Arizona Alliance for Retired Americans v. Kristin K. Mayes

Opinion

FOR PUBLICATION

UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT

ARIZONA ALLIANCE FOR No. 22-16490 RETIRED AMERICANS; VOTO LATINO; PRIORITIES USA, D.C. No. 2:22-cv-01374- Plaintiffs-Appellees, GMS

v. OPINION KRISTIN K. MAYES, in her official capacity as Attorney General for the State of Arizona,

Defendant-Appellant,

YUMA COUNTY REPUBLICAN COMMITTEE,

Intervenor-Defendant-Appellant,

and

KATIE HOBBS, in her official capacity as Secretary of State for the State of Arizona; LARRY NOBLE, Nominal Defendant, in his official capacity as Apache County Recorder; DAVID STEVENS, Nominal Defendant, in his official capacity as

Cochise County Recorder, previously named as David Stephens; PATTY HANSEN, Nominal Defendant, in her official capacity as Coconino County Recorder; SADIE JO BINGHAM, Nominal Defendant, in her official capacity as Gila County Recorder; WENDY JOHN, Nominal Defendant, in her official capacity as Graham County Recorder; SHARIE MILHEIRO, Nominal Defendant, in her official capacity as Greenlee County Recorder; RICHARD GARCIA, Nominal Defendant, in his official capacity as La Paz County Recorder; STEPHEN RICHER, Nominal Defendant, in his official capacity as Maricopa County Recorder; KRISTI BLAIR, Nominal Defendant, in her official capacity as Mohave County Recorder; MICHAEL SAMPLE, Nominal Defendant, in his official capacity as Navajo County Recorder; GABRIELLA CAZARES- KELLY, Nominal Defendant, in her official capacity as Pima County Recorder; DANA LEWIS, Nominal Defendant, in her official capacity as Pinal County Recorder; SUZANNE SAINZ, Nominal Defendant, in her official capacity as Santa Cruz County Recorder; MICHELLE BURCHILL, Nominal Defendant, in her official AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 3

capacity as Yavapai County Recorder; RICHARD COLWELL, Nominal Defendant, in his official capacity as Yuma County Recorder,

Defendants.

Appeal from the United States District Court for the District of Arizona G. Murray Snow, Chief District Judge, Presiding

Argued and Submitted En Banc June 25, 2025 Seattle, Washington

Filed August 7, 2026

Before: Mary H. Murguia, Chief Judge, and Consuelo M. Callahan, Mark J. Bennett, Ryan D. Nelson, Daniel A. Bress, Lawrence VanDyke, Jennifer Sung, Holly A. Thomas, Salvador Mendoza, Jr., Roopali H. Desai and Anthony D. Johnstone, Circuit Judges.*

Opinion by Judge Johnstone; Concurrence by Judge Bress; Concurrence by Judge VanDyke

* Judge Ikuta, who was originally a member of this en banc panel, died after the case was argued and submitted. Pursuant to Ninth Circuit General Order 5.1(b)(1), Judge Mendoza replaced her on the panel. Judge Mendoza reviewed all the case materials and watched the recording of oral arguments.

SUMMARY**

Elections / Standing

The en banc court vacated the district court’s preliminary injunction enjoining enforcement of two Arizona election law amendments: (1) a provision that requires a county recorder to cancel a voter’s registration in that county if the voter moves and registers to vote in a new county (“Cancellation Provision”); and (2) a provision that criminalizes knowingly providing a “mechanism for voting” to someone registered to vote in another state (“Felony Provision”). Each provision more stringently regulates the voting process in Arizona than did preexisting Arizona law. Plaintiffs the Arizona Alliance for Retired Americans, Voto Latino, and Priorities USA (“Organizations”) sued, claiming that the Cancellation and Felony Provisions would interfere with their voter-registration and voter-education efforts. Under FDA v. Alliance for Hippocratic Medicine, 602 U.S. 367 (2024), an organizational plaintiff must show that the challenged provision directly affects and interferes with its core business activities to establish an injury-in-fact. The en banc court overruled the organizational standing analyses in prior cases decided without the benefit of Hippocratic Medicine, which did not ask whether the challenged conduct “directly affected and interfered with” the plaintiff organizations’ core activities.

** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 5

The en banc court held that the Organizations lack standing to challenge the Cancellation Provision because they have not clearly shown that the Cancellation Provision affects or interferes with their voter-registration and voter- education efforts, and thus, have not established a sufficiently concrete and particularized injury-in-fact. The en banc court next held that the Organizations had standing to challenge the Felony Provision. Because the Felony Provision arguably proscribes voter registration and education, which are arguably affected with a constitutional interest, the Organizations face a credible threat of prosecution and cross the low threshold for standing in pre- enforcement challenges. Although the Organizations have standing to challenge the Felony Provision, the en banc court held that the district court erred in holding that the Organizations were likely to succeed on the merits of their claim. Although the stand- alone phrase “mechanism for voting” arguably encompasses voter registration and education, the full text and statutory context of the Felony Provision show that it likely does not criminalize those activities. Concurring in parts II.B and III and concurring in the judgment as to the Cancellation Provision, Judge Bress, joined by Judges Callahan, Bennett, R. Nelson, and VanDyke, wrote that the majority opinion reaches the right result, holding that certain public interest organizations lack Article III standing to challenge the Cancellation Provision, and correctly overrules the court’s past precedents—which expansively conferred organizational standing—as incompatible with Hippocratic Medicine, but failed to engage fully with the core reasoning of Hippocratic Medicine. He wrote that because all that the plaintiff

organizations have alleged (and could allege) is that the Cancellation Provision frustrates their organizational missions of enhancing voter registration, and that they have had to divert resources in responding to it, they do not establish organizational standing after Hippocratic Medicine. Concurring in parts II.B and III and concurring in the judgment, Judge VanDyke wrote that he would have taken this opportunity to slightly modify the standard set forth in this court’s decision in Miller v. Gammie, 335 F.3d 889, 893 (9th Cir. 2003) (en banc), which requires three-judge panels of this court to follow Ninth Circuit precedent unless it is “clearly irreconcilable” with subsequent Supreme Court precedent, to instead allow a three-judge panel to overturn a prior panel decision when an intervening Supreme Court decision is simply inconsistent with our court’s precedent. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 7

COUNSEL

Aria C. Branch (argued), Spencer W. Klein, Joel J. Ramirez, Christopher D. Dodge, Daniel J. Cohen, and Tina M. Morrison, Elias Law Group LLP, Washington, D.C.; Jonathan P. Hawley, Elias Law Group LLP, Seattle, Washington; Roy Herrera and Daniel A. Arellano, Herrera Arellano LLP, Phoenix, Arizona; for Plaintiffs-Appellees. Tracy A. Olson (argued), Brett W. Johnson, Eric H. Spencer, and Colin P. Ahler, Snell & Wilmer LLP, Phoenix, Arizona, for Intervenor-Defendant-Appellant. Alexander Samuels (argued), Principal Deputy Solicitor General; Joshua M. Whitaker and Jennifer J. Wright, Assistant Attorneys General; Drew C. Ensign, Deputy Solicitor General; Joseph A. Kanefield, Chief Deputy, Chief of Staff; Joshua Bendor, Solicitor General; Mark Brnovich, Former Attorney General of Arizona; Kristin K. Mayes, Attorney General of Arizona; Office of the Arizona Attorney General, Phoenix, Arizona; for Defendant-Appellant. Sean R. Janda (argued) and Daniel Tenny, Attorneys, Appellate Staff, Civil Division; Yaakov M. Roth, Acting Assistant Attorney General; United States Department of Justice, Washington, D.C.; Timothy Courchaine, United States Attorney, Office of the United States Attorney, United States Department of Justice, Phoenix, Arizona; for Amicus Curiae United States of America. Lila Miller, Reed Colfax, and Tara Ramchandani, Relman Colfax PLLC, Washington, D.C., for Amici Curiae Community Groups, Fair Housing Agencies, and Public Interest Organizations.

OPINION

JOHNSTONE, Circuit Judge: Arizona amended its election laws in 2022 with additional regulations of voter registration. Two provisions are at issue. The first requires a county recorder to cancel a voter’s registration in that county if the voter moves and registers to vote in a new county (“Cancellation Provision”). The second criminalizes knowingly providing a “mechanism for voting” to someone registered to vote in another state (“Felony Provision”). Three organizations—the Arizona Alliance for Retired Americans (“AARA”), Voto Latino, and Priorities USA (“Organizations”)—sued, claiming that the Cancellation and Felony Provisions would interfere with their voter-registration and voter-education efforts. The district court agreed and granted a preliminary injunction. We vacate the preliminary injunction and remand. First, we hold that the Organizations lack standing to challenge the Cancellation Provision. Under FDA v. Alliance for Hippocratic Medicine, 602 U.S. 367 (2024), an organizational plaintiff must show that the challenged provision “directly affect[s] and interfere[s] with [its] core business activities” to establish an injury-in-fact. Id. at 395. The Organizations assert that the Cancellation Provision will interfere with their voter-registration and voter-education efforts, and that, in response, they will need to divert resources to educate and assist voters whose registrations may be subject to cancellation. But because the Organizations have not clearly shown that the Cancellation Provision affects or interferes with their voter-registration and voter-education efforts, they have not established an injury-in-fact. Thus, we vacate the district court’s preliminary injunction of the Cancellation Provision. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 9

Second, we hold that the Organizations have standing to challenge the Felony Provision but are unlikely to succeed on the merits of their claim. The Organizations argue that they may be prosecuted under the Felony Provision because providing someone a “mechanism for voting” may include voter registration and education. We agree that the Felony Provision arguably proscribes these activities, which are arguably affected with a constitutional interest, and that the Organizations face a credible threat of prosecution. So the Organizations cross our low threshold for standing in pre- enforcement challenges. But the district court erred in holding that the Organizations were likely to succeed on the merits of their claim. Although the stand-alone phrase “mechanism for voting” arguably encompasses voter registration and education, the full text and statutory context of the Felony Provision show that it likely does not criminalize those activities. Thus, we also vacate the district court’s preliminary injunction of the Felony Provision. I. The Organizations challenge the 2022 amendments to Arizona’s voter-registration laws. Enacted in 2022, Senate Bill (SB) 1260 “[m]odifie[d] the criteria for voter registration cancellations, active early voting list regulations and violations associated with illegal voting.” Ariz. H.R. Summary S.B. 1260, 55th Reg. Sess., at 1 (2022). At issue are two provisions that SB 1260 added to Arizona’s Elections and Electors Code: (1) the Cancellation Provision, A.R.S. § 16-165(A)(11), (B), and (2) the Felony Provision, A.R.S. § 16-1016(12). Each provision more stringently regulates the voting process in Arizona than did preexisting Arizona law. The Cancellation Provision requires a county recorder to cancel a voter’s registration (1) “[w]hen the county recorder

receives confirmation from another county recorder that the person registered has registered to vote in that other county,” A.R.S. § 16-165(A)(11), or (2) “[i]f the county recorder receives credible information that a person has registered to vote in a different county” and the county recorder “confirm[s] the person’s voter registration with that other county,” A.R.S. § 16-165(B). The Felony Provision makes it a felony for anyone to “[k]nowingly provide[] a mechanism for voting to another person who is registered in another state.” A.R.S. § 16-1016(12). The statute does not define “mechanism for voting.” AARA, Voto Latino, and Priorities USA sued the Arizona Attorney General, the Secretary of State, and fifteen county recorders, challenging SB 1260 on preemption and constitutional grounds. AARA’s “mission is to ensure that retirees . . . have access to social and economic justice and full civil rights,” which includes ensuring its members can register to vote and meaningfully participate in Arizona elections. AARA regularly engages in voter-registration and voter-education efforts. Voto Latino’s “mission is to grow political engagement in historically underrepresented communities,” especially “young Latinx voters.” Priorities USA’s mission is to “engage Americans by educating, persuading, registering, and mobilizing citizens around issues and elections that affect their lives.” To fulfill these missions, the Organizations run public information campaigns to educate potential voters about why and how to register to vote in Arizona. AARA and Voto Latino also assist voters in registering to vote. The Yuma County Republican Committee intervened to defend the law. The Organizations moved to preliminarily enjoin the Cancellation and Felony Provisions. They claimed that the Cancellation Provision does not comply with the National AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 11

Voter Registration Act’s procedural requirements for canceling a voter’s registration. See 52 U.S.C. § 20507(a)(3)–(4), (d)(1). They also claimed that the Felony Provision is unconstitutionally vague and overbroad because it may criminalize voter registration and education. Agreeing with the Organizations, the district court preliminarily enjoined the enforcement of both provisions. The Arizona Attorney General timely appealed. A three- judge panel of our Court vacated the preliminary injunction. It held—over a dissent—that the Organizations lacked standing to challenge the Cancellation Provision and unanimously held that the Organizations failed to show a likelihood of success on the merits in their challenge to the Felony Provision. Ariz. All. for Retired Ams. v. Mayes, 117 F.4th 1165, 1185 (9th Cir. 2024), vacated and reh’g en banc granted, 130 F.4th 1177 (9th Cir. 2025). A majority of active judges voted to rehear the case en banc, vacating the panel opinion. Ariz. All., 130 F.4th 1177. We review for abuse of discretion the grant of a preliminary injunction, Am. Trucking Ass’ns, Inc. v. City of Los Angeles, 559 F.3d 1046, 1052 (9th Cir. 2009), but review de novo “[t]he district court’s interpretation of the underlying legal principles,” Sw. Voter Registration Educ. Project v. Shelley, 344 F.3d 914, 918 (9th Cir. 2003) (en banc) (per curiam). A plaintiff seeking a preliminary injunction must clearly show—not merely allege—that it is likely to prove each element of standing. Lopez v. Candaele, 630 F.3d 775, 785 (9th Cir. 2010); see also Murthy v. Missouri, 603 U.S. 43, 58 (2024).

II. The Organizations’ Article III standing depends on an injury-in-fact. Article III of the Constitution extends the judicial power of federal courts to “Cases” and “Controversies.” U.S. Const. art. III, § 1. “[A]n essential and unchanging part of the case-or-controversy requirement” is the doctrine of standing—the idea that a plaintiff can invoke the judicial power only to decide real disputes and not to answer abstract questions. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992). Standing reinforces the separation of powers between the judiciary and the political branches, see Spokeo, Inc. v. Robins, 578 U.S. 330, 338 (2016), and respects our federal system by “avoid[ing] undue federal interference with state government . . . [,]” 13A Wright & Miller’s Federal Practice & Procedure § 3531.3 n.22 (3d ed. 1998) (quoting Provo City Corp. v. Willden, 768 P.2d 455, 458 n.3 (Utah 1989)). There are three elements to standing: the plaintiff suffered an injury-in-fact, the defendant caused it, and the court can redress it. Lujan, 504 U.S. at 560–61. This standing inquiry is the same for individuals and for organizations. That is, “organizations must satisfy the usual standards for injury in fact, causation, and redressability that apply to individuals.” Hippocratic Med., 602 U.S. at 393–94 (citing Havens Realty Corp. v. Coleman, 455 U.S. 363, 378– 79 (1982)). “There is no question that an association may have standing in its own right to seek judicial relief from injury to itself and to vindicate whatever rights and immunities the association itself may enjoy.” Warth v. Seldin, 422 U.S. 490, 511 (1975). But just like any individual bringing a case in federal court, an organization needs a “direct stake in the outcome” of a case. Sierra Club v. Morton, 405 U.S. 727, 740 (1972). So what matters for standing is the nature of the injury, not the legal form of the AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 13

injured. See, e.g., Diamond Alt. Energy, LLC v. EPA, 606 U.S. 100, 113⁠–⁠14 (2025) (holding that business corporations producing fuel had standing to challenge regulations of automakers); Joint Anti-Fascist Refugee Comm. v. McGrath, 341 U.S. 123, 140–41 (1951) (holding that charitable organizations assisting refugees had standing to challenge allegedly defamatory statements). Organizations need the same standing as individuals to sue in federal courts, no more and no less.1 Hippocratic Med., 602 U.S. at 393–94 (citing Havens Realty, 455 U.S. at 378–79). The Organizations’ standing to challenge the Cancellation and Felony Provisions turns on whether they have suffered or would likely suffer an injury-in-fact: a concrete, particularized, and actual or imminent injury to a legally protected interest. See Lujan, 504 U.S. at 560. The Organizations claim that the Cancellation and Felony Provisions interfere with their voter-registration and voter- education efforts, but they present a different theory of standing for each. For the Cancellation Provision, we

1 An organization can also bring a claim on behalf of its members by showing “representational” (or “associational”) standing. See Hunt v. Wash. State Apple Advert. Comm’n, 432 U.S. 333, 342–43 (1977). An organization may have representational standing when: “(a) its members would otherwise have standing to sue in their own right; (b) the interests it seeks to protect are germane to the organization’s purpose; and (c) neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit.” Id. at 343. AARA argues that it has representational standing to challenge the Cancellation and Felony Provisions. But AARA has not identified a member who had a current voter registration canceled based on the Cancellation Provision or who intended to engage in protected conduct arguably prohibited by the Felony Provision. See Associated Gen. Contractors of Am., San Diego Chapter, Inc. v. Cal. Dep’t of Transp., 713 F.3d 1187, 1194–95 (9th Cir. 2013).

consider whether the law interferes with the Organizations’ core business activities, resulting in a concrete and particularized injury. For the Felony Provision, we consider whether the provision arguably regulates the Organizations’ conduct, resulting in a sufficiently imminent injury. The Organizations have made a clear showing of a likely injury- in-fact from the Felony Provision but have not made such a showing as to the Cancellation Provision. A. The Organizations lack standing to challenge the Cancellation Provision. The Organizations assert that the Cancellation Provision will interfere with their voter-registration and voter- education efforts. They contend that the law will cause county recorders to cancel voters’ new registrations instead of their old registrations when those voters register in a new county. The Organizations say that, to prevent the cancellation of new registrations, they must divert time and resources to ask voters about their past registrations and help voters cancel old registrations. They also allege that they must divert resources to educate voters on the harmful effects of the Cancellation Provision. To understand when an organization establishes a concrete and particularized injury-in-fact, we must look to the Supreme Court’s history of decisions in this area, culminating with its recent decision in Hippocratic Medicine. 602 U.S. at 393–96. 1. The doctrine of organizational injury. Effects on an entity’s core business activities have long been recognized as concrete and particularized injuries. In Village of Arlington Heights v. Metropolitan Housing Development Corp., the plaintiff nonprofit organization Metropolitan Housing Development Corporation (“MHDC”), a real estate developer, planned to build AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 15

moderate-income and low-income housing. 429 U.S. 252, 254 (1977). The Village of Arlington Heights denied a rezoning application that MHDC needed to build the housing. Id. MHDC sued, claiming the denial was racially discriminatory. Id. The Supreme Court held that MHDC had standing to challenge the denial of rezoning. Id. at 263. That denial stood “as an absolute barrier to constructing the housing MHDC had contracted” to build, which interfered with MHDC’s “interest in making suitable low-cost housing available in areas where such housing is scarce.” Id. at 261, 263. This injury was concrete, “not mere abstract concern about a problem of general interest,” and particular to MHDC’s specific project. Id. at 263. Five years later, the Court extended this analysis in Havens Realty. Plaintiff Housing Opportunities Made Equal (“HOME”) alleged that defendants’ racial steering practices “perceptibly impaired HOME’s ability to provide counseling and referral services for low-and moderate- income homeseekers.” 455 U.S. at 379. While the challenged conduct did not entirely prevent HOME from providing its counseling and referral services, “such concrete and demonstrable injury to the organization’s activities— with the consequent drain on the organization’s resources— constitute[d] far more than simply a setback to [HOME’s] abstract social interests.” Id. Although HOME could continue to counsel and refer its clients, the defendants’ conduct interfered with HOME’s ability to provide equal access to housing through its services. See id. Thus, the Court held that “there can be no question that the organization has suffered injury in fact” because of the defendants’ “concrete and demonstrable injury to the organization’s activities.” Id.

More recently in FDA v. Alliance for Hippocratic Medicine, the Supreme Court corrected a common misreading of Havens Realty. There, four pro-life medical associations challenged FDA’s relaxation of regulatory requirements for mifepristone, an abortion drug. 602 U.S. at 372–73, 376. The associations maintained that they had Article III standing because FDA “impaired their ability to provide services and achieve their organizational missions.” Id. at 394 (internal quotation marks omitted). They claimed that they “incurr[ed] costs to oppose FDA’s actions” by conducting their own studies on mifepristone’s risks and engaging in public advocacy and education. Id. The Supreme Court held that the associations lacked organizational standing to challenge FDA’s actions. It clarified that the mere diversion of resources “to gather information and advocate against the defendant’s action” that the organization opposes is not enough to create standing under Havens Realty. Id. at 394–95. So while the frustration of an organization’s mission with a consequent drain on resources sometimes may be a concrete and particularized injury, it is not always one. When an organization’s mission is frustrated just because it has a general objection to the challenged action, the organization has not suffered an injury-in-fact, even when it expends resources to voice that objection. See id. at 394. As the Court explained, pure issue advocacy cannot create a concrete or particularized injury because standing doctrine “screens out plaintiffs who might have only a general legal, moral, ideological, or policy objection to a particular” action. Id. at 381. General objections are not concrete because they reflect only abstract opposition, not a real harm to a legally protected interest. See id. Nor are they particularized because they do not specifically concern the AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 17

potential plaintiff, any more than a concerned bystander. See id. at 382. This is so even if that general objection arises from an organization’s mission because “standing is not measured by the intensity of the litigant’s interest or the fervor of his advocacy.” Valley Forge Christian Coll. v. Ams. United for Separation of Church & State, Inc., 454 U.S. 464, 486 (1982). Conversely, an organization has suffered an injury- in-fact if the frustration of its mission comes from an action that “directly affect[s] and interfere[s] with [its] core business activities.” Hippocratic Med., 602 U.S. at 395. To be clear, a “direct” effect does not mean, as the principal concurrence suggests, that an organizational plaintiff must be directly regulated by a challenged law. As the Court explained in Hippocratic Medicine, it “has identified a variety of familiar circumstances where government regulation of a third-party individual or business may be likely to cause injury in fact to an unregulated plaintiff.” Id. at 384. Whether an organizational plaintiff is unregulated by the action at issue usually goes to causation and redressability, not to injury. Id. at 385 n.2; see also Diamond Alt. Energy, 606 U.S. at 112 (“When the plaintiff is not the object of a government regulation . . . causation and redressability often depend on how regulated third parties not before the court will act in response to the government regulation or judicial relief.”); see, e.g., Bost v. Ill. State Bd. of Elections, 607 U.S. 71, 76⁠–⁠77 (2026) (citing Hippocratic Med., 602 U.S. at 379) (explaining that “an unlawful election rule [regulating ballot-counting, not candidates] can injure a candidate in several ways”). In this way, an injury that affects an organization’s core activities is “direct”—both concrete and particularized—because it results in a real harm that is specific to the organization. See

Arlington Heights, 429 U.S. at 263; Sierra Club, 405 U.S. at 740. It was not enough that the organizations in Hippocratic Medicine had diverted resources in response to the FDA’s actions. That is because “an organization that has not suffered a concrete injury caused by a defendant’s action cannot spend its way into standing simply by expending money to gather information and advocate against the defendant’s action.” Hippocratic Med., 602 U.S. at 394. Plaintiffs cannot manufacture standing in the absence of an independently sufficient injury-in-fact, even where that diversion is “costly and burdensome.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 415–16 (2013). Otherwise, organizations could spend their way to standing in challenges based solely on ideological opposition. See Hippocratic Med., 602 U.S. at 394–95. While a diversion of resources may be evidence of an injury, see Monsanto Co. v. Geertson Seed Farms, 561 U.S. 139, 153–54 (2010), it is not itself an injury-in-fact. Although the associations in Hippocratic Medicine invoked Havens Realty to argue that its diversion of resources could itself establish standing, the Court rejected that interpretation as “incorrect.” 602 U.S. at 395. Such a rule “would mean that all the organizations in America would have standing to challenge almost every federal policy that they dislike, provided they spend a single dollar opposing those policies.” Id. Havens Realty did not stand for that “expansive theory of standing.” Id. Prior to Hippocratic Medicine, we “read Havens to hold that an organization has direct standing to sue where it establishes that the defendant’s behavior has frustrated its mission and caused it to divert resources in response to that AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 19

frustration of purpose.” E. Bay Sanctuary Covenant v. Biden, 993 F.3d 640, 663 (9th Cir. 2021). But these earlier cases, decided without the benefit of Hippocratic Medicine, did not ask whether the challenged conduct “directly affected and interfered with” the plaintiff organizations’ core activities. 602 U.S. at 395. The organizational standing analyses in these cases are therefore overruled. See Nielsen v. Thornell, 101 F.4th 1164, 1170–71 (9th Cir. 2024); Fellowship of Christian Athletes v. San Jose Unified Sch. Dist. Bd. of Educ., 82 F.4th 664, 682–83 (9th Cir. 2023) (en banc); Sabra v. Maricopa Cnty. Cmty. Coll. Dist., 44 F.4th 867, 879–80 (9th Cir. 2022); E. Bay Sanctuary Covenant, 993 F.3d at 662–65; Nat’l Council of La Raza v. Cegavske, 800 F.3d 1032, 1040 (9th Cir. 2015); Fair Hous. Council of San Fernando Valley v. Roommate.com, LLC, 666 F.3d 1216, 1219 (9th Cir. 2012); El Rescate Legal Servs., Inc. v. Exec. Off. of Immigr. Rev., 959 F.2d 742, 748 (9th Cir. 1991).2 2. The Organizations have not clearly shown that the Cancellation Provision injures them. The Cancellation Provision states that “[i]f the county recorder receives credible information that a person has registered to vote in a different county, the county recorder shall confirm the person’s voter registration with that other county and, on confirmation, shall cancel the person’s registration.” A.R.S. § 16-165(B). The Organizations allege that under this provision “significant numbers of voters could be purged from the registration rolls and the early

2 In our order in Immigrant Defenders Law Center v. Noem, 145 F.4th 972 (9th Cir. 2025), we held that the plaintiff alleged organizational standing. To the extent that the Immigrant Defenders organizational standing analysis relied on East Bay Sanctuary, that analysis is overruled. Id. at 987⁠–⁠89.

voting lists without any notice or opportunity to contest their removal.” And they say that because the Cancellation Provision does not provide county recorders with a way to ensure that only a voter’s outdated registration is canceled, voters may be stripped of the new registration they intend to use for voting. The Organizations also say that they must reshape their voter education activities to “provide education and training on the harms of the Cancellation . . . Provision[].” For standing purposes, the Organizations contend that the provision will interfere with their voter-registration activities because they fear that the voters they help register might have their new registrations canceled. To prevent this harm, the Organizations argue, they must spend time and resources asking voters about their past registrations and assisting them with canceling those old registrations when they register in a new county. This fear cannot support standing because it “lacks record support and is highly speculative.” Hippocratic Med., 602 U.S. at 390; see id. at 381 (explaining that injury in fact “must be actual or imminent, not speculative”). Nothing in the Cancellation Provision arguably requires county recorders to cancel a voter’s new registration. And the record confirms that the state and counties have longstanding procedures in place to ensure that, if elections officials determine an individual is registered in multiple counties, only the old registration is canceled. The state’s Election Procedures Manual, which binds county recorders, codifies these procedures. See Sec’y Katie Hobbs, 2023 Elections Procedures Manual 25–26, 39 (2023); A.R.S. § 16-452(C). The Cancellation Provision aligns with this guidance, and the Organizations point to no evidence that elections officials will abandon it. Indeed, the Organizations conceded AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 21

at oral argument that there is no evidence in the record of the state canceling a voter’s new registration under these longstanding procedures.3 Without any showing of actual or imminent injury to the Organizations’ voter-registration activities, they are left only with the loss of time and resources they have spent manufacturing one. 602 U.S. at 394. But under Hippocratic Medicine, this is not a concrete injury. Id. And “an organization that has not suffered a concrete injury caused by a defendant’s action cannot spend its way into standing simply by expending money to gather information and advocate against the defendant’s action.” Id. The Organizations also contend that they must divert resources to educate voters and the public about the Cancellation Provision’s supposed effects. But without this harm to the Organizations’ voter-registration activities, their mere diversion of resources to educate and train voters on the potential effects of the Cancellation Provision cannot independently establish an injury-in-fact. As in Hippocratic Medicine, the Organizations’ expenditures to “better inform their members and the public about [the Cancellation Provision’s] risks” and to “engag[e] in public advocacy and public education” do not confer standing. 602 U.S. at 394. In the end, the Organizations have not clearly shown that it is likely that the provision affects or interferes with their core activities. So at this stage they have not established that

3 The principal concurrence mistakes this for a merits analysis. It is not. Determining whether a claimed injury is actual or imminent rather than speculative is exactly what Hippocratic Medicine requires. See, e.g., 602 U.S. at 381; Clapper, 568 U.S. at 401, 410 (holding that plaintiffs did not have standing because their asserted injury “rest[ed] on their highly speculative fear” that the challenged law operated as they claimed).

they suffered a sufficiently concrete and particularized injury-in-fact. Accordingly, the district court abused its discretion in preliminarily enjoining the Cancellation Provision. The principal concurrence, although it joins neither our analysis above nor the Conclusion below, agrees that the Organizations have not shown injury sufficient for standing to challenge the Cancellation Provision under Hippocratic Medicine. We part ways when the concurrence proceeds to opine on questions that this case has never presented. But the Supreme Court reminds us that “[c]ourts . . . do not, or should not, sally forth each day looking for wrongs to right. They wait for cases to come to them, and when cases arise, courts normally decide only questions presented by the parties.” See United States v. Sineneng-Smith, 590 U.S. 371, 376 (2020) (citation modified). Our duty is to decide the case before us, not the cases we may foresee. We trust that future panels will decide these questions faithfully when presented. B. The Organizations have standing to challenge the Felony Provision. The Felony Provision makes it a crime for any person to “[k]nowingly provide[] a mechanism for voting to another person who is registered in another state, including by forwarding an early ballot addressed to the other person.” A.R.S. § 16-1016(12). The Organizations argue that the Felony Provision is unconstitutionally vague and overbroad. They contend that the Felony Provision will chill their voter- registration and voter-mobilization activities because they worry that they could be prosecuted. If the Felony Provision does so, that kind of constitutional harm is a concrete and particularized injury. See TransUnion LLC v. Ramirez, 594 U.S. 413, 425 (2021) (“[T]raditional harms may also include harms specified by the Constitution itself.”). And because AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 23

the Organizations’ claims sound partly in the First Amendment, they “present unique standing considerations” that “tilt[] dramatically toward a finding of standing,” as the chilling of protected expression “is, itself, a constitutionally sufficient injury.” Libertarian Party of Los Angeles Cnty. v. Bowen, 709 F.3d 867, 870 (9th Cir. 2013) (quoting Ariz. Right to Life Pol. Action Comm. v. Bayless, 320 F.3d 1002, 1006 (9th Cir. 2003)). The Organizations have not been prosecuted under the Felony Provision, however, so we must also ask whether the pre-enforcement threat to their constitutional rights is sufficiently imminent. Plaintiffs have standing to challenge a law before it has been enforced against them when they show (1) “an intention to engage in a course of conduct arguably affected with a constitutional interest,” (2) that such conduct is “arguably . . . proscribed by a statute,” and (3) that “there exists a credible threat of prosecution” under the statute. Susan B. Anthony List v. Driehaus, 573 U.S. 149, 159 (2014) (quoting Babbitt v. United Farm Workers Nat’l Union, 442 U.S. 289, 298 (1979)). We have sometimes framed this inquiry in similar terms, asking “whether the plaintiffs have articulated a concrete plan to violate the law in question, whether the prosecuting authorities have communicated a specific warning or threat to initiate proceedings, and the history of past prosecution or enforcement under the challenged statute.” Thomas v. Anchorage Equal Rts. Comm’n, 220 F.3d 1134, 1139 (9th Cir. 2000) (en banc) (citation modified). We take this opportunity to clarify that we “adopt the Supreme Court’s framework.” Peace Ranch, LLC v. Bonta, 93 F.4th 482, 487 (9th Cir. 2024). Thus, we consider whether the plaintiffs intend to engage in conduct arguably protected by the Constitution, whether the law they challenge arguably

proscribes that conduct, and whether there is a credible threat of enforcement against the plaintiffs because of their conduct. Because the Organizations have clearly shown all three in their challenge to the Felony Provision, we hold that they have standing. 1. The Organizations engage in conduct arguably affected with a constitutional interest. The Organizations already engage in voter-registration and voter-mobilization efforts that the Constitution arguably protects. They claim that the Felony Provision is unconstitutionally vague under the Fourteenth Amendment because the undefined phrase “mechanism for voting” arguably encompasses a nearly limitless range of voting- related activities, including voter registration and voter mobilization. See City of Chicago v. Morales, 527 U.S. 41, 56 (1999). They further claim that the Felony Provision is unconstitutionally overbroad because by sweeping in registration and mobilization, it implicates “core political speech” protected by the First Amendment. See United States v. Williams, 553 U.S. 285, 292 (2008). Defendants do not contest that core political speech is involved. Encouraging someone to register to vote may involve “interactive communication concerning political change” similar to soliciting signatures on an initiative petition. Pierce v. Jacobsen, 44 F.4th 853, 859 (9th Cir. 2022) (quoting Meyer v. Grant, 486 U.S. 414, 422 (1988)). This means that the Organizations’ conduct is arguably affected with constitutional interests. 2. The Organizations’ conduct is arguably proscribed by the Felony Provision. The Felony Provision’s criminal ban on “knowingly provid[ing] a mechanism for voting to another person who AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 25

is registered in another state” arguably covers the Organizations’ voter registration and mobilization efforts. A.R.S. § 16-1016(12). For example, the provision could be read to include their initiatives to sign up voters and their websites that provide links allowing Arizona voters to request a ballot online. Outside of its more common usage to describe machinery, mechanism may mean “[a]n instrument or a process, physical or mental, by which something is done or comes into being.” Mechanism, American Heritage 1090 (5th ed. 2016); see also Mechanism, Merriam-Webster Online Dictionary (last visited July 31, 2026), https://www.merriam-webster.com/dictionary/mechanism [https://perma.cc/4JF4-G55Z] (“a process, technique, or system for achieving a result”). Thus, the Felony Provision arguably criminalizes knowingly providing someone with a process for voting. The Organizations help Arizonans register to vote, a necessary step in Arizona’s voting process. Their broad reading of “voting mechanism” to encompass voter registration is arguable and supports a pre-enforcement challenge. 3. The Organizations face a credible threat of prosecution. The Organizations also face a credible threat of prosecution. In a credible threat analysis, we consider any past enforcement of the challenged statute and the potential for future enforcement, including whether defendants have disavowed enforcement. Driehaus, 573 U.S. at 164–66; cf. Thomas, 220 F.3d at 1139 (considering a specific threat to enforce, or a history of enforcing, the challenged law). We also weigh whether the gravity of the threat is “sufficient to justify pre-enforcement review,” given the burden of enforcement proceedings and severity of sanctions available under the statute. Driehaus, 573 U.S. at 165. In other words,

we assess a credible threat by looking to the risk of enforcement based on past and future enforcement, and the burden of enforcement, including the enforcement process and possible punishment. No single factor is dispositive. See LSO, Ltd. v. Stroh, 205 F.3d 1146, 1155 (9th Cir. 2000). We begin by analyzing the burden of enforcement. Any felony prosecution gives rise to severe procedural burdens. See Dombrowski v. Pfister, 380 U.S. 479, 487 (1965) (noting that a chilling effect on the exercise of First Amendment rights “may derive from the fact of the prosecution” of a felony). Such burdens on “electoral speech are of particular concern,” including the diversion of time and resources “in the crucial days leading up to an election.” Driehaus, 573 U.S. at 165. The Felony Provision also is punishable by imprisonment of six months or more. A.R.S. § 13-702(D). Thus, the burden that enforcement of the Felony Provision would impose on the Organizations through their staff and members is substantial. Next we consider the risk that the Organizations will bear that burden. There is no prior enforcement here because the district court enjoined the Felony Provision one day after it took effect. So “the history of past enforcement carries little, if any weight.” Arizona v. Yellen, 34 F.4th 841, 850 (9th Cir. 2022); see Isaacson v. Mayes, 84 F.4th 1089, 1099 (9th Cir. 2023) (explaining that plaintiffs did not need to show a history of past enforcement because the challenged laws were recently enacted and then enjoined). Looking forward, the Attorney General disavowed future enforcement in this litigation. Counsel for the office told the district court that Organizations “are at no danger of prosecution for registration activities” because “mechanism for voting” means a “ballot and a ballot affidavit envelope and nothing else.” This disavowal in district court is the AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 27

Attorney General’s primary argument against the Organizations’ standing. A disavowal is “where the enforcing authority expressly interpreted the challenged law as not applying to the plaintiffs’ activities.” Lopez, 630 F.3d at 788. But while a “failure to disavow enforcement is sufficient to establish a credible threat of prosecution,” at least in First Amendment challenges, Matsumoto v. Labrador, 122 F.4th 787, 797 (9th Cir. 2024) (citation modified), the converse is not necessarily true, see Ind. Right to Life Victory Fund v. Morales, 66 F.4th 625, 631 (7th Cir. 2023) (“noting that state officials’ promises” not to enforce a statute receive less weight” than an expressed intent to enforce a challenged law, “especially when they cannot bind their successors in office”). For example, we have credited disavowals when they confirm that a plaintiff’s reading of the challenged statute is not arguable, or they bolster a long history of nonenforcement. See, e.g., Johnson v. Stuart, 702 F.2d 193, 195 (9th Cir. 1983) (crediting disavowal when challenged law, read in context, “was not intended to regulate the conduct of [plaintiffs]”); Foothills Christian Ministries v. Johnson, 148 F.4th 1040, 1050 (9th Cir. 2025) (crediting disavowal when the challenged law had “been on the books for over 40 years” and plaintiffs failed to identify a single instance of enforcement in the challenged circumstances). We have declined to credit disavowals, however, when the disavowal does not bind either the defendant or other potential enforcers and there is no corroborating history of nonenforcement. See, e.g., Am. Encore v. Fontes, 152 F.4th 1097, 1118–19 (9th Cir. 2025) (holding that the attorney general’s disavowal of challenged guidance did not bind the secretary of state’s authority over election workers); Isaacson, 84 F.4th at 1100–01 (holding that even if the

attorney general disavows enforcement, the possibility of at least one county attorney enforcing the challenged law “is reason enough” to establish credible threat); Yellen, 34 F.4th at 850–51 (holding that the defendant’s disavowal of enforcement of the challenged law “in the way feared by” plaintiffs did not preclude enforcement, which remained “sufficiently realistic and credible to confer standing under Driehaus”). We have also rejected disavowals that decline enforcement as a “mere litigation position.” Lopez, 630 F.3d at 788; see Am.-Arab Anti-Discrim. Comm. v. Thornburgh, 970 F.2d 501, 508 (9th Cir. 1991) (declining to credit dropped charges as a disavowal, when charges were dropped “not because they were considered inapplicable, but for tactical reasons”). Our approach follows federal courts’ general reluctance to allow a defendant’s disavowal to defeat a constitutional challenge. Even on the merits, courts do not “accept[] as ‘authoritative’ an Attorney General’s interpretation of state law when ‘the Attorney General does not bind the state courts or local law enforcement authorities.’” Stenberg v. Carhart, 530 U.S. 914, 940 (2000) (quoting Virginia v. Am. Booksellers Ass’n, 484 U.S. 383, 395 (1988)); see also United States v. Stevens, 559 U.S. 460, 480 (2010) (“We would not uphold an unconstitutional statute merely because the Government promised to use it responsibly.”). Nor do we allow a government’s voluntary change in policy to moot a case if it remains “free to return to [its] old ways.” Fikre v. FBI, 904 F.3d 1033, 1039 (9th Cir. 2018); cf. Finstuen v. Crutcher, 496 F.3d 1139, 1150–51 (10th Cir. 2007) (rejecting attempt by “an executive agency to moot an appeal by virtue of a statutory interpretation advanced only in the context of adversarial litigation, rather than a rule-making or other appropriate mechanism authorized by the legislature”). AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 29

So too for standing. See, e.g., Vt. Right to Life Comm., Inc. v. Sorrell, 221 F.3d 376, 383 (2d Cir. 2000) (“[T]here is nothing that prevents the State from changing its mind. It is not forever bound, by estoppel or otherwise, to the view of the law that it asserts in this litigation.”); N.C. Right to Life, Inc. v. Bartlett, 168 F.3d 705, 711 (4th Cir. 1999) (noting that even though a state board of elections disavowed enforcement against plaintiff organization, “[i]t has no guarantee that the Board might not tomorrow bring its interpretation more in line with the provision’s plain language,” thus giving rise to a “reasonable fear” of prosecution); Turtle Island Foods, S.P.C. v. Strain, 65 F.4th 211, 218 (5th Cir. 2023) (holding that the state’s “disclaimed intent to penalize . . . is not compelling” to defeat threat of enforcement because “nothing binds the Commissioner here—he (or a future holder of his office) could change his mind”); Ind. Right to Life Victory Fund, 66 F.4th at 631 (“It is only when a state agency acknowledges that it will not enforce a statute because it is plainly unconstitutional that [disavowals] might mean anything at all.” (citation modified)). When, for standing purposes, a plaintiff has already shown that the challenged law arguably applies to its conduct, a defendant’s disavowal must do more than deny that interpretation. The Arizona Attorney General’s district-court disavowal does not close the courthouse doors to the Organizations. The Felony Provision remains on the books, and no state court has construed it as inapplicable to the Organizations’ conduct. As the Attorney General concedes, the disavowal does not bind successors in that office, or current or future county prosecutors. No prosecutor has sought, nor has the Attorney General issued, even a nonbinding opinion narrowing the Felony Provision’s arguable scope. See

A.R.S. § 41-193(A)(7); Yes on Prop 200 v. Napolitano, 160 P.3d 1216, 1225 (Ariz. Ct. App. 2007) (holding that attorney general opinions are not legally binding on Arizona state courts or other officials). The Attorney General also argues in a footnote that its disavowal here likely would preclude a state prosecution by judicial estoppel. But speculation that the equities may tilt toward dismissal of a future prosecution does little to mitigate the Organizations’ fear of being subject to enforcement of the Felony Provision in the first place. And, as the Organizations point out, the Attorney General has not formally stipulated that the Felony Provision is unenforceable in any way. Thus, the Attorney General’s bare disavowal simply contradicts our holding that the Provision arguably applies to the Organizations’ voter- registration and voter-mobilization efforts. Even if the Attorney General could credibly commit to nonenforcement by that office, “the universe of potential complainants is not restricted to state officials.” Driehaus, 573 U.S. at 164. As in most states, independent county attorneys in Arizona have original jurisdiction to “conduct all prosecutions for public offenses.” A.R.S. § 11-532(A)(1). The Attorney General argues that such prosecutions under the Felony Provision are speculative because the Organizations did not name county attorneys as parties. Relying on Driehaus, we have rejected the argument that injury-in-fact is “short-circuited by the fact that there are multiple authorized enforcers of the statute” who are not named defendants. Matsumoto, 122 F.4th at 798. Requiring plaintiffs to name as a defendant all potential enforcers of the challenged law would invite “judicial scrutiny of every defendant prosecutor’s enforcement record, office policies, budgetary constraints, and a multitude of other considerations.” Id. at 798 n.7. It is the credibility of the AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 31

threat that plaintiffs will be prosecuted under the law, not the credibility of any particular prosecutor’s commitment to enforce the law against the plaintiffs, that informs whether a fear of prosecution for arguably protected conduct is sufficiently imminent to confer standing. Thus, a credible threat does not require plaintiffs to sue every person who might invoke a facially unconstitutional law against them. See Driehaus, 573 U.S. at 164. In sum, because the Organizations intend to engage in arguably protected conduct, the Felony Provision arguably proscribes that conduct, and there is a credible threat of enforcement against the Organizations, we hold that they have clearly shown that they are likely to suffer an injury-in- fact sufficient for Article III standing in their challenge to the Felony Provision. We also hold that the Organizations have met the remaining standing requirements. The chill to their expressive activities is an injury “fairly traceable” to the Attorney General’s potential enforcement of the Felony Provision against them. See Ass’n of Irritated Residents v. EPA, 10 F.4th 937, 943 (9th Cir. 2021) (“An injury is fairly traceable to a challenged action as long as the links in the proffered chain of causation are not hypothetical or tenuous and remain plausible.” (citation modified)). Their injury is also redressable because a preliminary injunction prohibiting the Attorney General’s enforcement of the Felony Provision would reduce the risk of prosecution by that office and thus alleviate the chill to the Organizations’ expression. See Matsumoto, 122 F.4th at 801 (“Where a state statute specifically grants enforcement powers to multiple government authorities, an injunction against the exercise of those powers by any one of those authorities suffices to establish redressability.”); see also Ass’n of Irritated

Residents, 10 F.4th at 944 (“[A] plaintiff can meet the redressability requirement by showing that it is likely, although not certain, that his injury can be redressed by a favorable decision.” (citation modified)). Again, “when the threatened enforcement effort implicates First Amendment rights, the inquiry tilts dramatically toward a finding of standing.” Stroh, 205 F.3d at 1155. Thus, the Organizations have standing to challenge the Felony Provision. III. The Organizations are not likely to succeed on the merits of their Felony Provision claim. Still, the Organizations are not likely to succeed on the merits of their challenge because their interpretation of the Felony Provision, while arguable for standing purposes, is wrong. Plaintiffs seeking a preliminary injunction must establish that: (1) they are “likely to succeed on the merits,” (2) they are “likely to suffer irreparable harm in the absence of preliminary relief,” (3) “the balance of equities tips in [their] favor,” and (4) “an injunction is in the public interest.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). Because the Felony Provision does not encompass voter registration and mobilization, the Organizations are unlikely to succeed on the merits of their claim. Thus, the district court abused its discretion by preliminarily enjoining the Felony Provision. The Organizations’ vagueness and overbreadth challenges both rest on whether the statutory phrase “mechanism for voting” includes voter registration and mobilization. A state’s own courts have the last word on what that state’s laws mean. See Cal. Tchrs. Ass’n v. State Bd. of Educ., 271 F.3d 1141, 1146 (9th Cir. 2001). Where, as here, a state court has not issued an authoritative interpretation of a statute, we look to that state’s traditional AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 33

tools of statutory construction to determine the statute’s “allowable meaning.” Id. at 1147 (quoting Grayned v. City of Rockford, 408 U.S. 104, 110 (1972)). Because “mechanism for voting” is not defined by the statute, we must give the words “their ordinary meaning.” Arizona ex rel. Brnovich v. Maricopa Cnty. Cmty. Coll. Dist. Bd., 416 P.3d 803, 805 (Ariz. 2018) (quoting Arizona v. Miller, 413 P.2d 757, 765 (Ariz. 1966)). In doing so we construe the phrase “in view of the entire text, considering the context.” Nicaise v. Sundaram, 432 P.3d 925, 927 (Ariz. 2019). The Organizations’ reading of the Felony Provision’s criminal ban on knowingly providing a “mechanism for voting” to include a “process” for voting like voter registration is arguable. In context, however, the better reading of “mechanism for voting” is an “instrument” or “technique” for voting, like a ballot. Mechanism, American Heritage 1090 (5th ed. 2016); see also Mechanism, Merriam- Webster Online Dictionary (last visited July 31, 2026), https://www.merriam-webster.com/dictionary/mechanism [https://perma.cc/4JF4-G55Z]. So it likely does not include voter registration and mobilization. First, in the context of the Felony Provision, “mechanism for voting” cannot sensibly include everything a person must do in preparation for voting. The process for registering to vote is a separate process from casting a vote, even though it is a necessary step to voting in Arizona. See A.R.S. § 16-122. Under the Organizations’ reading, “mechanism for voting” could include any number of precursors in the voting process, such as moving to Arizona with the intent to vote there. This is unlikely. And the only example of “mechanism for voting” the Felony Provision provides is an instrument, not a process: “an early ballot.”

A.R.S. § 16-1016(12). So “mechanism for voting” likely refers only to instruments used to cast a vote. Second, the section as a whole targets “illegal voting,” not a broader set of predicate steps to voting like registration. A.R.S. § 16-1016. The section criminalizes twelve distinct actions, all of which implicate only voting itself, such as voting without eligibility (subsection 1); voting more than once (subsections 2–4); and tampering with the ballot, ballot box, poll list, or vote totals (subsections 5–11). A.R.S. § 16- 1016(1)–(12). None of the listed actions implicate conduct before voting, such as registration, so “mechanism for voting” should be interpreted consistently and include only a ballot itself and a ballot affidavit envelope. See Schreiber v. Burlington N., Inc., 472 U.S. 1, 8 (1985) (“[W]ords grouped in a list should be given related meaning.” (quoting Secs. Indus. Ass’n v. Bd. of Governors of Fed. Rsrv. Sys., 468 U.S. 207, 218 (1984))). The Organizations are unlikely to succeed on the merits because the Felony Provision does not criminalize voter registration and mobilization. Thus, the district court abused its discretion by granting a preliminary injunction. IV. Conclusion We hold that the Organizations lack standing to challenge the Cancellation Provision. Hippocratic Medicine reaffirmed that standing doctrine is the same for organizations and individuals. An organizational plaintiff must establish that the challenged provision “directly affect[s] and interfere[s]” with its core business activities to establish a concrete and particularized injury. The Organizations have not clearly shown that the Cancellation Provision is likely to do so. At this stage, they have not established an injury-in-fact. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 35

We also hold that the Organizations have standing to challenge the Felony Provision. Under the Organizations’ interpretation, the Felony Provision arguably proscribes voter registration and mobilization. And the Attorney General’s disavowal of that interpretation does not defeat a credible threat of prosecution. But the Organizations are unlikely to succeed on the merits of their challenge to the Felony Provision. The plain meaning and statutory context of the Felony Provision confirm that “mechanism for voting” likely includes only a ballot itself and a ballot affidavit envelope. Because the Organizations have not clearly shown that they have standing to challenge the Cancellation Provision, and their challenge to the Felony Provision is not likely to succeed on the merits, we vacate the preliminary injunction and remand for further proceedings. VACATED AND REMANDED.

BRESS, Circuit Judge, with whom CALLAHAN, BENNETT, R. NELSON, and VANDYKE, Circuit Judges, join, concurring in parts II.B and III and concurring in the judgment: This case presented the en banc court with the opportunity to clarify our law of organizational standing following the Supreme Court’s decision in FDA v. Alliance for Hippocratic Medicine, 602 U.S. 367 (2024). Today’s majority opinion reaches the right result, holding that certain public interest organizations lack Article III standing to challenge A.R.S. § 16-165(A)(11), (B), an Arizona voter registration law known as the “Cancellation Provision.” The majority opinion also correctly overrules our past precedents—which expansively conferred organizational standing—as incompatible with Hippocratic Medicine. But while doing much that is good and right, the majority opinion fails to complete the mission. And in the process, it leaves confusion and uncertainty in its wake. The majority refuses to engage fully with the core reasoning of Hippocratic Medicine, including by not acknowledging key parts of that decision. And in explaining why the voter advocacy organizations here lack Article III standing, the majority introduces an unfamiliar form of analysis that keys standing to the merits, directly contrary to Supreme Court and Ninth Circuit precedent. While purporting to follow Hippocratic Medicine, the majority opinion in fact undermines it. We had the chance in this case to show lower courts how to conduct the organizational standing analysis post- Hippocratic Medicine. Today’s decision instead replaces our misguided organizational standing precedents with yet another misguided form of analysis that fails to respect AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 37

Supreme Court precedent and leaves our law of organizational standing as nebulous as we found it. Where Hippocratic Medicine sought to rein in organizational standing, the majority opinion yet again loosens the slack. As to the Cancellation Provision, I can therefore concur only in the judgment. I write separately to explain where the majority opinion goes awry and how this issue should have been resolved. I A “Plaintiffs bear the burden of establishing each of the three ‘irreducible’ elements of Article III standing.” Winsor v. Sequoia Benefits & Ins. Servs., LLC, 62 F.4th 517, 523 (9th Cir. 2023) (quoting Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992)). To this end, plaintiffs “must sufficiently allege ‘(i) that [they] suffered an injury in fact that is concrete, particularized, and actual or imminent; (ii) that the injury was likely caused by the defendant; and (iii) that the injury would likely be redressed by judicial relief.’” Id. (alteration in original) (quoting TransUnion LLC v. Ramirez, 594 U.S. 413, 423 (2021)). This case involves the question of when an organization has Article III standing to sue in its own right, on behalf of the organization itself, for injuries to the organization. See, e.g., Students for Fair Admissions, Inc. v. President & Fellows of Harvard Coll., 600 U.S. 181, 199 (2023) (explaining that “where the plaintiff is an organization, the standing requirements of Article III can be satisfied” by an organization “claim[ing] that it suffered an injury in its own right”). The Supreme Court has said that “organizations must satisfy the usual standards for injury in fact, causation,

and redressability that apply to individuals.” Hippocratic Med., 602 U.S. at 393–94. This is a point of law that the Supreme Court has long embraced. See, e.g., Havens Realty Corp. v. Coleman, 455 U.S. 363, 378 (1982) (explaining that to determine whether an organization has Article III standing, “we conduct the same inquiry as in the case of an individual”). Alas, lower courts turned the strictures of Article III into something much different—and much more permissive—for organizational plaintiffs. Pre-Hippocratic Medicine, in the Ninth Circuit and elsewhere, it was commonplace for lower courts to find organizational standing so long as the organization made two easy-to-satisfy showings: (1) the challenged law or policy frustrated the organization’s mission; and (2) the organization diverted resources in response to that frustration of purpose. See, e.g., Fellowship of Christian Athletes v. San Jose Unified Sch. Dist. Bd. of Educ., 82 F.4th 664, 682–83 (9th Cir. 2023) (en banc); Fair Hous. of Marin v. Combs, 285 F.3d 899, 905 (9th Cir. 2002); E. Bay Sanctuary Covenant v. Biden, 993 F.3d 640, 663 (9th Cir. 2021); Sabra v. Maricopa Cnty. Cmty. Coll. Dist., 44 F.4th 867, 879–80 (9th Cir. 2022); see also, e.g., PETA v. U.S. Dep’t of Agric., 797 F.3d 1087, 1094 (D.C. Cir. 2015); Tennessee Conf. of NAACP v. Lee, 139 F.4th 557, 563 (6th Cir. 2025) (collecting cases). Our adulteration of the Article III standards for organizational plaintiffs was effectively limitless, allowing any organization to challenge any law or policy within the organization’s general area of concern, so long as the organization could plausibly allege it had expended resources to address the issue or its effects. Because that is, of course, what an organization dedicated to a cause does, organizations enjoyed prefabricated standing under our case AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 39

law. See Fair Hous. Council of San Fernando Valley v. Roommate.com, LLC, 666 F.3d 1216, 1224, 1226 (9th Cir. 2012) (Ikuta, J., concurring and dissenting) (questioning how organizations could be “‘injured’ for standing purposes by the very expenses that advanced their mission”). Take a simple example. The government changes the asylum laws to make them less favorable to asylum-seekers. An organization dedicated to helping immigrants gain asylum sues to challenge the new rule. Because the new policy makes asylum more difficult to obtain, the organization claims it will need to spend more to help its clients, at the expense of other endeavors. Presto. The organization can sue, in its own right, to challenge the new policy. See E. Bay Sanctuary Covenant, 993 F.3d at 663– 64. There would even be Article III standing if the organization developed “new education and outreach campaigns” in response to the defendant’s actions. Sabra, 44 F.4th at 879 (quoting Fair Hous. Council, 666 F.3d at 1219). Our permissive standing jurisprudence could be summed up as follows: Challenged Law or Policy + Organizational Plaintiff = Article III Standing. This framework greatly eased the path for challenges to governmental actions, because organizations are ready-made litigants. They lack the complications associated with claims brought by individual plaintiffs who are actually subject to the challenged laws (lack of standing, mootness, unwillingness to be part of a lawsuit, and so on). Organizations could avoid all these standard impediments through the expedient of suing over the very issues they were created to pursue. The same story played out in case after case, ensuring that if a law or policy was enacted, there would be an

organization with standing to challenge it, right out of the gate. We conferred Article III standing on “any entity with a policy position and a dollar.” CASA de Maryland, Inc. v. Trump, 971 F.3d 220, 239–40 (4th Cir. 2020), vacated for reh’g en banc, 981 F.3d 311 (4th Cir. 2020) (dismissed Mar. 11, 2021) (Wilkinson, J.). As our revered late colleague Judge Ikuta observed, this circuit’s approach to organizational standing “look[ed] suspiciously” as though it permitted an organization to sue over “a harm that is simply a setback to the organization’s abstract social interests”—the “very thing” that Article III standing is supposed to guard against. Fair Hous. Council, 666 F.3d at 1226 (Ikuta, J., concurring and dissenting) (quotations and citation omitted).1 Our lax formulation for organizational standing was derived from a misreading of the Supreme Court’s decision in Havens Realty Corp. v. Coleman, 455 U.S. 363 (1982). In Havens, the Supreme Court held that a housing counseling organization, Housing Opportunities Made Equal (HOME), and one of its employees, Coleman, had standing to bring a claim against housing provider Havens Realty under the Fair Housing Act (FHA). Id. at 368, 374, 378–79. HOME alleged that the defendants engaged in prohibited racial steering by lying to Coleman (who was Black) about apartment availability. Id. at 366 & n.1, 367–68. Importantly, the FHA “conferred on all ‘persons’ a legal right to truthful information about available housing,” making it unlawful for any individual or firm “[t]o represent to any person because of race, color, religion, sex, or national origin that any dwelling is not available for

1 Judge Ikuta was originally a member of this en banc panel. Upon her passing, another judge was drawn to replace her. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 41

inspection, sale, or rental when such dwelling is in fact so available.” Id. at 373 (alteration in original) (quoting 42 U.S.C. § 3604(d)). “Person[s]” under the statute included “associations.” 42 U.S.C. § 3602(d); see also PETA, 797 F.3d at 1100 (Millett, J., dubitante); CASA de Maryland, 971 F.3d at 239. After determining that Coleman had suffered a “specific injury” from the defendants’ provision of misinformation, the Supreme Court held that HOME likewise had standing to bring its FHA claim. Havens, 455 U.S. at 373–74, 378– 79. Parsing HOME’s complaint, the Supreme Court explained that HOME’s alleged injury was based on defendants’ conduct that directly frustrated HOME’s “efforts to assist equal access to housing through counseling and other referral services,” requiring HOME “to devote significant resources to identify and counteract” the racial steering practices directed against it. Id. at 379. Under those circumstances, the Court held that “there can be no question that the organization has suffered injury in fact” because the defendants’ “steering practices have perceptibly impaired HOME’s ability to provide counseling and referral services for low-and moderate-income homeseekers.” Id. As Judge Millett of the D.C. Circuit has cogently explained, the result in Havens was “unsurprising” and “makes sense,” because “[f]ederal law vested HOME with a specific legal right to truthful, non-discriminatory housing information, and Havens Realty’s racially disparate misinformation targeted HOME along with the individuals it was aiding.” PETA, 797 F.3d at 1100 (Millett, J., dubitante). Havens Realty’s racist misinformation, delivered to a HOME employee, specifically impaired HOME’s own counseling services. See id. (“Put simply, what HOME used its own resources, information, and client

base to build up, Havens Realty’s racist lies tore down.”); id. (“Havens Realty’s racially disparate misinformation targeted HOME along with the individuals it was aiding.”). And “[w]hile the Court noted that HOME needed to divert resources as a result of these discriminatory practices, it cast HOME’s injury in terms of its ability to function.” CASA de Maryland, 971 F.3d at 239 (emphasis in original); see also Fair Hous. Council, 666 F.3d at 1225 (Ikuta, J., concurring and dissenting). Nevertheless, despite the decision’s clear limiting features, we broadly “read Havens to hold that an organization has direct standing to sue where it establishes that the defendant’s behavior has frustrated its mission and caused it to divert resources in response to that frustration of purpose.” E. Bay Sanctuary Covenant, 993 F.3d at 663. The Supreme Court firmly put an end to that excursion in FDA v. Alliance for Hippocratic Medicine, 602 U.S. 367 (2024). There, and as relevant here, pro-life medical associations sued FDA for relaxing the regulatory requirements for the abortion drug mifepristone. Id. at 372– 74. Although the new rules made it easier for doctors to prescribe mifepristone, the associations neither prescribed nor used the drug, and the FDA was “not requiring them to do or refrain from doing anything.” Id. at 374; see also id. at 385–86. But the associations contended that they had standing to challenge FDA’s action under Havens Realty. Id. at 394–95. In familiar fashion, the associations argued that FDA “impaired their ability to provide services and achieve their organizational missions,” and that the associations “incurr[ed] costs to oppose FDA’s actions,” including by conducting studies, drafting citizen petitions, and engaging in public advocacy and education in response to the FDA’s new policy. Id. at 394 (quotations omitted). In the associations’ view, it was sufficient under Havens Realty AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 43

that they had “divert[ed] [their] resources in response to [the] defendant’s actions.” Id. at 395. The medical associations in Hippocratic Medicine easily satisfied our court’s relaxed test for organizational standing. But the Supreme Court squarely rebuffed the associations’ (and our) improperly “expansive” reading of Havens Realty. Id. Such an interpretation, Hippocratic Medicine held, would “mean that all the organizations in America would have standing to challenge almost every federal policy that they dislike, provided they spend a single dollar opposing those policies.” Id. That understanding of Havens was “incorrect” because “an organization that has not suffered a concrete injury caused by a defendant’s action cannot spend its way into standing simply by expending money to gather information and advocate against the defendant’s action.” Id. at 394–95. Central to its reasoning, the Supreme Court in Hippocratic Medicine provided careful explication of Havens Realty, confirming the limits of that earlier decision (which should have been apparent long ago). As the Court explained, “[c]ritically, HOME not only was an issue- advocacy organization, but also operated a housing counseling service.” Id. at 395. So “when Havens gave HOME’s employees false information about apartment availability,” “Havens’s actions directly affected and interfered with HOME’s core business activities—not dissimilar to a retailer who sues a manufacturer for selling defective goods to the retailer.” Id. That was not the “kind of injury that the medical associations” alleged in Hippocratic Medicine, for FDA’s actions did “not impose[] any similar impediment to the medical associations’ advocacy businesses.” Id. Hippocratic Medicine could not have been clearer about the reach of Havens Realty,

cautioning that “Havens was an unusual case, and this Court has been careful not to extend the Havens holding beyond its context.” Id. at 396. B Applying Hippocratic Medicine to the case before us is straightforward. The plaintiffs are three public interest organizations that engage in voter registration, education, and outreach. They claim that Arizona’s Cancellation Provision violates federal election law, which imposes certain notice requirements before a state or political subdivision can remove a registered voter from its voting rolls. The Cancellation Provision provides that “[i]f the county recorder receives credible information that a person has registered to vote in a different county, the county recorder shall confirm the person’s voter registration with that other county and, on confirmation, shall cancel the person’s registration.” A.R.S. § 16-165(B). The plaintiff organizations fear that the Cancellation Provision will improperly lead to the cancellation of new voter registrations, without sufficient notice. To address this feared harm, the organizations contend that they will have to divert resources from other organizational activities to educate voters, provide trainings, and double-check voting registrations. These allegations plainly do not establish organizational standing after Hippocratic Medicine. The organizations have not alleged (and cannot allege) that the Cancellation Provision “directly” interferes with their “core business activities” of voter registration and mobilization in a way that is analogous to HOME’s injury in Havens Realty. See Hippocratic Med., 602 U.S. at 395. Unlike in Havens, the organizations here allege direct injuries to third parties, AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 45

namely, the voters who are their clients and whom they serve. The Cancellation Provision does not regulate or operate upon the organizations themselves. See A.R.S. § 16- 165(A)(11), (B). Nor does the Cancellation Provision regulate voters in a way that would directly injure the organizations as organizations within the meaning of Hippocratic Medicine (imagine, for example, a law that disallowed voters from speaking with public interest organizations, which would directly injure both the voters and the organizations in their organizational capacities). See Hippocratic Med., 602 U.S. at 384 (noting that “the Court has identified a variety of familiar circumstances where government regulation of a third-party individual or business may be likely to cause injury in fact to an unregulated plaintiff”). Instead, all the plaintiff organizations have alleged (and could allege) here is that the Cancellation Provision frustrates their organizational missions of enhancing voter registration, and that they have had to divert resources in responding to it. That theory used to be sufficient to establish organizational standing in this circuit. After Hippocratic Medicine, that is no longer the case. In assessing whether the Cancellation Provision directly affects or interferes with the organizations’ core business activities, it is therefore not enough to say that these organizations’ core business activities are education, outreach, and mobilization, and that the Cancellation Provision will lead it to engage in those efforts. The same could have been said in Hippocratic Medicine. But that was insufficient to confer standing there, and it is insufficient to confer standing here. Otherwise, “all the organizations in America would have standing to challenge almost every . . . policy that they dislike, provided they spend a single dollar

opposing those policies.” Hippocratic Med., 602 U.S. at 395. As the above discussion shows, the majority opinion is incorrect in asserting that I would allow organizational standing only when an organization challenges a law that directly regulates the organization. Maj. Op. 17. As I have indicated, a law may regulate third parties yet may directly affect the organization’s core business activities. See Hippocratic Med., 602 U.S. at 384. But that is not the case here. The plaintiff organizations’ alleged injuries differ critically from the injury in Havens Realty, which explains why Article III standing was established there but is lacking here. The injury in Havens was an injury to the organization in its functioning as an organization, because “[f]ederal law vested HOME with a specific legal right to truthful, non- discriminatory housing information, and Havens Realty’s racially disparate misinformation targeted HOME along with the individuals it was aiding.” PETA, 797 F.3d at 1100 (Millett, J., dubitante). The defendants’ actions in Havens impeded the organization from doing the very thing it was supposed to do as a counseling organization. It was in that specific sense that Havens’s actions “directly affected and interfered with HOME’s core business activities,” analogous to “a retailer who sues a manufacturer for selling defective goods to the retailer.” Hippocratic Med., 602 U.S. at 395. The nature of the injury to the organizational plaintiff in Havens was precisely what made it “an unusual case” of organizational standing. Id. at 396. Unlike in Havens, the plaintiff organizations here are challenging a law that allegedly harms others (voters) and that will allegedly result in harm to the organizations only AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 47

through the diversion of the organizations’ resources. The Cancellation Provision does not impede the plaintiff organizations’ functioning, but simply requires them to do more of the advocacy work they were created to do. Nothing in the Cancellation Provision “directly impairs [the organizations’] ability to provide counseling, referral, or other services” to voters. CASA de Maryland, 971 F.3d at 239. Instead, as in Hippocratic Medicine, the organizations here “are unregulated parties who seek to challenge [the] regulation of others.” 602 U.S. at 385 (emphasis in original). The organizations’ efforts to link themselves to the Cancellation Provision therefore must fail under a basic precept of standing law—equally applicable to organizational plaintiffs, see id. at 393–94—which is that a “plaintiff generally must assert his own legal rights and interests, and cannot rest his claim to relief on the legal rights or interests of third parties.” Warth v. Seldin, 422 U.S. 490, 499 (1975); see also CASA de Maryland, 971 F.3d at 240 (explaining that Havens does not permit organizations to have standing based on “gripes with how the law is enforced against other persons”). We can see how the plaintiff organizations’ theory must be wrong by comparing the organizations at issue here to a prototypical type of organization that helps others: law firms. It is well established that law firms lack a cognizable legal basis to challenge laws harming their clients. See Kowalski v. Tesmer, 543 U.S. 125, 134 n.5 (2004) (rejecting the contention that a medical malpractice attorney could challenge a tort reform statute or that an attorney specializing in Social Security cases could challenge the implementation of a Social Security regulation). Could we ever imagine, for example, a law firm specializing in oil and

gas law having organizational standing to challenge a new drilling regulation that will make its clients’ ability to dig oil wells more onerous, on the theory that the law firm will need to expend more resources advising their clients? Yet at oral argument in this case, the plaintiffs acknowledged that under their theory of Article III standing, a law firm would have standing to challenge laws that make it more difficult for their clients to achieve their legal objectives, because these laws would make it more difficult for the law firm to provide its services. It takes very little to see that by this theory of standing, Havens Realty would hardly be the “unusual case” not to be “extend[ed] . . . beyond its context,” Hippocratic Med., 602 U.S. at 396, but a blow-the-doors-open precedent that would permit any organization to challenge any law that regulates the people it serves. It may seem far-fetched to imagine a law firm having organizational standing to challenge a policy that falls on its clients, but one of our precedents found standing in that very situation. See Immigrant Defs. L. Ctr. v. Noem, 145 F.4th 972, 987–89 (9th Cir. 2025). That decision is clearly wrong under Hippocratic Medicine. But it is the very theory that the plaintiffs here would have us resurrect. Of course, that the plaintiff organizations here lack standing to challenge the Cancellation Provision, which operates on the voters they serve, does not mean the organizations will never have standing to challenge any law. In fact, the organizations’ remaining challenge in this case serves as a useful foil. The organizations also challenge A.R.S. § 16-1016(12), known as the “Felony Provision,” which criminalizes the knowing provision of “a mechanism for voting to another person who is registered in another state.” The Organizations allege that their First Amendment AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 49

activities, including voter registration and voter outreach, could constitute a “mechanism for voting” and lead to prosecution and criminal penalties. Unlike the harms associated with the Cancellation Provision, which fall on third-party voters alone, the organizations themselves fear criminal punishment and allege that their own educational and voting-related activities will be chilled based on the Felony Provision. By the organizations’ well-pleaded allegations, the Felony Provision constitutes “direct” interference with the organizations’ “core business activities,” analogous to the injuries that HOME alleged in Havens Realty. Hippocratic Med., 602 U.S. at 395. That is why the organizations have Article III standing to challenge the Felony Provision, as the majority opinion correctly concludes. But it also shows why the organizations lack Article III standing to challenge the Cancellation Provision. II Today’s majority opinion agrees that the plaintiff organizations lack Article III standing to challenge the Cancellation Provision. In fact, after taking this case en banc presumably to correct a grievous error of law at the three- judge panel level, all eleven judges on this en banc court agree with the three-judge panel majority on that critical point. And like the majority at the three-judge panel stage, the en banc court unanimously agrees that we must overrule our prior organizational standing precedents as incompatible with Hippocratic Medicine. Maj. Op. 18–19. This is welcome news and long overdue. But in our business, it is the reasoning that matters just as much as the result. And when it comes to the reasoning, the majority opinion is inadequate, leaving cracks from which more

expansive notions of organizational standing are destined to seep back through. The majority opinion gives us a limited and incomplete discussion of Havens Realty. It never explains the nature of the injury there or why, under the FHA, it was specific to the organization itself. See PETA, 797 F.3d at 1100 (Millett, J., dubitante). The majority discusses Hippocratic Medicine, as it must. But it never explains the analytical distinction between Hippocratic Medicine and Havens Realty, which made Havens “an unusual case.” Hippocratic Med., 602 U.S. at 396. Nor does the majority opinion discuss how the medical associations in Hippocratic Medicine were “unregulated parties who seek to challenge FDA’s regulation of others.” 602 U.S. at 385. In treating Hippocratic Medicine as a case to be quoted but not analyzed, the majority opinion fails to provide a framework for reasoned decision-making under the terms of Hippocratic Medicine. Of course, to say that the majority opinion quotes Hippocratic Medicine is only half right, because the majority also omits and alters key language from the Hippocratic Medicine decision. Today’s decision from the en banc court does not even acknowledge the Supreme Court’s critical statement in Hippocratic Medicine that Havens Realty “was an unusual case” that the Court “has been careful not to extend . . . beyond its context.” Id. at 396. The majority opinion also does not acknowledge the Supreme Court’s critical description of how HOME had standing in a manner “not dissimilar to a retailer who sues a manufacturer for selling defective goods to the retailer.” Id. at 395. These are not mere castaway comments from the Supreme Court, but reasoning integral to its analysis. Just as questionably, the majority opinion selectively dilutes Hippocratic Medicine’s AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 51

requirement that an organizational plaintiff show that the challenged provision “directly affect[s] and interfere[s] with [its] core business activities.” Id. at 395. In various places, the majority opinion curiously omits the word “directly” from “directly affected and interfered.” See Maj. Op. 8, 15, 20, 21. And it likewise sometimes omits “business” from “core business activities.” Id. at 19, 21. These omissions cannot be excused as efforts in concision when the majority opinion otherwise fails to acknowledge key reasoning in Hippocratic Medicine and when it fails to explain the critical distinction between that case and Havens Realty. Indeed, and in language that lower courts will likely struggle with for years, the majority opinion tells us that “while the frustration of an organization’s mission with a consequent drain on resources sometimes may be a concrete and particularized injury, it is not always one.” Maj. Op. 16. That hedging is inconsistent with the majority overruling our prior organizational standing cases. And it is likewise inconsistent with the Supreme Court’s instruction that an organization “cannot spend its way into standing simply by expending money to gather information and advocate against the defendant’s action.” Hippocratic Med., 602 U.S. at 394. Although purporting to enforce Hippocratic Medicine, the majority sprinkles seeds that are poised to uproot it. Perhaps nowhere is this more apparent than in the majority’s supposed application of Hippocratic Medicine to this case. Maj. Op. 19–22. After overruling our prior cases as inconsistent with Hippocratic Medicine, one might have expected the majority opinion to engage in analysis much like we have done above, through a careful exposition of why the asserted injuries here are not like those in Havens Realty. But the majority tellingly does not do that. Instead,

the majority opinion concludes that the plaintiff organizations lack standing to challenge the Cancellation Provision because their theory of how the Cancellation Provision could harm voters is too speculative. Maj. Op. 20– 21. The majority reasons:

Nothing in the Cancellation Provision arguably requires county recorders to cancel a voter’s new registration. And the record confirms that the state and counties have longstanding procedures in place to ensure that, if elections officials determine an individual is registered in multiple counties, only the old registration is canceled. The state’s Election Procedures Manual, which binds county recorders, codifies these procedures. See Sec’y Katie Hobbs, 2023 Elections Procedures Manual 25–26, 39 (2023); A.R.S. § 16-452(C). The Cancellation Provision aligns with this guidance, and the Organizations point to no evidence that elections officials will abandon it. Indeed, the Organizations conceded at oral argument that there is no evidence in the record of the state canceling a voter’s new registration under these longstanding procedures. Without any showing of actual or imminent injury to the Organizations’ voter-registration activities, they are left only with the loss of time and resources they have spent manufacturing one.

Maj. Op. 20–21. The majority concludes, in other words, that based on its interpretation of the Cancellation Provision, AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 53

the organizations’ theory of how the Cancellation Provision will operate is unsupported. Therefore, the organizations lack standing to raise their challenge. That is decidedly not the mode of analysis that Hippocratic Medicine undertook in finding no organizational standing there. Instead, what the majority in this case holds is that because the organizations’ theory of the Cancellation Provision fails on the merits, the organizational plaintiffs lack standing to advance it. That approach conflates standing and the merits, and it violates black-letter law. As we have previously explained, even if plaintiffs’ “theory may fail on the merits,” that “does not mean [they] lack standing to raise it.” Idaho Conservation League v. Bonneville Power Admin., 83 F.4th 1182, 1189 (9th Cir. 2023) (collecting cases); see also, e.g., Iten v. County of L.A., 81 F.4th 979, 984–86, 989–90 (9th Cir. 2023); Am. Encore v. Fontes, 152 F.4th 1097, 1110–11 (9th Cir. 2025); 13A Charles Alan Wright et al., Federal Practice and Procedure § 3531 (3d ed. Apr. 2026). The Supreme Court has likewise instructed that “[f]or standing purposes, we accept as valid the merits of [plaintiffs’] legal claims.” FEC v. Cruz, 596 U.S. 289, 298 (2022); see also, e.g., Warth, 422 U.S. at 500 (“[S]tanding in no way depends on the merits of the plaintiff’s contention that particular conduct is illegal.”). There is, of course, a limited exception, which is that a plaintiff lacks standing where his claim is “wholly insubstantial and frivolous” or “clearly appears to be immaterial and made solely for the purpose of obtaining jurisdiction.” Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 89 (1998) (citation omitted). But the majority opinion does not invoke this standard, for the obvious reason that it would not be met. Whatever one may think of the

merits of plaintiffs’ theory here, it does not rise to the level of a frivolous claim. Indeed, both the district court judge and our learned dissenting colleague at the three-judge panel level concluded that the plaintiffs showed a likelihood of success on their Cancellation Provision theory. See Ariz. All. for Retired Americans v. Mayes, 117 F.4th 1165, 1192, 1198–99 (9th Cir. 2024) (Nguyen, J., dissenting in part); Ariz. All. for Retired Americans v. Hobbs, 630 F. Supp. 3d 1180, 1189–94 (D. Ariz. 2022). The majority opinion advances an interpretation of the Cancellation Provision that differs from plaintiffs’ interpretation. It then uses this to deny standing. The majority opinion thereby ushers in a new form of standing analysis wholly foreign to existing law, in which an assertedly weak (but not frivolous) claim can be turned around and used as the basis for finding that the plaintiff lacks standing to raise it. The majority has apparently replaced the now-invalidated frustration of purpose / diversion of resources test with a focus on the merits, which violates longstanding principles of standing jurisprudence. That is troubling enough in its own right. But the implication of the majority opinion is that if the plaintiff organizations did have a viable theory as to why the Cancellation Provision could harm voters, then the plaintiffs would have organizational standing. Such a holding, however, would run headlong into Hippocratic Medicine, because under a proper understanding of that case, which we have set forth above, the plaintiff organizations would lack Article III standing no matter how compelling their interpretation of the Cancellation Provision. The majority asserts that it is not, in fact, conducting a merits analysis to deny standing, and that it has merely held that the plaintiff organizations lack an “actual or imminent” AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 55

injury. Maj. Op. 21 n.3. But this is not a case involving a “Rube Goldberg” set of contingencies or an attenuated causal chain. See Clapper v. Amnesty Int’l USA, 568 U.S. 398, 410 (2013). And that is not the lens through which Hippocratic Medicine analyzed the claims of the organizational plaintiffs in that case (the majority is relying on a portion of Hippocratic Medicine concerning individual doctor plaintiffs, not the organizational plaintiffs, see 602 U.S. at 390). All we have here is a statute that was enjoined the day after it took effect, and the state’s representation that it will not enforce the statute in the way the plaintiffs fear. The only reason the majority can assert that the organizations’ claimed injury is “speculative” in this pre- enforcement posture is because it is rejecting the plaintiffs’ interpretation of the Cancellation Provision. See Maj. Op. 20. (“Nothing in the Cancellation Provision arguably requires county recorders to cancel a voter’s new registration.”). That is a merits analysis. But see Ariz. All. for Retired Americans, 117 F.4th at 1192 (Nguyen, J., dissenting in part) (providing a different interpretation of the Cancellation Provision and arguing that “plaintiffs’ statutory interpretation is neither insubstantial nor frivolous”). That the majority opinion disclaims the merits-based nature of its own analysis will, if anything, make compliance with today’s decision even more challenging for lower courts. But regardless of the exact basis the majority is invoking for concluding that the harm caused by the Cancellation Provision is “speculative,” the implication once again is that if the harm to voters were not speculative in the way the majority conceives it, the plaintiff organizations would have standing. But as I have explained, that cannot be the case because the plaintiffs’ theory of organizational

standing fails for more fundamental reasons under Hippocratic Medicine. This is where the majority opinion fundamentally falls short. It overrules our past cases but leaves nothing viable in their stead. It says what the test is not but then applies either a lighter version of that discredited test or one that wrongly turns on the merits of the plaintiffs’ claims (or, apparently, something else entirely). The majority opinion’s assertion that this concurrence “opine[s] on questions that this case has never presented” is therefore manifestly incorrect and betrays a central flaw in the majority’s approach. Maj. Op. 22. What I have addressed here is what this case required us to address. It is the majority that is avoiding those questions through an inadequate analysis that fails to comport with Supreme Court precedent. District courts in our circuit can therefore be forgiven if they are unsure what to do next. But a possible result of all of this, and one the majority opinion perhaps not so subtly invites, is that courts will backslide to where we were before, with organizational standing free for the asking, or at least substantially easier to secure than Supreme Court precedent would direct. This is not where the Supreme Court left us after Hippocratic Medicine. Havens Realty was “an unusual case.” Hippocratic Med., 602 U.S. at 396. It was incumbent upon us to flesh out why that is, which is what will allow courts in our circuit to decide questions of organizational standing in a legally supported and principled manner, consistent with the Supreme Court’s directives. Whether today’s decision is merely a missed opportunity or a more directed deviation from Supreme Court precedent will be for time to tell. But every indication is that it is the latter. AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 57

VANDYKE, Circuit Judge, concurring in parts II.B and III and concurring in the judgment: I join Judge Bress’s excellent concurring opinion in full. As I’ve emphasized previously, see Sabra v. Maricopa Cnty. Cmty. Coll. Dist., 44 F.4th 867, 896 (9th Cir. 2022) (VanDyke, J., concurring), our organizational standing doctrine has long been in need of correction, and I share my colleagues’ concerns that the majority opinion, while purporting to apply Food & Drug Administration v. Alliance for Hippocratic Medicine, 602 U.S. 367 (2024), is actually taking the first step toward undercutting it. But in addition to resetting our organizational precedent and conforming it to the Supreme Court’s requirements as Judge Bress emphasizes, I would have taken this opportunity to slightly modify the standard set forth in this court’s decision in Miller v. Gammie, which requires three-judge panels of this court to follow Ninth Circuit precedent unless it is “clearly irreconcilable” with subsequent Supreme Court precedent. 335 F.3d 889, 893 (9th Cir. 2003) (en banc). Or, to state Gammie’s test another way, our precedent trumps Supreme Court precedent even when our precedent is (merely) “irreconcilable” with—or just inconsistent with, or at odds with, or based on reasoning that cannot be squared with— what the Supreme Court has subsequently told us. That is wrong. But that is not the only oddity created by our circuit’s choice to employ a robust clear statement rule before it deems intervening Supreme Court precedent worth following. One of the main arguments proffered for taking this case en banc was that the original three-judge panel majority had failed to properly comply with the strictures of Miller v. Gammie and therefore erred by complying with Supreme Court precedent. Presumably, that argument

resonated with at least some of my colleagues, since a majority of our court voted to grant en banc review. Yet while a supposed violation of the Miller v. Gammie standard is often employed as a reason to take a case en banc, once a case is taken en banc, that argument immediately becomes irrelevant. An en banc panel’s duty when rehearing a case is to get the law right, which means following the Supreme Court’s instructions regardless of whether those instructions are “clearly irreconcilable” with our precedent. Thus, even if a three-judge panel arguably erred in its Miller v. Gammie analysis, an en banc panel could reach the same decision by applying a standard that is less deferential to our own precedent (and more faithful to the Supreme Court’s). Indeed, that’s close to what happened here: We took the case en banc only to reach the same outcome. The en banc panel’s majority opinion levels no criticism against the three- judge panel’s opinion. That suggests that although the panel’s conclusion was correct, it was supposedly wrong for them to say so, yet it is correct for us to say so. That makes little sense. Supreme Court precedent should be supreme over three-judge panels and en banc panels alike. See U.S. Const. art. III, § 1 (“The judicial power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.” (emphasis added)). Three-judge panels should not need permission from the en banc court to follow intervening Supreme Court precedent. Accordingly, in addition to fixing our organizational standing precedent, I would have also taken this opportunity to modify our Miller v. Gammie rule to clarify that Supreme Court precedent controls whenever it is inconsistent with our own. Doing so would ensure that we need not consume the en banc court’s future resources determining whether an irreconcilability AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 59

between panel precedent and Supreme Court precedent is “clear” enough that the Supreme Court precedent is worth following. * * * Our law of the circuit doctrine forbids three-judge panels from overruling prior panel decisions, which can only be modified by a subsequent en banc panel or by the Supreme Court. See Hart v. Massanari, 266 F.3d 1155, 1171 (9th Cir. 2001). It is common, however, for the Supreme Court to issue decisions that do not directly overrule our panel precedent but do contradict or undermine the general reasoning or conclusions therein. This frequently presents panels with a dilemma: Do they follow the intervening Supreme Court precedent or our panel precedent? Our court answered this question in Miller v. Gammie by requiring that Ninth Circuit panels always prefer the precedent set by a three-judge panel to that of the Supreme Court unless the two decisions are “clearly irreconcilable.” 335 F.3d. at 893. This standard is an outlier in the federal courts of appeals and a major factor in the unpredictability and occasional inaccuracy of Ninth Circuit decisions. I would have used this en banc case as an opportunity to reconsider this rule so that any time our precedent is “merely” irreconcilable or inconsistent with Supreme Court precedent—including the mode or type of analysis used in reaching a conclusion—a panel must follow the Supreme Court precedent, even where it may not meet Miller v. Gammie’s “clearly irreconcilable” standard. Many other circuits, when faced with the issue of whether to follow the Supreme Court or their own earlier panel decisions, have adopted standards that more appropriately favor the precedent set by the Supreme Court.

For example, both the Second Circuit and the Eighth Circuit allow panels to overrule an earlier panel decision when a subsequent Supreme Court decision simply “casts doubt” on that prior decision. See, e.g., Union of Needletrades, Indus. & Textile Emps. v. U.S. INS, 336 F.3d 200, 210 (2d Cir. 2003); United States v. Villareal-Amarillas, 562 F.3d 892, 898 n.4 (8th Cir. 2009). The Sixth Circuit seems to employ a standard even more deferential to the Supreme Court. See United States v. Fields, 53 F.4th 1027, 1046–47 (6th Cir. 2022) (“One panel is usually bound by the ruling of a previous one, but there is an exception to that rule when the Supreme Court issues an intervening decision …. That exception applies even when the intervening Supreme Court decision is not ‘precisely on point’ but provides ‘directly applicable’ legal reasoning, or when it provides on-point dictum.” (citations omitted)). Other circuits similarly employ various iterations of standards less demanding than the Ninth Circuit’s. See, e.g., Riccio v. Sentry Credit, Inc., 954 F.3d 582, 592 (3d Cir. 2020) (“[C]ircuit precedent, though not directly overruled or superseded, nonetheless might crumble if case law postdating the original decision, although not directly controlling, nevertheless offers a sound reason for believing that the former panel, in light of fresh developments, would change its collective mind.” (internal quotation marks and citation omitted)). Indeed, at least in the context of intervening Supreme Court decisions, the D.C. Circuit has argued that stare decisis affirmatively requires a more deferential approach to the Supreme Court. See Save Jobs USA v. U.S. Dep’t of Homeland Sec., 111 F.4th 76, 80 (D.C. Cir. 2024) (“[V]ertical stare decisis ... requires a circuit panel to depart from a circuit precedent … if the circuit precedent’s reasoning was later eviscerated by the reasoning AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 61

in [an intervening Supreme Court decision].” (internal quotation marks and citation omitted)). So while different circuits vary modestly in their answers to the question we decided in Miller v. Gammie, generally speaking, the other circuits privilege their own precedent—vis-à-vis the Supreme Court’s—less than we privilege ours.1 The Ninth Circuit’s self-aggrandizing standard has also led to unnecessary circuit splits. For example, in the early 2000s the Third, Seventh, Eighth, and Ninth Circuits all adopted a standard that seemed to conflict with a subsequent Supreme Court decision, United States v. Booker, 543 U.S. 220 (2005). The Third, Seventh, and Eighth circuits all concluded that their circuit precedent was abrogated by Booker. See Villareal-Amarillas, 562 F.3d at 898 n.4; United States v. Fisher, 502 F.3d 293, 295 (3d Cir. 2007); United States v. Reuter, 463 F.3d 792, 793 (7th Cir. 2006). But the Ninth Circuit declined to do so. See United States v. Staten, 466 F.3d 708, 718–20 (9th Cir. 2006). In explaining the reason for this apparent split, the Eighth Circuit highlighted that, “[i]n the Ninth Circuit, a three-judge panel may

1 The Fourth Circuit’s standard appears most like the Ninth’s. See Short v. Hartman, 87 F.4th 593, 605 (4th Cir. 2023) (“A Supreme Court decision overrules or abrogates our prior precedent only if our precedent is impossible to reconcile with a subsequent Supreme Court decision.” (internal quotation marks and citation omitted)). Only one circuit, the Eleventh, has a standard that appears stricter than the Ninth Circuit’s. See United States v. Dubois, 94 F.4th 1284, 1293 (11th Cir. 2024) (“An intervening Supreme Court decision abrogates our precedent only if the intervening decision is both clearly on point and clearly contrary to our earlier decision.” (internal quotation marks and citation omitted)). Unlike the other circuits, the Federal Circuit does not appear to have a well-established standard for determining if the Supreme Court has overruled its circuit precedent. See Teva Pharms. USA, Inc. v. Novartis Pharms. Corp., 482 F.3d 1330, 1338 (Fed. Cir. 2007).

reexamine a prior panel decision only if a supervening Supreme Court decision is ‘clearly irreconcilable.’ By contrast, [the Eighth Circuit] may reconsider a prior panel’s decision if a supervening Supreme Court decision undermines or casts doubt on the earlier panel decision.” Villareal-Amarillas, 562 F.3d at 898 n.4 (internal quotation marks and citation omitted). Thus, the Ninth Circuit’s too- demanding “clearly irreconcilable” standard caused it to split with every other circuit faced with the same issue. Presumably, our peculiar standard is meant to foster consistency and predictability in our court’s decisions. See Gammie, 335 F.3d at 900 (“A goal of our circuit’s decisions, including panel and en banc decisions, must be to preserve the consistency of circuit law.”). But it is not clear at all that our “clearly irreconcilable” standard has made our decisions any more consistent or predictable. To the contrary, the Ninth Circuit has been regarded as one of the least predictable and most inconsistent federal appeals courts in the nation. See, e.g., Richard A. Posner, Is the Ninth Circuit Too Large? A Statistical Study of Judicial Quality, 29 J. LEGAL STUD. 711, 717 n.12 (2000) (“Among experienced litigators, 30 percent reported trouble predicting the outcome of an appeal in the Ninth Circuit, and this was a higher percentage than for any other circuit.” (citing Fed. Jud. Ctr., Survey of Judges and Appellate Counsel Conducted for the Commission on Structural Alternatives for the Federal Courts of Appeals 79 (1998))). Indeed, this very case helpfully demonstrates the unpredictability of decisions in our court. We granted en banc review in response to the argument that the three-judge panel erred by concluding that earlier Ninth Circuit precedent regarding organizational standing, see, e.g., Fair Hous. Council of San Fernando Valley v. Roommate.com, AZ ALLIANCE FOR RETIRED AMERICANS V. MAYES 63

LLC, 666 F.3d 1216, 1219 (9th Cir. 2012), was “clearly irreconcilable” with later Supreme Court precedent in Hippocratic Medicine, 602 U.S. 367. But ultimately, after granting en banc review, we have reached the same result as the panel because, of course, our duty as an en banc court— unconstrained by Miller v. Gammie’s straitjacket—is simply to faithfully follow the Supreme Court’s most recent guidance. This case thus helps illustrate that our Miller v. Gammie standard does not necessarily provide more predictability or stability. Instead, it encourages tangential court-wide disputes about whether the Supreme Court has made “clear” enough that our precedent is abrogated. Put differently, because of Miller v. Gammie, we have shifted the debate from simply whether our precedent is consistent with the Supreme Court’s guidance to whether it is so inconsistent that we should fix it. This brings me to yet another problem with the “clearly irreconcilable” standard—indeed, the elephant in the room. Our standard necessarily allows (indeed, requires) our court to apply precedent that is irreconcilable with Supreme Court decisions as long as it is not “clearly” so. This is an implicit admission that some of our precedents are irreconcilable with higher authority, yet we choose to follow those precedents anyway. If given a choice between Supreme Court precedent and our panel precedent, we should always choose the former. I would thus have taken this opportunity to slightly modify Miller v. Gammie’s “clearly irreconcilable” standard to allow a three-judge panel to overturn a prior panel decision when an intervening Supreme Court decision is simply inconsistent with our court’s precedent. Getting the law right should be more important than predictably getting the law wrong.

In the two decades since its implementation in Miller v. Gammie, the “clearly irreconcilable” standard has gained a reputation for being unpredictably and inconsistently applied and has placed the Ninth Circuit on the minority side of circuit splits. See, e.g., Posner, supra, at 717 n.12; Villareal-Amarillas, 562 F.3d at 898 n.4. I would have used this case to join our sister circuits in choosing Supreme Court precedent over our own when the two are “merely” inconsistent, and I hope that in a future en banc case our court will consider that question.