Americans for Prosperity Foundation v. Bonta, 594 U.S. ___ (2021)

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

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 19–251 and 19–255

_________________

AMERICANS FOR PROSPERITY FOUNDATION, PETITIONER

19–251v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

THOMAS MORE LAW CENTER, PETITIONER

19–255v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

on writs of certiorari to the united states court of appeals for the ninth circuit

[July 1, 2021]

Chief Justice Roberts delivered the opinion of the Court, except as to Part II–B–1.

To solicit contributions in California, charitable organizations must disclose to the state Attorney General’s Office the identities of their major donors. The State contends that having this information on hand makes it easier to police misconduct by charities. We must decide whether California’s disclosure requirement violates the First Amendment right to free association.

I

The California Attorney General’s Office is responsible for statewide law enforcement, including the supervision and regulation of charitable fundraising. Under state law, the Attorney General is authorized to “establish and maintain a register” of charitable organizations and to obtain “whatever information, copies of instruments, reports, and records are needed for the establishment and maintenance of the register.” Cal. Govt. Code Ann. §12584 (West 2018). In order to operate and raise funds in California, charities generally must register with the Attorney General and renew their registrations annually. §§12585(a), 12586(a). Over 100,000 charities are currently registered in the State, and roughly 60,000 renew their registrations each year.

California law empowers the Attorney General to make rules and regulations regarding the registration and renewal process. §§12585(b), 12586(b). Pursuant to this regulatory authority, the Attorney General requires charities renewing their registrations to file copies of their Internal Revenue Service Form 990, along with any attachments and schedules. See Cal. Code Regs., tit. 11, §301 (2020). Form 990 contains information regarding tax-exempt organizations’ mission, leadership, and finances. Schedule B to Form 990—the document that gives rise to the present dispute—requires organizations to disclose the names and addresses of donors who have contributed more than $5,000 in a particular tax year (or, in some cases, who have given more than 2 percent of an organization’s total contributions). See 26 CFR §§1.6033–2(a)(2)(ii)(f ), (iii) (2020).

The petitioners are tax-exempt charities that solicit contributions in California and are subject to the Attorney General’s registration and renewal requirements. Americans for Prosperity Foundation is a public charity that is “devoted to education and training about the principles of a free and open society, including free markets, civil liberties, immigration reform, and constitutionally limited government.” Brief for Petitioner Foundation 10. Thomas More Law Center is a public interest law firm whose “mission is to protect religious freedom, free speech, family values, and the sanctity of human life.” Brief for Petitioner Law Center 4. Since 2001, each petitioner has renewed its registration and has filed a copy of its Form 990 with the Attorney General, as required by Cal. Code Regs., tit. 11, §301. Out of concern for their donors’ anonymity, however, the petitioners have declined to file their Schedule Bs (or have filed only redacted versions) with the State.

For many years, the petitioners’ reluctance to turn over donor information presented no problem because the Attorney General was not particularly zealous about collecting Schedule Bs. That changed in 2010, when the California Department of Justice “ramped up its efforts to enforce charities’ Schedule B obligations, sending thousands of deficiency letters to charities that had not complied with the Schedule B requirement.” Americans for Prosperity Foundation v. Becerra, 903 F.3d 1000, 1006 (CA9 2018). The Law Center and the Foundation received deficiency letters in 2012 and 2013, respectively. When they continued to resist disclosing their contributors’ identities, the Attorney General threatened to suspend their registrations and fine their directors and officers.

The petitioners each responded by filing suit in the Central District of California. In their complaints, they alleged that the Attorney General had violated their First Amendment rights and the rights of their donors. The petitioners alleged that disclosure of their Schedule Bs would make their donors less likely to contribute and would subject them to the risk of reprisals. Both organizations challenged the disclosure requirement on its face and as applied to them.

In each case, the District Court granted preliminary injunctive relief prohibiting the Attorney General from collecting their Schedule B information. Americans for Prosperity Foundation v. Harris, 2015 WL 769778 (CD Cal., Feb. 23, 2015); App. to Pet. for Cert. in No. 19–255, pp. 90a–96a. The Ninth Circuit vacated and remanded. Americans for Prosperity Foundation v. Harris, 809 F.3d 536 (2015) (per curiam). The court held that it was bound by Circuit precedent to reject the petitioners’ facial challenge. Id., at 538 (citing Center for Competitive Politics v. Harris, 784 F.3d 1307, 1317 (2015)). And reviewing the petitioners’ as-applied claims under an “exacting scrutiny” standard, the panel narrowed the injunction, allowing the Attorney General to collect the petitioners’ Schedule Bs so long as he did not publicly disclose them. 809 F. 3d, at 538, 543.

On remand, the District Court held bench trials in both cases, after which it entered judgment for the petitioners and permanently enjoined the Attorney General from collecting their Schedule Bs. Americans for Prosperity Foundation v. Harris, 182 F. Supp. 3d 1049 (CD Cal. 2016); Thomas More Law Center v. Harris, 2016 WL 6781090 (CD Cal., Nov. 16, 2016). Applying exacting scrutiny, the District Court held that disclosure of Schedule Bs was not narrowly tailored to the State’s interest in investigating charitable misconduct. The court credited testimony from California officials that Schedule Bs were rarely used to audit or investigate charities. And it found that even where Schedule B information was used, that information could be obtained from other sources.

The court also determined that the disclosure regime burdened the associational rights of donors. In both cases, the court found that the petitioners had suffered from threats and harassment in the past, and that donors were likely to face similar retaliation in the future if their affiliations became publicly known. For example, the CEO of the Foundation testified that a technology contractor working at the Foundation’s headquarters had posted online that he was “inside the belly of the beast” and “could easily walk into [the CEO’s] office and slit his throat.” 182 F. Supp. 3d, at 1056. And the Law Center introduced evidence that it had received “threats, harassing calls, intimidating and obscene emails, and even pornographic letters.” 2016 WL 6781090, *4.

The District Court also found that California was unable to ensure the confidentiality of donors’ information. During the course of litigation, the Foundation identified nearly 2,000 confidential Schedule Bs that had been inadvertently posted to the Attorney General’s website, including dozens that were found the day before trial. One of the Foundation’s expert witnesses also discovered that he was able to access hundreds of thousands of confidential documents on the website simply by changing a digit in the URL. The court found after trial that “the amount of careless mistakes made by the Attorney General’s Registry is shocking.” 182 F. Supp. 3d, at 1057. And although California subsequently codified a policy prohibiting disclosure, Cal. Code Regs., tit. 11, §310(b)—an effort the District Court described as “commendable”—the court determined that “[d]onors and potential donors would be reasonably justified in a fear of disclosure given such a context” of past breaches. 2016 WL 6781090, *5.

The Ninth Circuit again vacated the District Court’s injunctions, and this time reversed the judgments and remanded for entry of judgment in favor of the Attorney General. 903 F.3d 1000. The court held that the District Court had erred by imposing a narrow tailoring requirement. Id., at 1008–1009. And it reasoned that the disclosure regime satisfied exacting scrutiny because the up-front collection of charities’ Schedule Bs promoted investigative efficiency and effectiveness. Id., at 1009–1012. The panel also found that the disclosure of Schedule Bs would not meaningfully burden donors’ associational rights, in part because the Attorney General had taken remedial security measures to fix the confidentiality breaches identified at trial. Id., at 1013–1019.

The Ninth Circuit denied rehearing en banc. Americans for Prosperity Foundation v. Becerra, 919 F.3d 1177 (2019). Judge Ikuta dissented, joined by four other judges. In her view, the panel had impermissibly overridden the District Court’s factual findings and evaluated the disclosure requirement under too lenient a degree of scrutiny. Id., at 1184–1187.

We granted certiorari. 592 U. S. ___ (2021).

II

A

The First Amendment prohibits government from “abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.” This Court has “long understood as implicit in the right to engage in activities protected by the First Amendment a corresponding right to associate with others.” Roberts v. United States Jaycees, 468 U.S. 609, 622 (1984). Protected association furthers “a wide variety of political, social, economic, educational, religious, and cultural ends,” and “is especially important in preserving political and cultural diversity and in shielding dissident expression from suppression by the majority.” Ibid. Government infringement of this freedom “can take a number of forms.” Ibid. We have held, for example, that the freedom of association may be violated where a group is required to take in members it does not want, see id., at 623, where individuals are punished for their political affiliation, see Elrod v. Burns, 427 U.S. 347, 355 (1976) (plurality opinion), or where members of an organization are denied benefits based on the organization’s message, see Healy v. James, 408 U.S. 169, 181–182 (1972).

We have also noted that “[i]t is hardly a novel perception that compelled disclosure of affiliation with groups engaged in advocacy may constitute as effective a restraint on freedom of association as [other] forms of governmental action.” NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 462 (1958). NAACP v. Alabama involved this chilling effect in its starkest form. The NAACP opened an Alabama office that supported racial integration in higher education and public transportation. Id., at 452. In response, NAACP members were threatened with economic reprisals and violence. Id., at 462. As part of an effort to oust the organization from the State, the Alabama Attorney General sought the group’s membership lists. Id., at 452–453. We held that the First Amendment prohibited such compelled disclosure. Id., at 466. We explained that “[e]ffective advocacy of both public and private points of view, particularly controversial ones, is undeniably enhanced by group association,” id., at 460, and we noted “the vital relationship between freedom to associate and privacy in one’s associations,” id., at 462. Because NAACP members faced a risk of reprisals if their affiliation with the organization became known—and because Alabama had demonstrated no offsetting interest “sufficient to justify the deterrent effect” of disclosure, id., at 463—we concluded that the State’s demand violated the First Amendment.

B

1

NAACP v. Alabama did not phrase in precise terms the standard of review that applies to First Amendment challenges to compelled disclosure. We have since settled on a standard referred to as “exacting scrutiny.” Buckley v. Valeo, 424 U.S. 1, 64 (1976) (per curiam). Under that standard, there must be “a substantial relation between the disclosure requirement and a sufficiently important governmental interest.” Doe v. Reed, 561 U.S. 186, 196 (2010) (internal quotation marks omitted). “To withstand this scrutiny, the strength of the governmental interest must reflect the seriousness of the actual burden on First Amendment rights.” Ibid. (internal quotation marks omitted). Such scrutiny, we have held, is appropriate given the “deterrent effect on the exercise of First Amendment rights” that arises as an “inevitable result of the government’s conduct in requiring disclosure.” Buckley, 424 U. S., at 65.

The Law Center (but not the Foundation) argues that we should apply strict scrutiny, not exacting scrutiny. Under strict scrutiny, the government must adopt “the least restrictive means of achieving a compelling state interest,” McCullen v. Coakley, 573 U.S. 464, 478 (2014), rather than a means substantially related to a sufficiently important interest. The Law Center contends that only strict scrutiny adequately protects the associational rights of charities. And although the Law Center acknowledges that we have applied exacting scrutiny in prior disclosure cases, it argues that those cases arose in the electoral context, where the government’s important interests justify less searching review.

It is true that we first enunciated the exacting scrutiny standard in a campaign finance case. See Buckley, 424 U. S., at 64–68. And we have since invoked it in other election-related settings. See, e.g., Citizens United v. Federal Election Comm’n, 558 U.S. 310, 366–367 (2010); Davis v. Federal Election Comm’n, 554 U.S. 724, 744 (2008). But exacting scrutiny is not unique to electoral disclosure regimes. To the contrary, Buckley derived the test from NAACP v. Alabama itself, as well as other nonelection cases. See 424 U. S., at 64 (citing Gibson v. Florida Legislative Investigation Comm., 372 U.S. 539 (1963); NAACP v. Button, 371 U.S. 415 (1963); Shelton v. Tucker, 364 U.S. 479 (1960); Bates v. Little Rock, 361 U.S. 516 (1960)). As we explained in NAACP v. Alabama, “it is immaterial” to the level of scrutiny “whether the beliefs sought to be advanced by association pertain to political, economic, religious or cultural matters.” 357 U. S., at 460–461. Regardless of the type of association, compelled disclosure requirements are reviewed under exacting scrutiny.

2

The Law Center (now joined by the Foundation) argues in the alternative that even if exacting scrutiny applies, such review incorporates a least restrictive means test similar to the one imposed by strict scrutiny. The United States and the Attorney General respond that exacting scrutiny demands no additional tailoring beyond the “substantial relation” requirement noted above. We think that the answer lies between those two positions. While exacting scrutiny does not require that disclosure regimes be the least restrictive means of achieving their ends, it does require that they be narrowly tailored to the government’s asserted interest.

The need for narrow tailoring was set forth early in our compelled disclosure cases. In Shelton v. Tucker, we considered an Arkansas statute that required teachers to disclose every organization to which they belonged or contributed. 364 U. S., at 480. We acknowledged the importance of “the right of a State to investigate the competence and fitness of those whom it hires to teach in its schools.” Id., at 485. On that basis, we distinguished prior decisions in which we had found “no substantially relevant correlation between the governmental interest asserted and the State’s effort to compel disclosure.” Ibid. But we nevertheless held that the Arkansas statute was invalid because even a “legitimate and substantial” governmental interest “cannot be pursued by means that broadly stifle fundamental personal liberties when the end can be more narrowly achieved.” Id., at 488; see also Louisiana ex rel. Gremillion v. NAACP, 366 U.S. 293, 296 (1961) (quoting same).

Shelton stands for the proposition that a substantial relation to an important interest is not enough to save a disclosure regime that is insufficiently tailored. This requirement makes sense. Narrow tailoring is crucial where First Amendment activity is chilled—even if indirectly—“[b]ecause First Amendment freedoms need breathing space to survive.” Button, 371 U. S., at 433.

Our more recent decisions confirm the need for tailoring. In McCutcheon v. Federal Election Commission, 572 U.S. 185 (2014), for example, a plurality of the Court explained:

“In the First Amendment context, fit matters. Even when the Court is not applying strict scrutiny, we still require a fit that is not necessarily perfect, but reasonable; that represents not necessarily the single best disposition but one whose scope is in proportion to the interest served, that employs not necessarily the least restrictive means but a means narrowly tailored to achieve the desired objective.” Id., at 218 (internal quotation marks and alterations omitted).

McCutcheon is instructive here. A substantial relation is necessary but not sufficient to ensure that the government adequately considers the potential for First Amendment harms before requiring that organizations reveal sensitive information about their members and supporters. Where exacting scrutiny applies, the challenged requirement must be narrowly tailored to the interest it promotes, even if it is not the least restrictive means of achieving that end.

The dissent reads our cases differently. It focuses on the words “broadly stifle” in the quotation from Shelton above, and it interprets those words to mean that narrow tailoring is required only for disclosure regimes that “impose a severe burden on associational rights.” Post, at 13 (opinion of Sotomayor, J.). Because, in the dissent’s view, the petitioners have not shown such a burden here, narrow tailoring is not required.

We respectfully disagree. The “government may regulate in the [ First Amendment] area only with narrow specificity,” Button, 371 U. S., at 433, and compelled disclosure regimes are no exception. When it comes to “a person’s beliefs and associations,” “[b]road and sweeping state inquiries into these protected areas . . . discourage citizens from exercising rights protected by the Constitution.” Baird v. State Bar of Ariz., 401 U.S. 1, 6 (1971) (plurality opinion). Contrary to the dissent, we understand this Court’s discussion of rules that are “broad” and “broadly stifle” First Amendment freedoms to refer to the scope of challenged restrictions—their breadth—rather than the severity of any demonstrated burden. That much seems clear to us from Shelton’s statement (in the sentence following the one quoted by the dissent) that “[t]he breadth of legislative abridgment must be viewed in the light of less drastic means for achieving the same basic purpose.” 364 U. S., at 488; see id., at 488, n. 9 (citing sources that support this reading). It also seems clear from the immediately preceding paragraph, which stressed that “[t]he scope of the inquiry required by [the law] is completely unlimited. . . . It requires [the teacher] to list, without number, every conceivable kind of associational tie—social, professional, political, avocational, or religious. Many such relationships could have no possible bearing upon the teacher’s occupational competence or fitness.” Id., at 488. In other words, the law was not narrowly tailored to the State’s objective.

Nor does our decision in Reed suggest that narrow tailoring is required only for laws that impose severe burdens. The dissent casts Reed as a case involving only “ ‘modest burdens,’ ” and therefore “a correspondingly modest level of tailoring.” Post, at 12 (quoting 561 U. S., at 201). But it was only after we concluded that various narrower alternatives proposed by the plaintiffs were inadequate, see 561 U. S., at 198–199, that we held that the strength of the government’s interest in disclosure reflected the burden imposed, see id., at 201. The point is that a reasonable assessment of the burdens imposed by disclosure should begin with an understanding of the extent to which the burdens are unnecessary, and that requires narrow tailoring.

III

The Foundation and the Law Center both argued below that the obligation to disclose Schedule Bs to the Attorney General was unconstitutional on its face and as applied to them. See 903 F. 3d, at 1007. The petitioners renew their facial challenge in this Court, and they argue in the alternative that they are entitled to as-applied relief. For the reasons below, we conclude that California’s blanket demand for Schedule Bs is facially unconstitutional.

A

As explained, exacting scrutiny requires that there be “a substantial relation between the disclosure requirement and a sufficiently important governmental interest,” Reed, 561 U. S., at 196 (internal quotation marks omitted), and that the disclosure requirement be narrowly tailored to the interest it promotes, see Shelton, 364 U. S., at 488. The Ninth Circuit found that there was a substantial relation between the Attorney General’s demand for Schedule Bs and a sufficiently strong governmental interest. 903 F. 3d, at 1008–1020. Of particular relevance, the court found that California had such an interest in preventing charitable fraud and self-dealing, and that “the up-front collection of Schedule B information improves the efficiency and efficacy of the Attorney General’s important regulatory efforts.” Id., at 1011–1012. The court did not apply a narrow tailoring requirement, however, because it did not read our cases to mandate any such inquiry. Id., at 1008–1009. That was error. And properly applied, the narrow tailoring requirement is not satisfied by the disclosure regime.

We do not doubt that California has an important interest in preventing wrongdoing by charitable organizations. It goes without saying that there is a “substantial governmental interest[ ] in protecting the public from fraud.” Schaumburg v. Citizens for Better Environment, 444 U.S. 620, 636 (1980) (internal quotation marks omitted). The Attorney General receives complaints each month that identify a range of misconduct, from “misuse, misappropriation, and diversion of charitable assets,” to “false and misleading charitable solicitations,” to other “improper activities by charities soliciting charitable donations.” App. in No. 19–255, p. 270 (alteration omitted). Such offenses cause serious social harms. And the Attorney General is the primary law enforcement officer charged with combating them under California law. See Cal. Govt. Code Ann. §12598.

There is a dramatic mismatch, however, between the interest that the Attorney General seeks to promote and the disclosure regime that he has implemented in service of that end. Recall that 60,000 charities renew their registrations each year, and nearly all are required to file a Schedule B. Each Schedule B, in turn, contains information about a charity’s top donors—a small handful of individuals in some cases, but hundreds in others. See App. in No. 19–251, p. 319. This information includes donors’ names and the total contributions they have made to the charity, as well as their addresses.

Given the amount and sensitivity of this information harvested by the State, one would expect Schedule B collection to form an integral part of California’s fraud detection efforts. It does not. To the contrary, the record amply supports the District Court’s finding that there was not “a single, concrete instance in which pre-investigation collection of a Schedule B did anything to advance the Attorney General’s investigative, regulatory or enforcement efforts.” 182 F. Supp. 3d, at 1055.

The dissent devotes much of its analysis to relitigating factual disputes that the District Court resolved against the Attorney General, see post, at 16–20, notwithstanding the applicable clear error standard of review, see Fed. Rule Civ. Proc. 52(a). For example, the dissent echoes the State’s argument that, in some cases, it relies on up-front Schedule B collection to prevent and police fraud. See post, at 18–19. But the record before the District Court tells a different story. See, e.g., App. in No. 19–251, at 397, 403, 417. And even if the State relied on up-front collection in some cases, its showing falls far short of satisfying the means-end fit that exacting scrutiny requires. California is not free to enforce any disclosure regime that furthers its interests. It must instead demonstrate its need for universal production in light of any less intrusive alternatives. Cf. Shelton, 364 U. S., at 488.

The Attorney General and the dissent contend that alternative means of obtaining Schedule B information—such as a subpoena or audit letter—are inefficient and ineffective compared to up-front collection. See post, at 19. It became clear at trial, however, that the Office had not even considered alternatives to the current disclosure requirement. See App. in No. 19–251, at 421 (“I see no reason to change what we’ve been doing.”). The Attorney General and the dissent also argue that a targeted request for Schedule B information could tip a charity off, causing it to “hide or tamper with evidence.” Brief for Respondent 43; see post, at 19–20. But again, the States’ witnesses failed to substantiate that concern. See, e.g., App. in No. 19–251, at 405–406; see also Board of Trustees of State Univ. of N. Y. v. Fox, 492 U.S. 469, 480 (1989) (“the State . . . must affirmatively establish the reasonable fit we require”). Nor do the actions of investigators suggest a risk of tipping off charities under suspicion, as the standard practice is to send audit letters asking for a wide range of information early in the investigative process. See App. in No. 19–251, at 406, 411, 418. Furthermore, even if tipoff were a concern in some cases, the State’s indiscriminate collection of Schedule Bs in all cases would not be justified.

The upshot is that California casts a dragnet for sensitive donor information from tens of thousands of charities each year, even though that information will become relevant in only a small number of cases involving filed complaints. See id., at 307–308. California does not rely on Schedule Bs to initiate investigations, and in all events, there are multiple alternative mechanisms through which the Attorney General can obtain Schedule B information after initiating an investigation. The need for up-front collection is particularly dubious given that California—one of only three States to impose such a requirement, see id., at 420—did not rigorously enforce the disclosure obligation until 2010. Certainly, this is not a regime “whose scope is in proportion to the interest served.” McCutcheon, 572 U. S., at 218 (internal quotation marks omitted).

In reality, then, California’s interest is less in investigating fraud and more in ease of administration. This interest, however, cannot justify the disclosure requirement. The Attorney General may well prefer to have every charity’s information close at hand, just in case. But “the prime objective of the First Amendment is not efficiency.” McCullen, 573 U. S., at 495. Mere administrative convenience does not remotely “reflect the seriousness of the actual burden” that the demand for Schedule Bs imposes on donors’ association rights. Reed, 561 U. S., at 196 (internal quotation marks omitted).

B

The foregoing discussion also makes clear why a facial challenge is appropriate in these cases. Normally, a plaintiff bringing a facial challenge must “establish that no set of circumstances exists under which the [law] would be valid,” United States v. Salerno, 481 U.S. 739, 745 (1987), or show that the law lacks “a plainly legitimate sweep,” Washington State Grange v. Washington State Republican Party, 552 U.S. 442, 449 (2008) (internal quotation marks omitted). In the First Amendment context, however, we have recognized “a second type of facial challenge, whereby a law may be invalidated as overbroad if a substantial number of its applications are unconstitutional, judged in relation to the statute’s plainly legitimate sweep.” United States v. Stevens, 559 U.S. 460, 473 (2010) (internal quotation marks omitted). We have no trouble concluding here that the Attorney General’s disclosure requirement is overbroad. The lack of tailoring to the State’s investigative goals is categorical—present in every case—as is the weakness of the State’s interest in administrative convenience. Every demand that might chill association therefore fails exacting scrutiny.

The Attorney General tries to downplay the burden on donors, arguing that “there is no basis on which to conclude that California’s requirement results in any broad-based chill.” Brief for Respondent 36. He emphasizes that “California’s Schedule B requirement is confidential,” and he suggests that certain donors—like those who give to noncontroversial charities—are unlikely to be deterred from contributing. Id., at 36–37. He also contends that disclosure to his office imposes no added burdens on donors because tax-exempt charities already provide their Schedule Bs to the IRS. Id., at 37.

We are unpersuaded. Our cases have said that disclosure requirements can chill association “[e]ven if there [is] no disclosure to the general public.” Shelton, 364 U. S., at 486. In Shelton, for example, we noted the “constant and heavy” pressure teachers would experience simply by disclosing their associational ties to their schools. Ibid. Exacting scrutiny is triggered by “state action which may have the effect of curtailing the freedom to associate,” and by the “possible deterrent effect” of disclosure. NAACP v. Alabama, 357 U. S., at 460–461 (emphasis added); see Talley v. California, 362 U.S. 60, 65 (1960) (“identification and fear of reprisal might deter perfectly peaceful discussions of public matters of importance” (emphasis added)). While assurances of confidentiality may reduce the burden of disclosure to the State, they do not eliminate it.[1]*

It is irrelevant, moreover, that some donors might not mind—or might even prefer—the disclosure of their identities to the State. The disclosure requirement “creates an unnecessary risk of chilling” in violation of the First Amendment, Secretary of State of Md. v. Joseph H. Munson Co., 467 U.S. 947, 968 (1984), indiscriminately sweeping up the information of every major donor with reason to remain anonymous. The petitioners here, for example, introduced evidence that they and their supporters have been subjected to bomb threats, protests, stalking, and physical violence. App. in No. 19–251, at 256, 291–292. Such risks are heightened in the 21st century and seem to grow with each passing year, as “anyone with access to a computer [can] compile a wealth of information about” anyone else, including such sensitive details as a person’s home address or the school attended by his children. Reed, 561 U. S., at 208 (Alito, J., concurring).

The gravity of the privacy concerns in this context is further underscored by the filings of hundreds of organizations as amici curiae in support of the petitioners. Far from representing uniquely sensitive causes, these organizations span the ideological spectrum, and indeed the full range of human endeavors: from the American Civil Liberties Union to the Proposition 8 Legal Defense Fund; from the Council on American-Islamic Relations to the Zionist Organization of America; from Feeding America—Eastern Wisconsin to PBS Reno. The deterrent effect feared by these organizations is real and pervasive, even if their concerns are not shared by every single charity operating or raising funds in California.

The dissent argues that—regardless of the defects in California’s disclosure regime—a facial challenge cannot succeed unless a plaintiff shows that donors to a substantial number of organizations will be subjected to harassment and reprisals. See post, at 21, n. 11. As we have explained, plaintiffs may be required to bear this evidentiary burden where the challenged regime is narrowly tailored to an important government interest. See supra, at 10–11. Such a demanding showing is not required, however, where—as here—the disclosure law fails to satisfy these criteria.

Finally, California’s demand for Schedule Bs cannot be saved by the fact that donor information is already disclosed to the IRS as a condition of federal tax-exempt status. For one thing, each governmental demand for disclosure brings with it an additional risk of chill. For another, revenue collection efforts and conferral of tax-exempt status may raise issues not presented by California’s disclosure requirement, which can prevent charities from operating in the State altogether. See Regan v. Taxation With Representation of Wash., 461 U.S. 540, 545 (1983); see also Schaumburg, 444 U. S., at 633 ( First Amendment protects right to solicit charitable contributions).

We are left to conclude that the Attorney General’s disclosure requirement imposes a widespread burden on donors’ associational rights. And this burden cannot be justified on the ground that the regime is narrowly tailored to investigating charitable wrongdoing, or that the State’s interest in administrative convenience is sufficiently important. We therefore hold that the up-front collection of Schedule Bs is facially unconstitutional, because it fails exacting scrutiny in “a substantial number of its applications . . . judged in relation to [its] plainly legitimate sweep.” Stevens, 559 U. S., at 473 (internal quotation marks omitted).

The dissent concludes by saying that it would be “sympathetic” if we “had simply granted as-applied relief to petitioners based on [our] reading of the facts.” Post, at 25. But the pertinent facts in these cases are the same across the board: Schedule Bs are not used to initiate investigations. That is true in every case. California has not considered alternatives to indiscriminate up-front disclosure. That is true in every case. And the State’s interest in amassing sensitive information for its own convenience is weak. That is true in every case. When it comes to the freedom of association, the protections of the First Amendment are triggered not only by actual restrictions on an individual’s ability to join with others to further shared goals. The risk of a chilling effect on association is enough, “[b]ecause First Amendment freedoms need breathing space to survive.” Button, 371 U. S., at 433.

*  *  *

The District Court correctly entered judgment in favor of the petitioners and permanently enjoined the Attorney General from collecting their Schedule Bs. The Ninth Circuit erred by vacating those injunctions and directing entry of judgment for the Attorney General. The judgment of the Ninth Circuit is reversed, and the cases are remanded for further proceedings consistent with this opinion.

It is so ordered.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 19–251 and 19–255

_________________

AMERICANS FOR PROSPERITY FOUNDATION, PETITIONER

19–251v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

THOMAS MORE LAW CENTER, PETITIONER

19–255v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

on writs of certiorari to the united states court of appeals for the ninth circuit

[July 1, 2021]

Justice Thomas, concurring in Parts I, II–A, II–B–2, and III–A, and concurring in the judgment.

The Court correctly holds that California’s disclosure requirement violates the First Amendment. It also correctly concludes that the District Court properly enjoined California’s attorney general from collecting the forms at issue, which contain sensitive donor information. But, while I agree with much of the Court’s opinion, I would approach three issues differently.

First, the bulk of “our precedents . . . require application of strict scrutiny to laws that compel disclosure of protected First Amendment association.” Doe v. Reed, 561 U.S. 186, 232 (2010) (Thomas, J., dissenting). California’s law fits that description. Although the Court rightly holds that even the less demanding “exacting scrutiny” standard requires narrow tailoring for laws that compel disclosure, ante, at 9–11, invoking exacting scrutiny is at odds with our repeated recognition “that privacy of association is protected under the First Amendment.” 561 U. S., at 240 (Thomas, J., dissenting). The text and history of the Assembly Clause suggest that the right to assemble includes the right to associate anonymously. See 4 Annals of Cong. 900–902, 941–942 (1795) (defending the Democratic-Republican societies, many of which met in secret, as exercising individuals’ “leave to assemble”); see also Brief for Becket Fund for Religious Liberty as Amicus Curiae 13–20; Reply Brief in No. 19–251, pp. 3–5; NAACP v. Alabama ex rel. Patterson, 357 U.S. 449 (1958) (discussing the history of anonymous publications). And the right to associate anonymously often operates as a vehicle to protect other First Amendment rights, such as the freedom of the press. McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 361–367 (1995) (Thomas, J., concurring) (“Founding-era Americans” understood the freedom of the press to include the right of printers and publishers not to be compelled to disclose the authors of anonymous works). Laws directly burdening the right to associate anonymously, including compelled disclosure laws, should be subject to the same scrutiny as laws directly burdening other First Amendment rights. Doe, 561 U. S., at 232, 240.

Second, the Court holds the law “overbroad” and, thus, invalid in all circumstances. Ante, at 16. But I continue to have “doubts about [the] origins and application” of our “overbreadth doctrine.” United States v. Sineneng-Smith, 590 U. S. ___, ___ (2020) (Thomas, J., concurring) (slip op., at 1). That doctrine purports to grant federal courts the power to invalidate a law “if a substantial number of its applications are unconstitutional, judged in relation to the statute’s plainly legitimate sweep.” Ibid. (internal quotation marks omitted). However, the Court has no power to enjoin the lawful application of a statute just because that statute might be unlawful as-applied in other circumstances. Id., at ___ (slip op., at 9); Borden v. United States, 593 U. S. ___, ___ (2021) (Thomas, J., concurring) (slip op., at 4) (“a court cannot, consistent with separation of powers, enjoin enforcement of a statute where enforcement would be lawful”). And the principle that application of a law is always unlawful if “ ‘a substantial number of its applications are unconstitutional’ ” “lacks any basis in the Constitution’s text” and “contravenes traditional standing principles.” Sineneng-Smith, 590 U. S., at ___ (Thomas, J., concurring) (slip op., at 1).

Third, and relatedly, this Court also lacks the power “to ‘pronounce that the statute is unconstitutional in all applications,’ ” even if the Court suspects that the law will likely be unconstitutional in every future application as opposed to just a substantial number of its applications. Borden, 593 U. S., at ___ (Thomas, J., concurring) (slip op., at 3) (quoting Chicago v. Morales, 527 U. S 41, 77 (1999) (Scalia, J., dissenting)). A declaration that the law is “facially” unconstitutional “seems to me no more than an advisory opinion—which a federal court should never issue at all.” 593 U. S., at ___ (Thomas, J., concurring) (slip op., at 3). Courts cannot “strike down statutory text” or resolve the legal rights of litigants not before them. Ibid. (brackets and internal quotation marks omitted).

Despite the Court’s use of the term “facially unconstitutional,” I join Part III–A, which finds that California’s law fails exacting scrutiny, because the Court does not say that it is “provid[ing] relief beyond the parties to the case.” Trump v. Hawaii, 585 U. S. ___, ___ (2018) (Thomas, J., concurring) (slip op., at 6). The Court simply (and correctly) holds that the District Court properly enjoined the law as applied to petitioners. Ante, at 19. The Court’s judgment is also not dependent on its overbreadth determination. Ante, at 12–15. One can understand the Court’s reasoning as based on the fundamental legal problems with the law (that are obvious in light of the facts of this suit) that will, in practice, prevent California from lawfully applying the disclosure requirement against a substantial number of entities, including petitioners.

With those points of difference clarified, I join Parts I, II–A, II–B–2, and III–A of the majority’s opinion and concur in the judgment.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 19–251 and 19–255

_________________

AMERICANS FOR PROSPERITY FOUNDATION, PETITIONER

19–251v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

THOMAS MORE LAW CENTER, PETITIONER

19–255v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

on writs of certiorari to the united states court of appeals for the ninth circuit

[July 1, 2021]

Justice Alito, with whom Justice Gorsuch joins, concurring in Parts I, II–A, II–B–2, and III, and concurring in the judgment.

I am pleased to join most of The Chief Justice’s opinion. In particular, I agree that the exacting scrutiny standard drawn from our election-law jurisprudence has real teeth. It requires both narrow tailoring and consideration of alternative means of obtaining the sought-after information. See ante, at 9–11, 14–15 (opinion of the Court). For the reasons The Chief Justice explains, California’s blunderbuss approach to charitable disclosures fails exacting scrutiny and is facially unconstitutional. See ante, at 12–19 (opinion of the Court). The question is not even close. And for the same reasons, California’s approach necessarily fails strict scrutiny.

The Chief Justice would hold that the particular exacting scrutiny standard in our election-law jurisprudence applies categorically “to First Amendment challenges to compelled disclosure.” Ante, at 7 (plurality opinion). Justice Thomas, by contrast, would hold that strict scrutiny applies in all such cases. See ante, at 1–2 (concurring opinion). I am not prepared at this time to hold that a single standard applies to all disclosure requirements. And I do not read our cases to have broadly resolved the question in favor of exacting scrutiny. This Court decided its seminal compelled disclosure cases before it developed modern strict scrutiny doctrine. See Fallon, Strict Judicial Scrutiny, 54 UCLA L. Rev. 1267, 1284 (2007) (“Before 1960, what we would now call strict judicial scrutiny . . . did not exist”); id., at 1282 (contending that modern strict scrutiny’s “first unambiguous appearance” in a majority opinion occurred in 1969). Accordingly, nothing in those cases can be understood as rejecting strict scrutiny. If anything, their language and reasoning—requiring a compelling interest and a minimally intrusive means of advancing that interest—anticipated and is fully in accord with contemporary strict scrutiny doctrine. See, e.g., Shelton v. Tucker, 364 U.S. 479, 488 (1960) (the government’s purpose “cannot be pursued by means that broadly stifle fundamental personal liberties when the end can be more narrowly achieved”); NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 463 (1958) (requiring a “compelling” interest (internal quotation marks omitted)). Similarly, Buckley v. Valeo, 424 U.S. 1 (1976) (per curiam), and its progeny should not be read to have broadly cabined our earlier decisions merely by relying on them in one particular context.

Because the choice between exacting and strict scrutiny has no effect on the decision in these cases, I see no need to decide which standard should be applied here or whether the same level of scrutiny should apply in all cases in which the compelled disclosure of associations is challenged under the First Amendment.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 19–251 and 19–255

_________________

AMERICANS FOR PROSPERITY FOUNDATION, PETITIONER

19–251v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

THOMAS MORE LAW CENTER, PETITIONER

19–255v.

ROB BONTA, ATTORNEY GENERAL OF CALIFORNIA

on writs of certiorari to the united states court of appeals for the ninth circuit

[July 1, 2021]

Justice Sotomayor, with whom Justice Breyer and Justice Kagan join, dissenting.

Although this Court is protective of First Amendment rights, it typically requires that plaintiffs demonstrate an actual First Amendment burden before demanding that a law be narrowly tailored to the government’s interests, never mind striking the law down in its entirety. Not so today. Today, the Court holds that reporting and disclosure requirements must be narrowly tailored even if a plaintiff demonstrates no burden at all. The same scrutiny the Court applied when NAACP members in the Jim Crow South did not want to disclose their membership for fear of reprisals and violence now applies equally in the case of donors only too happy to publicize their names across the websites and walls of the organizations they support.

California oversees nearly a quarter of this Nation’s charitable assets. As part of that oversight, it investigates and prosecutes charitable fraud, relying in part on a registry where it collects and keeps charitable organizations’ tax forms. The majority holds that a California regulation requiring charitable organizations to disclose tax forms containing the names and contributions of their top donors unconstitutionally burdens the right to associate even if the forms are not publicly disclosed.

In so holding, the Court discards its decades-long requirement that, to establish a cognizable burden on their associational rights, plaintiffs must plead and prove that disclosure will likely expose them to objective harms, such as threats, harassment, or reprisals. It also departs from the traditional, nuanced approach to First Amendment challenges, whereby the degree of means-end tailoring required is commensurate to the actual burdens on associational rights. Finally, it recklessly holds a state regulation facially invalid despite petitioners’ failure to show that a substantial proportion of those affected would prefer anonymity, much less that they are objectively burdened by the loss of it.

Today’s analysis marks reporting and disclosure requirements with a bull’s-eye. Regulated entities who wish to avoid their obligations can do so by vaguely waving toward First Amendment “privacy concerns.” Ante, at 17. It does not matter if not a single individual risks experiencing a single reprisal from disclosure, or if the vast majority of those affected would happily comply. That is all irrelevant to the Court’s determination that California’s Schedule B requirement is facially unconstitutional. Neither precedent nor common sense supports such a result. I respectfully dissent.

I

Charitable organizations that wish to solicit tax-deductible contributions from California residents must maintain membership in a registry managed by the California attorney general. Cal. Govt. Code Ann. §§12584, 12585 (West 2018). As a condition of membership, the attorney general requires charities to submit a complete copy of Internal Revenue Service (IRS) Form 990, including Schedule B, on which 501(c)(3) organizations report the names and contributions of their major donors. See Cal. Code Regs., tit. 11, §301 (2021); 26 CFR §§1.6033–2(a)(2)(ii), (iii) (2020). California regulations expressly require that Schedule Bs remain confidential, Cal. Code Regs., tit. 11, §310(b), and the attorney general’s office has implemented enhanced protocols to ensure confidentiality.[1] California relies on Schedule Bs to investigate fraud and other malfeasance.

After the attorney general’s office stepped up its efforts to enforce California’s Schedule B reporting requirement, petitioners Americans for Prosperity Foundation (Foundation) and Thomas More Law Center (Law Center) sought an injunction against the requirement. They alleged that the requirement “unconstitutionally burden[ed] their First Amendment right to free association by deterring individuals from financially supporting them.” Americans for Prosperity Foundation v. Becerra, 903 F.3d 1000, 1006 (CA9 2018). They pointed to evidence that their supporters experienced threats, reprisals, and harassment when their identities and associations became publicly known in other contexts. Importantly, however, the Foundation and Law Center failed to show that such consequences would result from the confidential submission of their top donors’ identities to California’s attorney general’s office in light of the security mechanisms the office has now implemented.

II

Because the freedom to associate needs “breathing space to survive,” NAACP v. Button, 371 U.S. 415, 433 (1963), this Court has recognized that associational rights must be “protected not only against heavy-handed frontal attack, but also from being stifled by more subtle governmental interference,” Bates v. Little Rock, 361 U.S. 516, 523 (1960). Publicizing individuals’ association with particular groups might expose members to harassment, threats, and reprisals by opponents of those organizations. Individuals may choose to disassociate themselves from a group altogether rather than face such backlash.

Acknowledging that risk, this Court has observed that “privacy in group association may in many circumstances be indispensable to preservation of freedom of association, particularly where a group espouses dissident beliefs.” NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 462 (1958). That observation places special emphasis on the risks actually resulting from disclosure. Privacy “may” be indispensable to the preservation of freedom of association, but it need not be. It depends on whether publicity will lead to reprisal. For example, privacy can be particularly important to “dissident” groups because the risk of retaliation against their supporters may be greater. For groups that promote mainstream goals and ideas, on the other hand, privacy may not be all that important. Not only might their supporters feel agnostic about disclosing their association, they might actively seek to do so.

Given the indeterminacy of how disclosure requirements will impact associational rights, this Court requires plaintiffs to demonstrate that a requirement is likely to expose their supporters to concrete repercussions in order to establish an actual burden. It then applies a level of means-end tailoring proportional to that burden. The Court abandons that approach here, instead holding that narrow tailoring applies to disclosure requirements across the board, even if there is no evidence that they burden anyone at all.

A

Before today, to demonstrate that a reporting or disclosure requirement would chill association, litigants had to show “a reasonable probability that the compelled disclosure of . . . contributors’ names will subject them to threats, harassment, or reprisals from either Government officials or private parties.” Buckley v. Valeo, 424 U.S. 1, 74 (1976) (per curiam). Proof could include “specific evidence of past or present harassment of members due to their associational ties, or of harassment directed against the organization itself,” ibid., as well as evidence that “fear of community hostility and economic reprisals that would follow public disclosure . . . had discouraged new members from joining” an organization or caused “former members to withdraw,” Bates, 361 U. S., at 524. Although the Court has never imposed an “unduly strict requiremen[t] of proof,” Buckley, 424 U. S., at 74, it has consistently required at least some record evidence demonstrating a risk of such objective harms. See Bates, 361 U. S., at 523–524; Patterson, 357 U. S., at 462–463.

Indeed, the Court has expressly held that parties do not have standing to bring claims where they assert nothing more than that government action will cause a “subjective ‘chill.’ ” Laird v. Tatum, 408 U.S. 1, 13–14 (1972). It does not matter if an individual perceives a government regulation “as inappropriate,” or believes “it is inherently dangerous for the [government] to be concerned with” a particular activity, or has “generalized yet speculative apprehensiveness that the [government] may at some future date misuse the information in some way that would cause direct harm” to her. Id., at 13. She must still allege a risk of objective harm. See id., at 14; see also Clapper v. Amnesty Int’l USA, 568 U.S. 398, 417–418 (2013).

Consistent with this approach, the Court has carefully scrutinized record evidence to determine whether a disclosure requirement actually risks exposing supporters to backlash. See Patterson, 357 U. S., at 462 (compelled disclosure of NAACP members “entail[ed] the likelihood of a substantial restraint” on association in light of “an uncontroverted showing” that past disclosures exposed members “to economic reprisal, loss of employment, threat of physical coercion, and other manifestations of public hostility”); Bates, 361 U. S., at 523–524 (compelled disclosure of NAACP membership “would work a significant interference with the freedom of association” based on “uncontroverted evidence” that past identification “had been followed by harassment and threats of bodily harm”); Shelton v. Tucker, 364 U.S. 479, 486 (1960) (disclosure of teachers’ organizational affiliations impaired association because record evidence substantiated a “fear of public disclosure” and a “constant and heavy” pressure on teachers “to avoid any ties which might displease those who control [their] professional destin[ies]”); Buckley, 424 U. S., at 69–70 (“any serious infringement” on associational rights caused by the compelled disclosure of contributors was “highly speculative” on the record before the Court).

Hence, in Doe v. Reed, 561 U.S. 186 (2010), the Court rejected a facial challenge to the public disclosure of referenda signatories on the ground that the “typical referendum” concerned revenue, budget, and tax policies unlikely to incite threats or harassment. Id., at 200–201. Any judge who has witnessed local fights over raising taxes, funding schools, building sewer systems, or rerouting roads can surely envisage signatories with reason to keep their support for such measures private. But in Reed, such subjective reasons did not suffice to establish a cognizable burden on associational rights.

Today, the Court abandons the requirement that plain-tiffs demonstrate that they are chilled, much less that they are reasonably chilled. Instead, it presumes (contrary to the evidence, precedent, and common sense) that all disclosure requirements impose associational burdens. For example, the Court explains that there is a risk of chill in this suit because the government requires disclosure of the identity of any donor “with reason to remain anonymous.” Ante, at 17. The Court does not qualify that statement, nor does it require record evidence of such reasons. If the Court did, it would not be able to strike California’s Schedule B requirement down in all its applications, because the only evidence in the record of donors with any reason to remain anonymous is that of petitioners’.[2]

At best, then, a subjective preference for privacy, which previously did not confer standing, now subjects disclosure requirements to close scrutiny. Of course, all disclosure requires some loss of anonymity, and courts can always imagine that someone might, for some reason, prefer to keep their donations undisclosed. If such speculation is enough (and apparently it is), then all disclosure requirements ipso facto impose cognizable First Amendment burdens.

Indeed, the Court makes obvious its presumption that all disclosure requirements are burdensome by beginning its analysis of “burden” with an evaluation of means-end fit instead. “[A] reasonable assessment of the burdens imposed by disclosure,” the Court explains, “should begin with an understanding of the extent to which the burdens are unnecessary, and that requires narrow tailoring.” Ante, at 11; see also ante, at 17–18 (excusing plaintiffs from showing any burden if a disclosure law is not narrowly tailored).

I disagree. A reasonable assessment of the burdens imposed by disclosure should begin by determining whether those burdens even exist. If a disclosure requirement imposes no burdens at all, then of course there are no “unnecessary” burdens. Likewise, if a disclosure requirement imposes no burden for the Court to remedy, there is no need for it to be closely scrutinized. By forgoing the requirement that plaintiffs adduce evidence of tangible burdens, such as increased vulnerability to harassment or reprisals, the Court gives itself license to substitute its own policy preferences for those of politically accountable actors.

B

All this would be less troubling if the Court still required means-end tailoring commensurate to the actual burden imposed. It does not. Instead, it adopts a new rule that every reporting or disclosure requirement be narrowly tailored. See ante, at 9 (“While exacting scrutiny does not require that disclosure regimes be the least restrictive means of achieving their ends, it does require that they be narrowly tailored to the government’s asserted interest”).

1

Disclosure requirements burden associational rights only indirectly and only in certain contexts. For that reason, this Court has never necessarily demanded such requirements to be narrowly tailored. Rather, it has reserved such automatic tailoring for state action that “directly and immediately affects associational rights.” Boy Scouts of America v. Dale, 530 U.S. 640, 659 (2000); see also Buckley, 424 U. S., at 22, 25 (requiring a “closely drawn” fit for political contribution limits, which directly “limit one important means of associating with a candidate or committee”). When it comes to reporting and disclosure requirements, the Court has instead employed a more flexible approach, which it has named “exacting scrutiny.” See ante, at 7–8 (opinion of Roberts, C. J.).

Exacting scrutiny requires two things: first, there must be “ ‘a “substantial relation” between the disclosure requirement and a “sufficiently important” government interest,’ ” and second, “ ‘the strength of the governmental interest must reflect the seriousness of the actual burden on First Amendment rights.’ ” Reed, 561 U. S., at 196. Exacting scrutiny thus incorporates a degree of flexibility into the means-end analysis. The more serious the burden on First Amendment rights, the more compelling the government’s interest must be, and the tighter must be the fit between that interest and the government’s means of pursuing it. By contrast, a less substantial interest and looser fit will suffice where the burden on First Amendment rights is weaker (or nonexistent). In other words, to decide how closely tailored a disclosure requirement must be, courts must ask an antecedent question: How much does the disclosure requirement actually burden the freedom to associate?

This approach reflects the longstanding principle that the requisite level of scrutiny should be commensurate to the burden a government action actually imposes on First Amendment rights. See, e.g., Burdick v. Takushi, 504 U.S. 428, 434 (1992) (“[T]he rigorousness of our inquiry . . . depends upon the extent to which a challenged regulation burdens” First Amendment rights); Board of Trustees of State Univ. of N. Y. v. Fox, 492 U.S. 469, 477 (1989) (“[C]ommercial speech enjoys a limited measure of protection, commensurate with its subordinate position in the scale of First Amendment values, and is [thus] subject to modes of regulation that might be impermissible in the realm of noncommercial expression” (internal quotation marks and alterations omitted)); see also Fulton v. Philadelphia, 593 U. S. ___, ___ (2021) (Barrett, J., concurring) (slip op., at 2) (noting the “nuanced” approach the Court generally takes in the “resolution of conflicts between generally applicable laws and . . . First Amendment rights”).

Compare, for instance, the Court’s approaches in Shelton v. Tucker and Doe v. Reed. At issue in Shelton was an Arkansas statute passed in 1958 that compelled all public school teachers, as a condition of employment, to submit annually a list of every organization to which they belonged or regularly contributed. 364 U. S., at 480–481. The Court held that the disclosure requirement “comprehensive[ly] interfere[d] with associational freedom,” because record evidence demonstrated a significant risk that the information would be publicly disclosed, and such disclosure could lead to public pressure on school boards “to discharge teachers who belong to unpopular or minority organizations.” Id., at 486–487, 490. Arkansas’s statute did not require that the information remain confidential; each school board was “free to deal with the information as it wishe[d].” Id., at 486. Indeed, “a witness who was a member of the Capital Citizens[’] Council” (an organization dedicated to resisting school integration)[3] “testified that his group intended to gain access” to the teachers’ affidavits “with a view to eliminating from the school system persons who supported organizations unpopular with the group.” Ibid., n. 7. Moreover, a starkly asymmetric power dynamic existed between teachers, who were “hired on a year-to-year basis,” and the hiring authorities to whom their membership lists were submitted. Id., at 482. The Arkansas Legislature had made no secret of its desire for teachers’ disclosures to be used for hiring and firing decisions. One year after enacting the disclosure requirement at issue in Shelton, the legislature enacted another provision that made it outright unlawful for state governmental bodies to employ members of the NAACP. Shelton v. McKinley, 174 F. Supp. 351, 353–354 (ED Ark. 1959). It is thus unsurprising that the Court found that Arkansas teachers would feel a “constant and heavy” pressure “to avoid any ties which might displease those who control [their] professional destin[ies].” Id., at 486; see also Keyishian v. Board of Regents of Univ. of State of N. Y., 385 U.S. 589, 604 (1967) (“When one must guess what conduct or utterance may lose him his position, one necessarily will steer far wider of the [impermissible] zone” (internal quotation marks omitted)). Because Arkansas’s purpose (ensuring teachers’ fitness) was “pursued by means that broadly stifle fundamental personal liberties,” the Court demanded that Arkansas “more narrowly achiev[e]” its interest. Shelton, 364 U. S., at 488.

Now consider this Court’s approach in Reed. Reed involved a facial challenge to a Washington law permitting the public disclosure of referendum petitions that included signatories’ names and addresses. 561 U. S., at 190–191. The Court found that Washington had a number of other mechanisms in place to pursue its stated interest in preventing fraudulent referendum signatures. For instance, the secretary of state was charged with verifying and canvassing the names on referendum petitions, advocates and opponents of a measure could observe the canvassing process, and citizens could challenge the secretary’s actions in court. Id., at 198. Publicly disclosing referendum signatories was thus a mere backstop, giving citizens the opportunity to catch the secretary’s mistakes. Had Washington been required to achieve its interests narrowly, as in Shelton, it is unlikely the disclosure requirement would have survived.[4]

In crucial contrast to Shelton, however, the Reed Court found “scant evidence” that disclosure exposed signatories of typical referendums to “threats, harassment, or reprisals from either Government officials or private parties.” 561 U. S., at 200–201 (internal quotation marks omitted). Given the “modest burdens” imposed by the requirement, id., at 201, the Court required a correspondingly modest level of tailoring. Under that standard, the disclosure requirement passed muster, and the Court refused to facially strike it down.

The public disclosure regimes in both Shelton and Reed served important government goals. Yet the Court’s assessment of each differed considerably because the First Amendment burdens differed. This flexible approach is necessary because not all reporting and disclosure regimes burden associational rights in the same way.

2

The Court now departs from this nuanced approach in favor of a “one size fits all” test. Regardless of whether there is any risk of public disclosure, and no matter if the burdens on associational rights are slight, heavy, or nonexistent, disclosure regimes must always be narrowly tailored.

The Court searches in vain to find a foothold for this new approach in precedent. The Court first seizes on Shelton’s statement that a governmental interest “ ‘cannot be pursued by means that broadly stifle fundamental personal liberties when the end can be more narrowly achieved.’ ” Ante, at 9 (quoting 364 U. S., at 488). The Court could not have

cherry-picked a less helpful quote. By its own terms, Shelton held that an end must be “more narrowly achieved” only if the means “broadly stifle” First Amendment liberties, that is, only if the means impose a severe burden on associational rights.[5]

In any event, the Court need not read a few isolated sentences from that opinion to divine Shelton’s meaning. As described, see Part II–B–1, supra, the Court in Shelton concluded that a reasonable “fear of public disclosure” and an asymmetric power dynamic with hiring authorities would result in a “constant and heavy” pressure on teachers “to avoid any ties which might displease those who control [their] professional destin[ies].” 364 U. S., at 486. Recall that a witness had testified that his white supremacist organization would seek to obtain the identities of teachers working on civil rights issues in order to eradicate them from the school system, and that just a year after Arkansas enacted its disclosure law, it enacted a law prohibiting the hiring of members of the NAACP as public school teachers. The problem was not the breadth of the inquiry; it was the significant risk that teachers would face serious repercussions for their disclosed associations.[6]

The Court next looks to McCutcheon v. Federal Election Comm’n, 572 U.S. 185 (2014), which addressed political contribution limits, not disclosure regimes. It is no surprise that the Court subjected the former to narrow tailoring, as Buckley had already held that contribution limits directly “impinge on protected associational freedoms.” 424 U. S., at 22; see also McCutcheon, 572 U. S., at 204 (explaining that aggregate limits on contributions “diminish an individual’s right of political association” by “limit[ing] the number of candidates he supports” or the amount of money he gives). Buckley itself distinguished the First Amendment burdens of disclosure requirements and contribution limits. See 424 U. S., at 64 (noting that, unlike contribution limits, “disclosure requirements impose no ceiling on campaign-related activities” and concluding only that compelled disclosure “can” infringe associational rights). Apparently, those distinctions no longer matter.

Neither Shelton nor McCutcheon, then, supports the idea that all disclosure requirements must be narrowly tailored. McCutcheon arose in the context of a direct limit on associational freedoms, while the law in Shelton “broadly stifle[d]” associational rights. Ignoring these distinctions, the Court decides that it will indiscriminately require narrow tailoring for every single disclosure regime. The Court thus trades precision for blunt force, creating a significant risk that it will topple disclosure regimes that should be constitutional, and that, as in Reed, promote important governmental interests.

III

A

Under a First Amendment analysis that is faithful to this Court’s precedents, California’s Schedule B requirement is constitutional. Begin with the burden it imposes on associational rights. Petitioners have unquestionably provided evidence that their donors face a reasonable probability of threats, harassment, and reprisals if their affiliations are made public. See ante, at 4. California’s Schedule B regulation, however, is a nonpublic reporting requirement, and California has implemented security measures to ensure that Schedule B information remains confidential.[7]

Nor have petitioners shown that their donors, or any organization’s donors, will face threats, harassment, or reprisals if their names remain in the hands of a few California state officials. The Court notes that, under Shelton, disclosure requirements can chill association even absent public disclosure. See ante, at 16. In Shelton, however, there was a serious concern that hiring authorities would punish teachers for their organizational affiliations. See 364 U. S., at 486. By contrast, the Court in no way suggests that California officials will use Schedule B information to retaliate against any organization’s donors. If California’s reporting requirement imposes any burden at all, it is at most a very slight one.

B

1

Given the modesty of the First Amendment burden, California may justify its Schedule B requirement with a correspondingly modest showing that the means achieve its ends. See Reed, 561 U. S., at 201. California easily meets this standard.

California collects Schedule Bs to facilitate supervision of charities that operate in the State. As the Court acknowledges, this is undoubtedly a significant governmental interest. See ante, at 12–13. In the United States, responsibility for overseeing charities has historically been vested in States’ attorneys general, who are tasked with prosecuting charitable fraud, self-dealing, and misappropriation of charitable funds. Effective policing is critical to maintaining public confidence in, and continued giving to, charitable organizations. California’s interest in exercising such oversight is especially compelling given the size of its charitable sector. Nearly a quarter of the country’s charitable assets are held by charities registered in California. Brief for Scholars of the Law of Non-Profit Organizations as Amici Curiae 10; see ibid. (“As of June 2018, charities registered in California reported $295 billion in annual income and net assets of $851 billion”).

The Schedule B reporting requirement is properly tailored to further California’s efforts to police charitable fraud. See Reed, 561 U. S., at 198–199 (noting that disclosure “helps” combat fraud, even if it is not the least restrictive method of doing so). The IRS Schedule B form requires organizations to disclose the names and addresses of their major donors, the total amount of their contributions, and whether the donation was cash or in-kind. See 26 CFR §§1.6033–2(a)(2)(ii), (iii). If the gift is in-kind, Schedule B requires a description of the property and its fair market value.

Schedule B and other parts of Form 990 help attorneys in the Charitable Trusts Section of the California Department of Justice (Section) uncover whether an officer or director of a charity is engaged in self-dealing, or whether a charity has diverted donors’ charitable contributions for improper use. Appellant-Cross-Appellee’s Excerpts of Record in No. 16–55727 etc. (CA9), pp. 575, 716–718 (16–55727 ER). It helps them determine whether a donor is using the charity as a pass-through entity, including as a source of improper loans that the donor repays as a contribution. Id., at 577–578, 716–718. It helps them identify red flags, such as discrepancies in reporting contributions across different schedules. Id., at 578–579, 716–718. And it helps them determine whether a charity has inflated the value of a donor’s in-kind contribution in order, for instance, to overstate how efficiently the charity expends resources. Id., at 721–727.

As a former head of the Section described it, Schedule B combined with the rest of Form 990 provides “[a] roadmap to the rest of the investigation that follows.” Id., at 717. Indeed, having Schedule Bs on hand is important to attorneys’ decisions regarding whether to advance an investigation at all. One of the first things an auditor or lawyer does upon receiving a complaint is review the entire Form 990, including Schedule B. Id., at 969–970, 996–997, 1062–1063. One Section leader testified that she used Schedule Bs “[a]ll the time” for this purpose. App. in No. 19–251, p. 413.

In sum, the evidence shows that California’s confidential reporting requirement imposes trivial burdens on petitioners’ associational rights and plays a meaningful role in Section attorneys’ ability to identify and prosecute charities engaged in malfeasance. That is more than enough to satisfy the First Amendment here.

2

Much of the Court’s tailoring analysis is categorically inappropriate under the correct standard of review. In any event, the Court greatly understates the importance to California of collecting information on charitable organizations’ top donors.

The Court claims that the collection of Schedule Bs does not form an “integral” part of California’s fraud detection efforts and has never done “ ‘anything’ ” to advance investigative efforts.[8] Ante, at 13. The record reveals otherwise. As discussed, Section leaders report that they use Schedule Bs “[a]ll the time” and rely on them to create roadmaps for their investigations. App. in No. 19–251, at 413; see also 16–55727 ER, at 717. The Court further complains that California does not rely on Schedule Bs to “initiate” investigations. Ante, at 15, 19. But disclosure assists California in its decisions whether to advance or end an investigation. Perhaps the Court’s main concern is that California has not identified enough instances in which Schedule B played a unique role in prosecuting charitable malfeasance. But “[l]ike a jigsaw puzzle,” investigations often advance “only by placing in the proper place the many pieces of evidence that, taken singly, would show comparatively little.” Andresen v. Maryland, 427 U.S. 463, 481, n. 10 (1976).

The Court next insists that California can rely on alternative mechanisms, such as audit letters or subpoenas, to obtain Schedule B information. But the Section receives as many as 100 charity-related complaints a month. App. in No. 19–251, at 307. It is not feasible for the Section, which has limited staff and resources, to conduct that many audits. See Appellant-Cross-Appellee’s Excerpts of Record in No. 16–56855 etc. (CA9), pp. 512–513. The subpoena process is also time consuming: Letters must go through multiple layers of review and waiting for a response causes further delays during which a charity can continue its malfeasance. App. in No. 19–251, at 412.

Implicitly acknowledging that audits and subpoenas are more cumbersome and time consuming, the Court trivializes the State’s interest in what it calls “ease of administration.” Ante, at 15. Yet in various contexts, the Court has recognized that an interest in “efficiency” is critical to the effective operation of public agencies.[9] See, e.g., Bailey v. United States, 568 U.S. 186, 200 (2013) (“[T]he law enforcement interests in conducting a safe and efficient search” justify detaining “occupants on the premises during the execution of a search warrant”); Civil Service Comm’n v. Letter Carriers, 413 U.S. 548, 564 (1973) (seeking a constitutional balance between the interests of a government employee in commenting on matters of public concern and the interest of the government in the efficiency of the services it performs).

In addition to being burdensome, audit letters and subpoenas can also significantly undercut the Section’s work by alerting an organization to the existence of an investigation, giving it a chance to hide assets or tamper with evidence. The Court dismisses this concern as unsubstantiated. Yet one Section head reported that this had “happened several times,” and another testified to her personal experience with organizations “fabricat[ing]” and “destroy[ing] records” after being tipped off to an investigation. 16–55727 ER, at 590, 998–999.[10] A State surely has a compelling interest in ensuring that the subject of an investigation does not destroy evidence or hide funds before investigators have an opportunity to find them. Cf. Kentucky v. King, 563 U.S. 452, 460 (2011) (“[T]he need to prevent the imminent destruction of evidence has long been recognized as a sufficient justification for a warrantless search” (internal quotation marks omitted)). The Court ignores those interests here.

IV

In a final coup de grâce, the Court concludes that California’s reporting requirement is unconstitutional not just as applied to petitioners, but on its very face. “In the First Amendment context,” such broad relief requires proof that the requirement is unconstitutional in “ ‘a substantial number of . . . applications . . . , judged in relation to the statute’s plainly legitimate sweep.’ ” United States v. Stevens, 559 U.S. 460, 473 (2010). “Facial challenges are disfavored for several reasons,” prime among them because they “often rest on speculation.” Washington State Grange v. Washing-

ton State Republican Party, 552 U.S. 442, 450 (2008). Speculation is all the Court has. The Court points to not a single piece of record evidence showing that California’s reporting requirement will chill “a substantial number” of top donors from giving to their charities of choice.[11] Yet it strikes the requirement down in every application.

The average donor is probably at most agnostic about having their information confidentially reported to California’s attorney general. A significant number of the charities registered in California engage in uncontroversial pursuits. They include hospitals and clinics; educational institutions; orchestras, operas, choirs, and theatrical groups; museums and art exhibition spaces; food banks and other organizations providing services to the needy, the elderly, and the disabled; animal shelters; and organizations that help maintain parks and gardens. Brief for Public Citizen et al. as Amici Curiae 12–13. It is somewhat hard to fathom that donors to the Anderson Elementary School PTA, the Loomis-Eureka Lakeside Little League, or the Santa Barbara County Horticultural Society (“[c]elebrating plants since 1880”) are less likely to give because their do-

nations are confidentially reported to California’s Charitable Trusts Section.[12]

In fact, research shows that the vast majority of donors prefer to publicize their charitable contributions. See Drennan, Where Generosity and Pride Abide: Charitable Naming Rights, 80 U. Cin. L. Rev. 45, 50 (2011) (“Research reveals that anonymous largesse from the wealthy has become rare”); Posner, Altruism, Status, and Trust in the Law of Gifts and Gratuitous Promises, 1997 Wis. L. Rev. 567, 574, n. 17 (“[C]haritable gifts are rarely made anonymously”). One study found that anonymous gifting accounted for less than 1% of all donations to Yale Law School, Harvard Law School, and Carnegie Mellon University. Glazer & Konrad, A Signaling Explanation for Charity, 86 Am. Econ. Rev. 1019, 1021 (1996). Symptomatic of this trend is the explosion in charitable naming rights since the mid-1990s. Drennan, 80 U. Cin. L. Rev., at 50, 55. As one author has recounted, “every nook and cranny of [public] buildings” is now “tagged by some wealthy, generous and obviously not publicity-shy donor.” Isherwood, The Graffiti of the Philanthropic Class, N. Y. Times, Dec. 2, 2007.

Of course, it is always possible that an organization is inherently controversial or for an apparently innocuous organization to explode into controversy. The answer, however, is to ensure that confidentiality measures are sound or, in the case of public disclosures, to require a procedure for governments to address requests for exemptions in a timely manner. It is not to hamper all government law enforcement efforts by forbidding confidential disclosures en masse.

Indeed, this Court has already rejected such an indiscriminate approach in the specific context of disclosure requirements. Just over a decade ago, in Reed, petitioners demonstrated that their own supporters would face reprisal if their opposition to expanding domestic partnership laws became public. That evidence did not support a facial challenge to Washington’s public disclosure law, however, because the “typical referendum petitio[n] concern[ed] tax policy, revenue, budget, or other state law issues,” and “there [was] no reason to assume that any burdens imposed by disclosure of typical referendum petitions would be remotely like the burdens plaintiffs fear in this case.” 561 U. S., at 200–201 (internal quotation marks omitted); see also id., at 202–203 (Alito, J., concurring) (“Many referendum petitions concern relatively uncontroversial matters, and plaintiffs have provided no reason to think that disclosure of signatory information in those contexts would significantly chill the willingness of voters to sign. Plaintiffs’ facial challenge therefore must fail” (citation omitted)).

So too here. Many charitable organizations “concern relatively uncontroversial matters” and petitioners “have provided no reason to think that” confidential disclosure of donor information “would significantly chill the willingness of ” most donors to give. Nor does the Court provide such a reason. It merely highlights threats that public disclosure would pose to these two petitioners’ supporters. Those threats provide “scant evidence” of anything beyond “the specific harm” that petitioners’ donors might experience were their Schedule B information publicly disclosed. Id., at 200–201. Petitioners’ “facial challenge therefore must fail.” Id., at 203 (Alito, J., concurring).

How, then, can their facial challenge succeed? Only because the Court has decided, in a radical departure from precedent, that there no longer need be any evidence that a disclosure requirement is likely to cause an objective burden on First Amendment rights before it can be struck down.

*  *  *

Today’s decision discards decades of First Amendment jurisprudence recognizing that reporting and disclosure requirements do not directly burden associational rights. There is no other explanation for the Court’s conclusion that, first, plaintiffs do not need to show they are actually burdened by a disclosure requirement; second, every disclosure requirement demands narrow tailoring; and third, a facial challenge can succeed in the absence of any evidence a state law burdens the associational rights of a substantial proportion of affected individuals.

That disclosure requirements directly burden associational rights has been the view of Justice Thomas, see id., at 232 (dissenting opinion), but it has never been the view of this Court. Just 11 years ago, eight Members of the Court, including two Members of the current majority, recognized that disclosure requirements do not directly interfere with First Amendment rights. See id., at 196 (majority opinion). In an opinion barely mentioned in today’s decision, the Court in Reed did the opposite of what the Court does today. First, it demanded objective evidence that disclosure risked exposing supporters to threats and reprisals; second, it required only a loose means-end fit in light of the “modest” burden it found; and third, it rejected a facial challenge given petitioners’ failure to establish that signatories to the “typical” referendum had any reason to fear disclosure. See id., at 200–201. In so doing, the Court ensured that it would not “short circuit the democratic process by preventing laws embodying the will of the people from being implemented in a manner consistent with the Constitution.” Washington State Grange, 552 U. S., at 451.

The Court 11 years later apparently has a different view of its role. It now calls upon the federal courts to serve “as virtually continuing monitors of the wisdom and soundness of [governmental] action.” Laird, 408 U. S., at 15. There is no question that petitioners have shown that their donors reasonably fear reprisals if their identities are publicly exposed. The Court and I, however, disagree about the likelihood of that happening and the role Schedule Bs play in the investigation of charitable malfeasance. If the Court had simply granted as-applied relief to petitioners based on its reading of the facts, I would be sympathetic, although my own views diverge. But the Court’s decision is not nearly so narrow or modest. Instead, the Court jettisons completely the longstanding requirement that plaintiffs demonstrate an actual First Amendment burden before the Court will subject government action to close scrutiny. It then invalidates a regulation in its entirety, even though it can point to no record evidence demonstrating that the regulation is likely to chill a substantial proportion of donors. These moves are wholly inconsistent with the Court’s precedents and our Court’s long-held view that disclosure requirements only indirectly burden First Amendment rights. With respect, I dissent.