are likely to win on the merits of their First Amendment claim, a preliminary injunction is proper.” Minn. Citizens Concerned for Life, Inc., 692 F.3d at 877 (stating that although “a high level of deference is appropriate,” a statute “suppressing or restricting speech must be judged by the sometimes inconvenient principles of the First Amendment” (quoting United States v. Alvarez, 567 U.S. 709, 715 (2012))); Phelps-Roper v. Troutman, 662 F.3d 485, 488, 490 (8th Cir. 2011) (per curiam), vacated on reh’g on other grounds, 705 F.3d 845 (8th Cir. 2012); see also, e.g., Sanders Cty. Republican Cent. Comm. v. Bullock, 698 F.3d 741, 749 (9th Cir. 2012) (reversing the district court’s denial of the appellants’ motion for a preliminary injunction and concluding that where a Montana statute criminalized political party endorsement of judicial candidates, and therefore violated the First Amendment on its face, a broad preliminary injunction against its enforcement was appropriate). In the present case, the scope of the preliminary injunction reflects no abuse of discretion on the part of the district court. Rodgers and Dilbeck argued to the district court the impracticability of more narrow relief, and the district court questioned what, if any, arguments were left for parties to address at trial. In short, we do not view the district court as having applied a default rule or as having abused its discretion.

IV.

For the foregoing reasons, we affirm the district court’s grant of a statewide preliminary injunction barring enforcement of Arkansas’s anti-loitering law.

STRAS, Circuit Judge, concurring in part and dissenting in part.

I agree with the court that Rodgers and Dilbeck have adequately alleged standing, that they will likely succeed in proving that Arkansas’s anti-loitering law violates the First Amendment, and that the law should not be enforced against them in the meantime. But I do not agree that the district court could prevent the entire Arkansas State Police force from enforcing the law against anyone, anywhere, at any

-12- time based on the harm faced by two individual plaintiffs. Because such broad relief was neither within the court’s power to grant nor justified by its reasoning, I would vacate the portion of the injunction preventing Arkansas police officers from enforcing the law against everyone else.

Let’s be clear about what the district court did here: it granted a universal 5 preliminary injunction to Rodgers and Dilbeck without asking whether it could, or even should. Whether it could depends on what the “High Court of Chancery in England” had the authority to do in a case like this one “at the time of the adoption of the Constitution and the enactment of the original Judiciary Act, 1789.” Grupo Mexicano de Desarrollo S.A. v. All. Bond Fund, Inc., 527 U.S. 308, 318 (1999) (citation omitted); see also Boyle v. Zacharie, 31 U.S. 648, 658 (1832) (explaining that “the remedies in equity are to be administered . . . according to the practice of courts of equity in the parent country”). Whether it should have rests on the district court’s consideration of several case-specific factors. See Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d 109, 113 (8th Cir. 1981) (en banc) (laying out four categories of factors for the district court to consider and weigh).

On the first point, history tells us that the Court of Chancery could not have granted a universal preliminary injunction under these circumstances. And on the

5 The injunction is “universal” in the sense that it prohibits enforcement of the law against anyone and everyone, whether they are parties or not. See Trump v. Hawaii, 138 S. Ct. 2392, 2425 n.1 (2018) (Thomas, J., concurring) (explaining why this term captures what makes such injunctions “distinctive”). I acknowledge that a universal injunction of this type does not necessarily raise all of the same policy concerns as those directed at federal laws or nationwide programs. See generally, e.g., Samuel L. Bray, Multiple Chancellors: Reforming the National Injunction, 131 Harv. L. Rev. 417 (2017); Amanda Frost, In Defense of Nationwide Injunctions, 93 N.Y.U. L. Rev. 1065 (2018). But my concern is with the limits of judicial power, not the wisdom of universal injunctions. -13- second, the district court abused whatever equitable discretion it had by presuming that Rodgers and Dilbeck were entitled to one.

In 1789, as a general rule, the English Court of Chancery granted injunctions for one overarching purpose: to prevent violations of the moving party’s rights. See 1 George Spence, The Equitable Jurisdiction of the Court of Chancery *668–69 (Philadelphia, Lea & Blanchard 1846) (tracing the development of injunctions to dissatisfaction with the English common law’s inability to provide “preventive justice by direct means”); see also 1 John Norton Pomeroy, A Treatise on Equity Jurisprudence as Administered in the United States of America § 91, at 76 (San Francisco, A. L. Bancroft & Co. 1881) (explaining that “the function and object” of courts of equity was to “maintain and preserve inviolate the primary rights and duties of the litigant parties” by “conferring the remedies adapted to the injury” (emphasis added)); id. § 112, at 93 (describing the “essential elements” of different types of injunctions). To that end, injunctions were “directed only to the parties,” 2 Joseph Story, Commentaries on Equity Jurisprudence as Administered in England and America § 875, at 166 (Boston, Hilliard, Gray, & Co. 1836), and would require them “to do a particular thing, or to refrain from doing a particular thing, according to the exigency of the writ,” id. § 861, at 154. So, for example, injunctions could

stay proceedings in courts of law, in the spiritual courts, the courts of admiralty, or in some other court of equity; . . . restrain the indorsement or negotiation of notes and bills of exchange, the sale of land, the sailing of a ship, the transfer of stock, or the alienation of a specific chattel; . . . prevent the wasting of assets or other property pending litigation; . . . restrain a trustee from assigning the legal estate, from setting up a term of years, or assignees from making a dividend; . . . prevent the removing out of the jurisdiction, marrying, or having any intercourse which the court disapproves of, with a ward; . . . restrain the commission of every species of waste to houses, mines, timber, or any other part of the inheritance; . . . prevent the infringement of patents, and the violation of copyright either by publication or theatrical

-14- representation; . . . suppress the continuance of public or private nuisances; and by the various modes of interpleader, restrain[] [a] multiplicity of suits, or quiet[] possession before the hearing, to stop the progress of vexatious litigation.

Robert Henley Eden, A Treatise on the Law of Injunctions 1 (New York, William Gould & Co. 1822) (adding that “[t]hese . . . [were] far from being all the instances in which this species of equitable interposition [was] obtained”).

The common thread was that courts of equity tailored injunctions to the particular harms that the moving party faced, whether during litigation or after. Injunctions to stay proceedings in courts of law—“one of the most ordinary modes of equitable interposition” at the time—illustrate the point. Id. at 3. The basic idea was that a party facing an unjust claim or judgment in a common-law court could seek an injunction from a court of equity “to prevent an unfair use being made of the process.” 2 Story, Equity Jurisprudence, supra, § 875, at 166. The injunction might “stay trial; or after verdict[,] stay judgment; or after judgment, stay execution; or, if the execution ha[d] been effected, stay the money in the hands of the sheriff[,] . . . according to the exigency of the particular case.” Id. § 874, at 165.

This party-centered understanding of injunctions was also reflected in equity’s requirement “that all persons materially interested, either legally, or beneficially, in the subject-matter of a suit, are to be made parties to it, . . . however numerous they may be.” Joseph Story, Commentaries on Equity Pleadings, and the Incidents Thereof, According to the Practice of the Courts of Equity, of England and America § 72, at 86 (Boston, Charles C. Little & James Brown 1848). The rule that each party could only get an injunction (or other relief) to protect its own rights meant that courts of equity, “to do complete justice,” had to insist that the plaintiffs join everyone with an interest in the litigation. 1 Edmund Robert Daniell, A Treatise on the Practice of the High Court of Chancery, with Some Practical Observations on the Pleadings in That Court 169 (Harrisburg, I.G. M’Kinley & J.M.G. Lescure 1845). Indeed, the idea that an injunction should just be the flipside of the threatened harm to the moving party was so deeply engrained that the central debate around the -15- time of the Founding was about “when an injunction may be granted,” 3 John Norton Pomeroy, A Treatise on Equity Jurisprudence as Administered in the United States of America § 1338, at 368 (San Francisco, A. L. Bancroft & Co. 1883) (emphasis added), not what it could do or how. See generally, e.g., Eden, supra; 3 Pomeroy, supra, §§ 1337–65, at 366–402; 2 Story, Equity Jurisprudence, supra, §§ 861–959, at 154–227. 6 It was natural at the time that an injunction would be “based upon and exactly fitted to” the violation it was meant to prevent or undo. 1 Pomeroy, supra, § 108, at 90; see also 3 Thomas Atkins Street, Federal Equity Practice §§ 2371–72, at 1387–88 (1909) (explaining that an injunction “should be sufficiently extensive to protect the right of the plaintiff that the defendant is violating” but “must not be given a broader scope than the case made in the bill”).

To be sure, preventing a violation of the plaintiff’s rights sometimes had the effect of protecting others too, such as in cases involving a public nuisance. But the benefit to nonparties in those cases was incidental, cf. 3 Edmund Robert Daniell, A Treatise on the Practice of the High Court of Chancery, with Some Practical Observations on the Pleadings in That Court 191 (Harrisburg, I.G. M’Kinley & J.M.G. Lescure 1846) (explaining that it was up to the Attorney General to sue for an injunction to protect the public generally), and only necessary because it was impossible to peel off just the portion of the nuisance that harmed the plaintiff. A court accordingly had no choice but to enjoin the whole thing. But as far as the court was concerned, the injunction was still about protecting the plaintiff’s “private rights,” Eden, supra, at 163, not preventing harm to “the public at large,” 3 Pomeroy, supra, § 1349, at 380–81.

6 Cf. 3 Edmund Robert Daniell, A Treatise on the Practice of the High Court of Chancery, with Some Practical Observations on the Pleadings in That Court 166 (Harrisburg, I.G. M’Kinley & J.M.G. Lescure 1846) (transcribing the standard form of an injunction to stay proceedings in a court of law: “We, therefore, in consideration of the premises, do strictly enjoin and command you, the said [defendants] . . . that you and every of you do absolutely desist from all further proceedings at law against the said complainant touching any of the matters in the said bill complained of” (emphasis added)); Eden, supra, at 265–72 (same, for various other types of injunctions). -16- B.

So far, the focus has been on individual lawsuits. In those cases, an injunction from the Court of Chancery would target only the specific harms faced by the moving party—typically the plaintiffs—unless addressing those harms necessarily required it to do something more. The rules were different, however, if the plaintiffs were suing on behalf of others too.

Representative suits arose out of the equitable practice of using a single case to resolve issues that would have otherwise resulted in a “multiplicity” of individual lawsuits (or none at all, if each standalone case would not have been worth bringing on its own). John Mitford, A Treatise on the Pleadings in Suits in the Court of Chancery by English Bill 127–29 (Philadelphia, P. Byrne 1812); 3 William Blackstone, Commentaries *438–39; see also, e.g., Mayor of York v. Pilkington (1737) 26 Eng. Rep. 180, 181; 1 Atk. 282, 284. If joining everyone with an interest in the litigation was impracticable because there were too many parties to manage, then a court of equity might allow just a few to sue (or be sued) on behalf of others. See 1 Daniell, supra, at 198 (describing “[t]he practice adopted by the Court [of Chancery] of permitting one or more persons to represent in a suit all who have similar interests”); Story, Pleadings, supra, §§ 94–96, 120, at 119–23, 152; see also, e.g., Adair v. New River Co. (1805) 32 Eng. Rep. 1153, 1159; 11 Ves. jun. 429, 445 (stating that it was only necessary to join enough parties “that it can be justly said, they will fairly and honestly try the legal right”); Brown v. Howard (1701) 21 Eng. Rep. 960, 960; 1 Eq. Ca. Abr. 163, 163 (justifying this approach on the ground that “else, where there are such Numbers, no Right could be done”). The smaller representative group would then seek an injunction or some other type of equitable relief for the benefit of the whole group. See Frederic Calvert, A Treatise upon the Law Respecting Parties to Suits in Equity 25 (Philadelphia, John S. Littell 1837) (discussing the general rule that “when a large number of persons have a common interest in the entire object of a suit in its nature beneficial to them all, one or more of them may sue on behalf of all”); Zechariah Chafee, Jr., Some Problems of Equity 200–13 (1950) (recounting the “origins and development of representative suits”

-17- (capitalization omitted)). The procedural device through which this occurred was usually referred to as a “[b]ill of [p]eace.” E.g., 2 Story, Equity Jurisprudence, supra, at §§ 853–54, at 148–49.

The procedures for filing bills of peace in federal court were formalized in Federal Equity Rule 48, which came into existence in 1842 and later became Federal Equity Rule 38. See James Love Hopkins, The New Federal Equity Rules 105 (1918) (characterizing Equity Rule 48 as “announc[ing] a pre-existing rule of equity procedure” (capitalization omitted)); id. at 203 (calling Equity Rule 38 “[a] new rule, in affirmance of an old principle”). When law and equity merged, the bill of peace evolved into the modern class action, now governed by Federal Rule of Civil Procedure 23. See Fed. R. Civ. P. 23 advisory committee’s note to 1937 adoption (“This is a substantial restatement of former Equity Rule 38 (Representatives of Class) as that rule has been construed.”); see also Supreme Tribe of Ben Hur v. Cauble, 255 U.S. 356, 363, 366 (1921) (linking “class suits” in federal courts to representative cases in Chancery); 7A Charles Alan Wright et al., Federal Practice & Procedure § 1751.

From the start, these rules provided the exclusive mechanism for filing run- of-the-mill representative suits.7 The first two equity rules simply declared that federal courts had the power to act “[w]here the parties on either side [were] very numerous, and [could] not, without manifest inconvenience and oppressive delays in the suit, be all brought before [the court]”—a near picture-perfect reflection of what was required to bring a “bill of peace” before the Court of Chancery. Fed. R. Equity P. 48 (1842); see also Fed. R. Equity P. 38 (1912) (“When the question is

7 To be sure, other types of lawsuits are arguably “representative” too. See, e.g., Fed. R. Civ. P. 17(a) (suits by an individual on behalf of a “real party in interest”); Fed. R. Civ. P. 23.1 (shareholder-derivative actions); 7A Wright et al., supra, § 1807 (collective actions under the Fair Labor Standards Act). But none of these other procedural devices even arguably apply here, so their existence has no bearing on whether the district court exceeded its equitable powers in granting a universal injunction in this particular case. -18- one of common or general interest to many persons constituting a class so numerous as to make it impracticable to bring them all before the court, one or more may sue or defend for the whole.”). The 1937 adoption of Rule 23 added structure, but continued to cover the universe of cases in which “one or more . . . may, on behalf of all, sue or be sued.” Fed. R. Civ. P. 23(a) (1937). None of these formulations carved out any exceptions.

The current version of Rule 23—although more restrictive of representative suits than the original bills of peace or the federal rules of equity—is even more clearly comprehensive than its predecessors. By providing that members of a class may sue “as representative parties on behalf of all members only if” they satisfy certain conditions, the rule leaves plaintiffs with no room to argue that they can use some other procedure to seek relief on behalf of others. Fed. R. Civ. P. 23(a)–(b) (emphasis added). The absence of any discussion of Rule 23 in this case, either by the district court or the parties, confirms that Rodgers and Dilbeck filed an individual suit, not one allowing them to represent the interests of others. Cf. Taylor v. Sturgell, 553 U.S. 880, 901 (2008) (rejecting a theory of “virtual representation” that would have “allowed courts to create de facto class actions at will” by “recogniz[ing], in effect, a common-law kind of class action” (internal quotation marks and citation omitted)).

In fact, Rodgers and Dilbeck have never claimed to be suing on behalf of anyone else. See Calvert, supra, at 19 (“The bill [of peace] must contain a specific allegation that the plaintiffs are suing . . . on behalf of themselves and others.” (footnote omitted)); 1 Daniell, supra, at 199 (“[I]n all cases where one or a few individuals of a large number, institute a suit on behalf of themselves and the others, they must so describe themselves in the bill . . . .”); Story, Equity Pleadings, supra, § 126, at 158–59. Neither their complaint nor their subsequent filings say so. Cf. 1 Daniell, supra, at 200 (noting that the court sometimes allowed plaintiffs to amend their pleadings to add “the words, ‘on behalf of himself [and others],’” when the omission “was evidently a mere slip”). And vague allegations in the introductory paragraph of their complaint that they are not “alone” and that others “suffer th[e]

-19- same government persecution for their speech” did not transform their individual action into a representative suit, particularly because they have never made an attempt to represent or join others. See Fed. R. Civ. P. 23(b) (explaining the requirements for “maintain[ing]” a class action).

Moreover, even if Rodgers and Dilbeck had decided to bring a representative suit, it would have been unfamiliar to the Court of Chancery. See Grupo Mexicano, 527 U.S. at 318. First, the link between Rodgers and Dilbeck and the diffuse class of Arkansas beggars who could benefit from the universal injunction is so attenuated that it would have stretched the limits of what courts of equity could do, even in representative suits. See 1 Pomeroy, supra, §§ 268–69, at 292–93 (explaining that traditionally “a community of interest merely in the question of law or of fact involved” was not good enough, although the doctrine was later “extended”); Story, Equity Pleadings, supra, § 120, at 152 (explaining that although the parties did not have to be in “privity,” they did have to share “a common interest,” “common right,” or “general claim or privilege”); cf. Pilkington, 26 Eng. Rep. at 181; 1 Atk. at 284. Second, a traditional feature of representative suits, at least in England, was that unnamed and unjoined members of the class would be bound by an unfavorable decision, just as they would benefit from a favorable one. 8 See, e.g., Adair, 32 Eng. Rep. at 1159; 11 Ves. jun. at 445; Brown v. Vermuden (1676) 22 Eng. Rep. 796, 797; 1 Chan. Cas. 272, 272 (“If the Defendant should not be bound, Suits of this Nature . . . would be infinite, and impossible to be ended.”); 7A Wright et al., supra, § 1751 (“[T]he English practice was to treat a class-action judgment as binding on everyone in the group.”). Here, by contrast, even if Rodgers and Dilbeck were to ultimately lose this case, others in Arkansas could keep challenging the anti-loitering law until

8 Despite English practice, this issue was unsettled under federal law for a long time. Compare Fed. R. Equity 48 (1842) (providing that in representative suits, “the decree shall be without prejudice to the rights and claims of all the absent parties”), with Smith v. Swormstedt, 57 U.S. 288, 303 (1853) (ignoring Rule 48 and stating that “the decree binds all of [the class members] the same as if all were before the court”), and Fed. R. Equity 38 (1912) (dropping the “without prejudice” clause from Rule 48). -20- some judge finally granted relief. See Canady v. Allstate Ins. Co., 282 F.3d 1005, 1014–16 (8th Cir. 2002) (describing the requirements for res judicata and collateral estoppel under federal law). Third, although less clear, some authority suggests that even when a bill of peace was brought, the Court of Chancery would only extend relief to nonparties through a perpetual injunction issued at the end of the case, not by ordering a temporary, preliminary injunction like the one here. See 3 Daniell, supra, at 217.

* * *

The bottom line is that the relief ordered in this case does not resemble what was “traditionally accorded” to plaintiffs like these in cases like this one. Grupo Mexicano, 527 U.S. at 319. If the Court of Chancery could not grant a universal injunction in 1789, then neither can the district court today.

C.

There is another problem too. Even if the district court had the power to grant a universal preliminary injunction, it should not have presumed that Rodgers and Dilbeck were entitled to one. A preliminary injunction of any scope is “an extraordinary remedy,” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24 (2008), and before granting one, a district court is required to consider and weigh several factors: “(1) the threat of irreparable harm to the movant; (2) the state of the balance between this harm and the injury that granting the injunction will inflict on other parties litigant; (3) the probability that [the] movant will succeed on the merits; and (4) the public interest.” See Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d 109, 113 (8th Cir. 1981) (en banc).

So what did the district court have to say about these factors here? Close to nothing. In less than a page of analysis, with only a line or two—at most—touching on the breadth of the injunction, the court enjoined the entire Arkansas State Police force from enforcing the law against everyone statewide, solely because, in its view,

-21- the anti-loitering law “plainly” violated the First Amendment. The court barely discussed any other Dataphase factor, and certainly not in the context of deciding how broad the injunction should be. Yet remarkably, this court says that this truncated analysis was “sufficient to justify” a universal preliminary injunction. Ante at 10. In my view, this universal-injunction-by-default approach is “plainly” wrong.

To be sure, as the court claims, showing a strong likelihood of success on the merits generally entitles a plaintiff to a preliminary injunction in First Amendment cases. See Telescope Media Grp. v. Lucero, 936 F.3d 740, 762 (8th Cir. 2019). But this maxim is not absolute and only gets the court so far. After all, a preliminary injunction “does not follow as a matter of course from a plaintiff’s showing of a likelihood of success on the merits,” even in First Amendment cases, and the district court must still consider the other factors and show its work. Benisek v. Lamone, 138 S. Ct. 1942, 1944 (2018) (per curiam) (citation omitted) (addressing a First Amendment retaliation claim); id. (explaining that district courts must consider whether the plaintiffs “[are] likely to suffer irreparable harm in the absence of preliminary relief, [whether] the balance of equities tips in [their] favor, and [whether] an injunction is in the public interest”). Here, the district court did not do either, much less explain why a universal preliminary injunction was appropriate or necessary. Id. at 1943 (reiterating that a preliminary injunction of any type is never “awarded as of right” (quoting Winter, 555 U.S. at 24)).

Examining the other Dataphase factors, it becomes apparent that such a broad injunction was not necessary. See Dataphase, 640 F.2d at 113. Though Rodgers and Dilbeck personally faced “the threat of irreparable harm,” there is every reason to believe that a narrower injunction would have fully remedied it. Indeed, neither this court nor the district court has provided any reason to believe that safeguarding Rodgers’s and Dilbeck’s right to speak somehow depends on preventing

-22- enforcement of the anti-loitering law against everyone else.9 Cf. Reprod. Health Servs. of Planned Parenthood of the St. Louis Region, Inc. v. Nixon, 428 F.3d 1139, 1145 (8th Cir. 2005) (vacating a preliminary injunction that “went beyond what was needed to protect [the plaintiffs] from the . . . threat of irreparable injury shown”).

Not to mention that a universal preliminary injunction potentially injures Arkansas, its citizens, and the overall public interest. See Dataphase, 640 F.2d at 113. Even if the district court is right that the harm caused by prohibiting enforcement of an unconstitutional law cannot tip the balance of the equities against injunctive relief, it was wrong when it said that it could “think of no injury caused by preventing [Arkansas] from enforcing [the anti-loitering law].” There has been no final determination yet, so the court’s decision necessarily rested on an educated guess about the outcome of the case, based on a limited record and arguments that may not be fully developed. See Univ. of Tex. v. Camenisch, 451 U.S. 390, 395 (1981). Given the possibility that the law will survive strict scrutiny once Arkansas has a full opportunity to defend it—no matter how unlikely that may have seemed to the district court—a federal court order preventing the State from enforcing its law against thousands of potential violators is a significant encroachment on its police powers. See New Motor Vehicle Bd. of Cal. v. Orrin W. Fox Co., 434 U.S. 1345, 1351 (Rehnquist, Circuit Justice 1977) (staying a statewide injunction and noting that “any time a State is enjoined by a court from effectuating statutes enacted by representatives of its people, it suffers a form of irreparable injury”); cf. Planned Parenthood Minn., N.D., S.D. v. Rounds, 530 F.3d 724, 733 (8th Cir. 2008) (en banc) (recognizing the need for “an appropriately deferential analysis” before “thwart[ing] a state’s presumptively reasonable democratic processes”).

In short, the district court imposed an unnecessarily broad remedy with no affirmative justification, a serious misunderstanding of the harm it could cause, and

9 To be sure, Rodgers and Dilbeck “argued to the district court” that “narrow[er] relief” was impracticable. Ante at 12. Even so, the court still had an obligation to make its own findings. See Benisek, 138 S. Ct. at 1943–44. -23- an explanation that was cursory at best. If this was not an abuse of discretion, it is hard to fathom what would be.

None of this should be surprising. Both the limits of the district court’s equitable powers and the constraints we impose on the exercise of its discretion rest on the bedrock principle “that injunctive relief should be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.” Califano v. Yamasaki, 442 U.S. 682, 702 (1979).10 Rodgers and Dilbeck sued to vindicate their

10 The court relies on a different “principle” from Califano: “the scope of injunctive relief is dictated by the extent of the violation established, not by the geographical extent of the plaintiff class.” 442 U.S. at 702. Out of context, as the court presents it, this statement seems to support the position that “injunctive relief should extend statewide because the violation established . . . impacts the entire state of Arkansas.” Ante at 10. In context, however, its meaning is different.

The line in question comes from Califano’s discussion of whether to reject a class of Social Security beneficiaries from across the country on the ground that a nationwide class action would lead to overly burdensome injunctive relief. See 442 U.S. at 689, 701–02. Without ever questioning the premise that injunctive relief should go no further “than necessary to redress the complaining parties,” the Supreme Court explained that there was no need to worry, because a nationwide class would still need to show nationwide harm before it could receive nationwide relief. Id. at 702. In other words, when Califano linked “the scope of injunctive relief” to “the extent of the violation,” the point was that the injunction could be narrower than “the geographical extent of the plaintiff class,” not broader. Id. Relying on that line to justify granting a universal injunction to Rodgers and Dilbeck, who have not even tried to certify a class of any size, turns Califano on its head. It also ignores the very next paragraph of the opinion, which instructed district courts to “take care” to certify nationwide classes only when “nationwide relief is indeed appropriate.” Id. That advice would make little sense if nationwide relief in the form of a universal injunction were generally available to any plaintiff who established a broad enough violation, regardless of the size, or even the existence, of a class.

-24- own rights under the First Amendment, so complete relief for them means protection from arrest under Arkansas’s likely-unconstitutional anti-loitering law. An injunction directing the state police not to arrest them would, as far as the record shows, provide them with complete relief. The district court made a mistake by giving them more. Cf. Lewis v. Casey, 518 U.S. 343, 359–60 (1996) (explaining that constitutional violations suffered by two individual prisoners were “patently inadequate” to justify an injunction mandating sweeping changes to the legal- assistance programs in Arizona’s state prison system).

It is true, I admit, that we have allowed broad injunctions before with no more justification than the district court offered here. See, e.g., Minn. Citizens Concerned for Life, Inc. v. Swanson, 692 F.3d 864, 870, 877 (8th Cir. 2012) (en banc); Phelps- Roper v. Troutman, 662 F.3d 485, 488, 490 (8th Cir. 2011) (per curiam), vacated on reh’g on other grounds, 705 F.3d 845 (8th Cir. 2012); see also Ashcroft v. Am. Civil Liberties Union, 542 U.S. 656 (2004).11 But we have never held that a universal injunction is available by default to plaintiffs who are likely to prevail on a First Amendment challenge. Nor have we directly addressed a challenge to the power of

11 The court’s endorsement of the universal preliminary injunction in this case appears to hinge largely on Ashcroft v. American Civil Liberties Union, 542 U.S. 656 (2004), which it says supports the issuance of “a nationwide preliminary injunction” when a “plaintiff brings a facial challenge to a statute under the First Amendment.” Ante at 11. Factually speaking, Ashcroft involved both a universal preliminary injunction and a facial challenge. But that is where the similarities end. Only two issues were in play in Ashcroft: the merits of the plaintiffs’ First Amendment challenge and whether there were “important practical reasons to let the injunction stand pending a full trial on the merits.” Id. at 670–71. Nowhere did the Supreme Court address the scope of the injunction, much less decide that universal preliminary injunctions are appropriate whenever a facial First Amendment challenge is likely to succeed. See id. In short, neither issue came up. Id.; see also United States v. Verdugo-Urquidez, 494 U.S. 259, 272 (1990) (“The Court often grants certiorari to decide particular legal issues while assuming without deciding the validity of antecedent propositions, and such assumptions—even on jurisdictional issues—are not binding in future cases that directly raise the questions.” (internal citations omitted)). -25- a court to issue such a sweeping injunction. So even if we have “implicit[ly] resol[ved]” the issue before, it does not prevent us from reaching the right answer now. Streu v. Dormire, 557 F.3d 960, 964 (8th Cir. 2009) (“[W]e are generally not bound by a prior panel’s implicit resolution of an issue that was neither raised by the parties nor discussed by the panel.”).

Because the district court neither could nor should have granted a universal preliminary injunction, I respectfully dissent. ______________________________

-26-