������ýapp�ٷ�

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

_________________

No. 19–631

_________________

WILLIAM P. BARR, ATTORNEY GENERAL, et al., PETITIONERS v. AMERICAN ASSOCIATION OF POLITICAL CONSULTANTS, INC., et al.

on writ of certiorari to the united states court of appeals for the fourth circuit

[July 6, 2020]

Justice Kavanaugh announced the judgment of the Court and delivered an opinion, in which The Chief Justice and Justice Alito join, and in which Justice Thomas joins as to Parts I and II.

Americans passionately disagree about many things. But they are largely united in their disdain for robocalls. The Federal Government receives a staggering number of complaints about robocalls—3.7 million complaints in 2019 alone. The States likewise field a constant barrage of complaints.

For nearly 30 years, the people’s representatives in Congress have been fighting back. As relevant here, the Telephone Consumer Protection Act of 1991, known as the TCPA, generally prohibits robocalls to cell phones and home phones. But a 2015 amendment to the TCPA allows robocalls that are made to collect debts owed to or guaranteed by the Federal Government, including robocalls made to collect many student loan and mortgage debts.

This case concerns robocalls to cell phones. Plaintiffs in this case are political and nonprofit organizations that want to make political robocalls to cell phones. Invoking the First Amendment, they argue that the 2015 government-debt exception unconstitutionally favors debt-collection speech over political and other speech. As relief from that unconstitutional law, they urge us to invalidate the entire 1991 robocall restriction, rather than simply invalidating the 2015 government-debt exception.

Six Members of the Court today conclude that Congress has impermissibly favored debt-collection speech over political and other speech, in violation of the First Amendment. See infra, at 6–9; post, at 1–2 (Sotomayor, J., concurring in judgment); post, at 1, 3 (Gorsuch, J., concurring in judgment in part and dissenting in part). Applying traditional severability principles, seven Members of the Court conclude that the entire 1991 robocall restriction should not be invalidated, but rather that the 2015 government-debt exception must be invalidated and severed from the remainder of the statute. See infra, at 10–25; post, at 2 (Sotomayor, J., concurring in judgment); post, at 11–12 (Breyer, J., concurring in judgment with respect to severability and dissenting in part). As a result, plaintiffs still may not make political robocalls to cell phones, but their speech is now treated equally with debt-collection speech. The judgment of the U. S. Court of Appeals for the Fourth Circuit is affirmed.

I

A

In 1991, Congress passed and President George H. W. Bush signed the Telephone Consumer Protection Act. The Act responded to a torrent of vociferous consumer complaints about intrusive robocalls. A growing number of telemarketers were using equipment that could automatically dial a telephone number and deliver an artificial or prerecorded voice message. At the time, more than 300,000 solicitors called more than 18 million Americans every day. TCPA, §2, ¶¶3, 6, 105Stat. 2394, note following 47 U. S. C. §227. Consumers were “outraged” and considered robocalls an invasion of privacy “regardless of the content or the initiator of the message.” ¶¶6, 10.

A leading Senate sponsor of the TCPA captured the zeitgeist in 1991, describing robocalls as “the scourge of modern civilization. They wake us up in the morning; they interrupt our dinner at night; they force the sick and elderly out of bed; they hound us until we want to rip the telephone right out of the wall.” 137 Cong. Rec. 30821 (1991).

In enacting the TCPA, Congress found that banning robocalls was “the only effective means of protecting telephone consumers from this nuisance and privacy invasion.” TCPA §2, ¶12. To that end, the TCPA imposed various restrictions on the use of automated telephone equipment. §3(a), 105Stat. 2395. As relevant here, one restriction prohibited “any call (other than a call made for emergency purposes or made with the prior express consent of the called party) using any automatic telephone dialing system or an artificial or prerecorded voice” to “any telephone number assigned to a paging service, cellular telephone service, specialized mobile radio service, or other radio common carrier service, or any service for which the called party is charged for the call.” Id., at 2395–2396 (emphasis added). That provision is codified in §227(b)(1)(A)(iii) of Title 47 of the U. S. Code.

In plain English, the TCPA prohibited almost all robocalls to cell phones.[1]

Twenty-four years later, in 2015, Congress passed and President Obama signed the Bipartisan Budget Act. In addition to making other unrelated changes to the U. S. Code, that Act amended the TCPA’s restriction on robocalls to cell phones. It stated:

“(a) In General.—Section 227(b) of the Communications Act of 1934 (47 U. S. C. 227(b)) is amended—

(1) in paragraph (1)—

(A) in subparagraph (A)(iii), by inserting ‘, unless such call is made solely to collect a debt owed to or guaranteed by the United States’ after ‘charged for the call.’ ” 129Stat. 588.[2]

In other words, Congress carved out a new government-debt exception to the general robocall restriction.

The TCPA imposes tough penalties for violating the robocall restriction. Private parties can sue to recover up to $1,500 per violation or three times their actual monetary losses, which can add up quickly in a class action. §227(b)(3). States may bring civil actions against robocallers on behalf of their citizens. §227(g)(1). And the Federal Communications Commission can seek forfeiture penalties for willful or repeated violations of the statute. §503(b).

B

Plaintiffs in this case are the American Association of Political Consultants and three other organizations that participate in the political system. Plaintiffs and their members make calls to citizens to discuss candidates and issues, solicit donations, conduct polls, and get out the vote. Plaintiffs believe that their political outreach would be more effective and efficient if they could make robocalls to cell phones.[3] But because plaintiffs are not in the business of collecting government debt, §227(b)(1)(A)(iii) prohibits them from making those robocalls.

Plaintiffs filed a declaratory judgment action against the U. S. Attorney General and the FCC, claiming that §227(b)(1)(A)(iii) violated the First Amendment. The U. S. District Court for the Eastern District of North Carolina determined that the robocall restriction with the government-debt exception was a content-based speech regulation, thereby triggering strict scrutiny. But the court concluded that the law survived strict scrutiny, even with the content-based exception, because of the Government’s compelling interest in collecting debt.

The U. S. Court of Appeals for the Fourth Circuit vacated the judgment. American Assn. of Political Consultants, Inc. v. FCC, 923 F.3d 159 (2019). The Court of Appeals agreed with the District Court that the robocall restriction with the government-debt exception was a content-based speech restriction. But the court held that the law could not withstand strict scrutiny and was therefore unconstitutional. The Court of Appeals then applied traditional severability principles and concluded that the government-debt exception was severable from the underlying robocall restriction. The Court of Appeals therefore invalidated the government-debt exception and severed it from the robocall restriction.

The Government petitioned for a writ of certiorari because the Court of Appeals invalidated part of a federal statute—namely, the government-debt exception. Plaintiffs supported the petition, arguing from the other direction that the Court of Appeals did not go far enough in providing relief and should have invalidated the entire 1991 robocall restriction rather than simply invalidating the 2015 government-debt exception. We granted certiorari. 589 U. S. ___ (2020).

II

Ratified in 1791, the First Amendment provides that Congress shall make no law “abridging the freedom of speech.” Above “all else, the First Amendment means that government” generally “has no power to restrict expression because of its message, its ideas, its subject matter, or its content.” Police Dept. of Chicago v. Mosley, 408 U.S. 92, 95 (1972).

The Court’s precedents allow the government to “constitutionally impose reasonable time, place, and manner regulations” on speech, but the precedents restrict the government from discriminating “in the regulation of expression on the basis of the content of that expression.” Hudgens v. NLRB, 424 U.S. 507, 520 (1976). Content-based laws are subject to strict scrutiny. See Reed v. Town of Gilbert, 576 U.S. 155, 163–164 (2015). By contrast, content-neutral laws are subject to a lower level of scrutiny. Id., at 166.

Section 227(b)(1)(A)(iii) generally bars robocalls to cell phones. Since the 2015 amendment, the law has exempted robocalls to collect government debt. The initial First Amendment question is whether the robocall restriction, with the government-debt exception, is content-based. The answer is yes.

As relevant here, a law is content-based if “a regulation of speech ‘on its face’ draws distinctions based on the message a speaker conveys.” Reed, 576 U. S., at 163. That description applies to a law that “singles out specific subject matter for differential treatment.” Id., at 169. For example, “a law banning the use of sound trucks for political speech—and only political speech—would be a content-based regulation, even if it imposed no limits on the political viewpoints that could be expressed.” Ibid.; see, e.g., Simon & Schuster, Inc. v. Members of N. Y. State Crime Victims Bd., 502 U.S. 105, 116 (1991); Arkansas Writers’ Project, Inc. v. Ragland, 481 U.S. 221, 229–230 (1987); Widmar v. Vincent, 454 U.S. 263, 265, 276–277 (1981); Carey v. Brown, 447 U.S. 455, 459–463 (1980); Erznoznik v. Jacksonville, 422 U.S. 205, 211–212 (1975); Mosley, 408 U. S., at 95–96.

Under §227(b)(1)(A)(iii), the legality of a robocall turns on whether it is “made solely to collect a debt owed to or guaranteed by the United States.” A robocall that says, “Please pay your government debt” is legal. A robocall that says, “Please donate to our political campaign” is illegal. That is about as content-based as it gets. Because the law favors speech made for collecting government debt over political and other speech, the law is a content-based restriction on speech.

The Government advances three main arguments for deeming the statute content-neutral, but none is persuasive.

First, the Government suggests that §227(b)(1)(A)(iii) draws distinctions based on speakers (authorized debt collectors), not based on content. But that is not the law in front of us. This statute singles out calls “made solely to collect a debt owed to or guaranteed by the United States,” not all calls from authorized debt collectors.

In any event, “the fact that a distinction is speaker based” does not “automatically render the distinction content neutral.” Reed, 576 U. S., at 170; Sorrell v. IMS Health Inc., 564 U.S. 552, 563–564 (2011). Indeed, the Court has held that “ ‘ laws favoring some speakers over others demand strict scrutiny when the legislature’s speaker preference reflects a content preference.’ ” Reed, 576 U. S., at 170 (quoting Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 658 (1994)).

Second, the Government argues that the legality of a robocall under the statute depends simply on whether the caller is engaged in a particular economic activity, not on the content of speech. We disagree. The law here focuses on whether the caller is speaking about a particular topic. In Sorrell, this Court held that a law singling out pharmaceutical marketing for unfavorable treatment was content-based. 564 U. S., at 563–564. So too here.

Third, according to the Government, if this statute is content-based because it singles out debt-collection speech, then so are statutes that regulate debt collection, like the Fair Debt Collection Practices Act. See 15 U. S. C. §1692 et seq.[4] That slippery-slope argument is unpersuasive in this case. As we explained in Sorrell, “the First Amendment does not prevent restrictions directed at commerce or conduct from imposing incidental burdens on speech.” 564 U. S., at 567. The law here, like the Vermont law in Sorrell, “does not simply have an effect on speech, but is directed at certain content and is aimed at particular speakers.” Ibid. The Government’s concern is understandable, but the courts have generally been able to distinguish impermissible content-based speech restrictions from traditional or ordinary economic regulation of commercial activity that imposes incidental burdens on speech. The issue before us concerns only robocalls to cell phones. Our decision today on that issue fits comfortably within existing First Amendment precedent. Our decision is not intended to expand existing First Amendment doctrine or to otherwise affect traditional or ordinary economic regulation of commercial activity.

In short, the robocall restriction with the government-debt exception is content-based. Under the Court’s precedents, a “law that is content based” is “subject to strict scrutiny.” Reed, 576 U. S., at 165. The Government concedes that it cannot satisfy strict scrutiny to justify the government-debt exception. We agree. The Government’s stated justification for the government-debt exception is collecting government debt. Although collecting government debt is no doubt a worthy goal, the Government concedes that it has not sufficiently justified the differentiation between government-debt collection speech and other important categories of robocall speech, such as political speech, charitable fundraising, issue advocacy, commercial advertising, and the like.[5]

III

Having concluded that the 2015 government-debt exception created an unconstitutional exception to the 1991 robocall restriction, we must decide whether to invalidate the entire 1991 robocall restriction, or instead to invalidate and sever the 2015 government-debt exception. Before we apply ordinary severability principles, we must address plaintiffs’ broader initial argument for why the entire 1991 robocall restriction is unconstitutional.

A

Plaintiffs correctly point out that the Government’s asserted interest for the 1991 robocall restriction is consumer privacy. But according to plaintiffs, Congress’s willingness to enact the government-debt exception in 2015 betrays a newfound lack of genuine congressional concern for consumer privacy. As plaintiffs phrase it, the 2015 exception “undermines the credibility” of the Government’s interest in consumer privacy. Tr. of Oral Arg. 38. Plaintiffs further contend that if Congress no longer has a genuine interest in consumer privacy, then the underlying 1991 robocall restriction is no longer justified (presumably under any level of heightened scrutiny) and is therefore now unconstitutional.

Plaintiffs’ argument is not without force, but we ultimately disagree with it. It is true that the Court has recognized that exceptions to a speech restriction “may diminish the credibility of the government’s rationale for restricting speech in the first place.” City of Ladue v. Gilleo, 512 U.S. 43, 52 (1994). But here, Congress’s addition of the government-debt exception in 2015 does not cause us to doubt the credibility of Congress’s continuing interest in protecting consumer privacy.

After all, the government-debt exception is only a slice of the overall robocall landscape. This is not a case where a restriction on speech is littered with exceptions that substantially negate the restriction. On the contrary, even after 2015, Congress has retained a very broad restriction on robocalls. The pre-1991 statistics on robocalls show that a variety of organizations collectively made a huge number of robocalls. And there is no reason to think that the incentives for those organizations—and many others—to make robocalls has diminished in any way since 1991. The continuing robocall restriction proscribes tens of millions of would-be robocalls that would otherwise occur every day. Congress’s continuing broad prohibition of robocalls amply demonstrates Congress’s continuing interest in consumer privacy.

The simple reality, as we assess the legislative developments, is that Congress has competing interests. Congress’s growing interest (as reflected in the 2015 amendment) in collecting government debt does not mean that Congress suddenly lacks a genuine interest in restricting robocalls. Plaintiffs seem to argue that Congress must be interested either in debt collection or in consumer privacy. But that is a false dichotomy, as we see it. As is not infrequently the case with either/or questions, the answer to this either/or question is “both.” Congress is interested both in collecting government debt and in protecting consumer privacy.

Therefore, we disagree with plaintiffs’ broader initial argument for holding the entire 1991 robocall restriction unconstitutional.

B

Plaintiffs next focus on ordinary severability principles. Applying those principles, the question before the Court is whether (i) to invalidate the entire 1991 robocall restriction, as plaintiffs want, or (ii) to invalidate just the 2015 government-debt exception and sever it from the remainder of the statute, as the Government wants.

We agree with the Government that we must invalidate the 2015 government-debt exception and sever that exception from the remainder of the statute. To explain why, we begin with general severability principles and then apply those principles to this case.

1

When enacting a law, Congress sometimes expressly addresses severability. For example, Congress may include a severability clause in the law, making clear that the unconstitutionality of one provision does not affect the rest of the law. See, e.g., 12 U. S. C. §5302; 15 U. S. C. §78gg; 47 U. S. C. §608. Alternatively, Congress may include a nonseverability clause, making clear that the unconstitutionality of one provision means the invalidity of some or all of the remainder of the law, to the extent specified in the text of the nonseverability clause. See, e.g., 4 U. S. C. §125; note following 42 U. S. C. §300aa–1; 94Stat. 1797.

When Congress includes an express severability or nonseverability clause in the relevant statute, the judicial inquiry is straightforward. At least absent extraordinary circumstances, the Court should adhere to the text of the severability or nonseverability clause. That is because a severability or nonseverability clause leaves no doubt about what the enacting Congress wanted if one provision of the law were later declared unconstitutional. A severability clause indicates “that Congress did not intend the validity of the statute in question to depend on the validity of the constitutionally offensive provision.” Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 686 (1987). And a nonseverability clause does the opposite.

On occasion, a party will nonetheless ask the Court to override the text of a severability or nonseverability clause on the ground that the text does not reflect Congress’s “actual intent” as to severability. That kind of argument may have carried some force back when courts paid less attention to statutory text as the definitive expression of Congress’s will. But courts today zero in on the precise statutory text and, as a result, courts hew closely to the text of severability or nonseverability clauses. See Seila Law LLC v. Consumer Financial Protection Bureau, ante, at 33 (plurality opinion); cf. Milner v. Department of Navy, 562 U.S. 562, 569–573 (2011).[6]

Of course, when enacting a law, Congress often does not include either a severability clause or a nonseverability clause.

In those cases, it is sometimes said that courts applying severability doctrine should search for other indicia of congressional intent. For example, some of the Court’s cases declare that courts should sever the offending provision unless “the statute created in its absence is legislation that Congress would not have enacted.” Alaska Airlines, 480 U. S., at 685. But experience shows that this formulation often leads to an analytical dead end. That is because courts are not well equipped to imaginatively reconstruct a prior Congress’s hypothetical intent. In other words, absent a severability or nonseverability clause, a court often cannot really know what the two Houses of Congress and the President from the time of original enactment of a law would have wanted if one provision of a law were later declared unconstitutional.

The Court’s cases have instead developed a strong presumption of severability. The Court presumes that an unconstitutional provision in a law is severable from the remainder of the law or statute. For example, in Free Enterprise Fund v. Public Company Accounting Oversight Bd., the Court set forth the “normal rule”: “Generally speaking, when confronting a constitutional flaw in a statute, we try to limit the solution to the problem, severing any problematic portions while leaving the remainder intact.” 561 U.S. 477, 508 (2010) (internal quotation marks omitted); see also Seila Law, ante, at 32 (same). In Regan v. Time, Inc., the plurality opinion likewise described a “presumption” in “favor of severability” and stated that the Court should “refrain from invalidating more of the statute than is necessary.” 468 U.S. 641, 652–653 (1984).

The Court’s power and preference to partially invalidate a statute in that fashion has been firmly established since Marbury v. Madison. There, the Court invalidated part of §13 of the Judiciary Act of 1789. 1 Cranch 137, 179–180 (1803). The Judiciary Act did not contain a severability clause. But the Court did not proceed to invalidate the entire Judiciary Act. As Chief Justice Marshall later explained, if any part of an Act is “unconstitutional, the provisions of that part may be disregarded while full effect will be given to such as are not repugnant to the constitution of the United States.” Bank of Hamilton v. Lessee of Dudley, 2 Pet. 492, 526 (1829); see also Dorchy v. Kansas, 264 U.S. 286, 289–290 (1924) (“A statute bad in part is not necessarily void in its entirety. Provisions within the legislative power may stand if separable from the bad”); Loeb v. Columbia Township Trustees, 179 U.S. 472, 490 (1900) (“one section of a statute may be repugnant to the Constitution without rendering the whole act void”).

From Marbury v. Madison to the present, apart from some isolated detours mostly in the late 1800s and early 1900s, the Court’s remedial preference after finding a provision of a federal law unconstitutional has been to salvage rather than destroy the rest of the law passed by Congress and signed by the President. The Court’s precedents reflect a decisive preference for surgical severance rather than wholesale destruction, even in the absence of a severability clause.

The Court’s presumption of severability supplies a workable solution—one that allows courts to avoid judicial policymaking or de facto judicial legislation in determining just how much of the remainder of a statute should be invalidated.[7] The presumption also reflects the confined role of the Judiciary in our system of separated powers—stated otherwise, the presumption manifests the Judiciary’s respect for Congress’s legislative role by keeping courts from unnecessarily disturbing a law apart from invalidating the provision that is unconstitutional. Furthermore, the presumption recognizes that plaintiffs who successfully challenge one provision of a law may lack standing to challenge other provisions of that law. See Murphy v. National Collegiate Athletic Assn., 584 U. S. ___, ___–___ (2018) (Thomas, J., concurring) (slip op., at 5–6).

Those and other considerations, taken together, have steered the Court to a presumption of severability. Applying the presumption, the Court invalidates and severs unconstitutional provisions from the remainder of the law rather than razing whole statutes or Acts of Congress. Put in common parlance, the tail (one unconstitutional provision) does not wag the dog (the rest of the codified statute or the Act as passed by Congress). Constitutional litigation is not a game of gotcha against Congress, where litigants can ride a discrete constitutional flaw in a statute to take down the whole, otherwise constitutional statute. If the rule were otherwise, the entire Judiciary Act of 1789 would be invalid as a consequence of Marbury v. Madison.[8]

Before severing a provision and leaving the remainder of a law intact, the Court must determine that the remainder of the statute is “capable of functioning independently” and thus would be “fully operative” as a law. Seila Law, ante, at 33; see Murphy, 584 U. S., at ___–___ (slip op., at 25–30). But it is fairly unusual for the remainder of a law not to be operative.[9]

2

We next apply those general severability principles to this case.

Recall how this statute came together. Passed by Congress and signed by President Franklin Roosevelt in 1934, the Communications Act is codified in Title 47 of the U. S. Code. The TCPA of 1991 amended the Communications Act by adding the robocall restriction, which is codified at §227(b)(1)(A)(iii) of Title 47. The Bipartisan Budget Act of 2015 then amended the Communications Act by adding the government-debt exception, which is codified along with the robocall restriction at §227(b)(1)(A)(iii) of Title 47.

Since 1934, the Communications Act has contained an express severability clause: “If any provision of this chapter or the application thereof to any person or circumstance is held invalid, the remainder of the chapter and the application of such provision to other persons or circumstances shall not be affected thereby.” 47 U. S. C. §608 (emphasis added). The “chapter” referred to in the severability clause is Chapter 5 of Title 47. And Chapter 5 in turn encompasses §151 to §700 of Title 47, and therefore covers §227 of Title 47, the provision with the robocall restriction and the government-debt exception.[10]

Enacted in 2015, the government-debt exception added an unconstitutional discriminatory exception to the robocall restriction. The text of the severability clause squarely covers the unconstitutional government-debt exception and requires that we sever it.

To get around the text of the severability clause, plaintiffs point out that the Communications Act’s severability clause was enacted in 1934, long before the TCPA’s 1991 robocall restriction and the 2015 government-debt exception. But a severability clause must be interpreted according to its terms, regardless of when Congress enacted it. See n. 6, supra.

Even if the severability clause did not apply to the government-debt provision at issue in this case (or even if there were no severability clause in the Communications Act), we would apply the presumption of severability as described and applied in cases such as Free Enterprise Fund. And under that presumption, we likewise would sever the 2015 government-debt exception, the constitutionally offending provision.

With the government-debt exception severed, the remainder of the law is capable of functioning independently and thus would be fully operative as a law. Indeed, the remainder of the robocall restriction did function independently and fully operate as a law for 20-plus years before the government-debt exception was added in 2015.

The Court’s precedents further support severing the 2015 government-debt exception. The Court has long applied severability principles in cases like this one, where Congress added an unconstitutional amendment to a prior law. In those cases, the Court has treated the original, pre-amendment statute as the “valid expression of the legislative intent.” Frost v. Corporation Comm’n of Okla., 278 U.S. 515, 526–527 (1929). The Court has severed the “exception introduced by amendment,” so that “the original law stands without the amendatory exception.” Truax v. Corrigan, 257 U.S. 312, 342 (1921).

For example, in Eberle v. Michigan, the Court held that “discriminatory wine-and-cider amendments” added in 1899 and 1903 were severable from the underlying 1889 state law generally prohibiting the manufacture of alcohol. 232 U.S. 700, 704–705 (1914). In Truax, the Court ruled that a 1913 amendment prohibiting Arizona courts from issuing injunctions in labor disputes was invalid and severable from the underlying 1901 law authorizing Arizona courts to issue injunctions generally. 257 U. S., at 341–342. In Frost, the Court concluded that a 1925 amendment exempting certain corporations from making a showing of “public necessity” in order to obtain a cotton gin license was invalid and severable from the 1915 law that required that showing. 278 U. S., at 525–528. Echoing Marbury, the Court in Frost explained that an unconstitutional statutory amendment “is a nullity” and “void” when enacted, and for that reason has no effect on the original statute. 278 U. S., at 526–527 (internal quotation marks omitted).[11]

Similarly, in 1932, Congress enacted the Federal Kidnaping Act, and then in 1934, added a death penalty provision to the Act. The death penalty provision was later declared unconstitutional by this Court. In considering severability, the Court stated that the “law as originally enacted in 1932 contained no capital punishment provision.” United States v. Jackson, 390 U.S. 570, 586 (1968). And when Congress amended the Act in 1934 to add the death penalty, “the statute was left substantially unchanged in every other respect.” Id., at 587–588. The Court found it “difficult to imagine a more compelling case for severability.” Id., at 589. So too here.

In sum, the text of the Communications Act’s severability clause requires that the Court sever the 2015 government-debt exception from the remainder of the statute. And even if the text of the severability clause did not apply here, the presumption of severability would require that the Court sever the 2015 government-debt exception from the remainder of the statute.

3

One final severability wrinkle remains. This is an equal-treatment case, and equal-treatment cases can sometimes pose complicated severability questions.

The “ First Amendment is a kind of Equal Protection Clause for ideas.” Williams-Yulee v. Florida Bar, 575 U.S. 433, 470 (2015) (Scalia, J., dissenting). And Congress violated that First Amendment equal-treatment principle in this case by favoring debt-collection robocalls and discriminating against political and other robocalls.

When the constitutional violation is unequal treatment, as it is here, a court theoretically can cure that unequal treatment either by extending the benefits or burdens to the exempted class, or by nullifying the benefits or burdens for all. See, e.g., Heckler v. Mathews, 465 U.S. 728, 740 (1984). Here, for example, the Government would prefer to cure the unequal treatment by extending the robocall restriction and thereby proscribing nearly all robocalls to cell phones. By contrast, plaintiffs want to cure the unequal treatment by nullifying the robocall restriction and thereby allowing all robocalls to cell phones.

When, as here, the Court confronts an equal-treatment constitutional violation, the Court generally applies the same commonsense severability principles described above. If the statute contains a severability clause, the Court typically severs the discriminatory exception or classification, and thereby extends the relevant statutory benefits or burdens to those previously exempted, rather than nullifying the benefits or burdens for all. In light of the presumption of severability, the Court generally does the same even in the absence of a severability clause. The Court’s precedents reflect that preference for extension rather than nullification. See, e.g., Sessions v. Morales-Santana, 582 U. S. ___, ___ (2017) (slip op., at 25); Califano v. Westcott, 443 U.S. 76, 89–91 (1979); Califano v. Goldfarb, 430 U.S. 199, 202–204, 213–217 (1977) (plurality opinion); Jimenez v. Weinberger, 417 U.S. 628, 637–638 (1974); Department of Agriculture v. Moreno, 413 U.S. 528, 529, 537–538 (1973); Frontiero v. Richardson, 411 U.S. 677, 678–679, 690–691 (1973) (plurality opinion); Welsh v. United States, 398 U.S. 333, 361–367 (1970) (Harlan, J., concurring in result).

To be sure, some equal-treatment cases can raise complex questions about whether it is appropriate to extend benefits or burdens, rather than nullifying the benefits or burdens. See, e.g., Morales-Santana, 582 U. S. ___. For example, there can be due process, fair notice, or other independent constitutional barriers to extension of benefits or burdens. Cf. Miller v. Albright, 523 U.S. 420, 458–459 (1998) (Scalia, J., concurring in judgment); see generally Ginsburg, Some Thoughts on Judicial Authority to Repair Unconstitutional Legislation, 28 Clev. St. L. Rev. 301 (1979). There also can be knotty questions about what is the exception and what is the rule. But here, we need not tackle all of the possible hypothetical applications of severability doctrine in equal-treatment cases. The government-debt exception is a relatively narrow exception to the broad robocall restriction, and severing the government-debt exception does not raise any other constitutional problems.

Plaintiffs insist, however, that a First Amendment equal-treatment case is different. According to plaintiffs, a court should not cure “a First Amendment violation by outlawing more speech.” Brief for Respondents 34. The implicit premise of that argument is that extending the robocall restriction to debt-collection robocalls would be unconstitutional. But that is wrong. A generally applicable robocall restriction would be permissible under the First Amendment. Extending the robocall restriction to those robocalls raises no First Amendment problem. So the First Amendment does not tell us which way to cure the unequal treatment in this case. Therefore, we apply traditional severability principles. And as we have explained, severing the 2015 government-debt exception cures the unequal treatment and constitutes the proper result under the Court’s traditional severability principles. In short, the correct result in this case is to sever the 2015 government-debt exception and leave in place the longstanding robocall restriction.[12]

4

Justice Gorsuch’s well-stated separate opinion makes a number of important points that warrant this respectful response.

Justice Gorsuch suggests that our decision provides “no relief” to plaintiffs. Post, at 6. We disagree. Plaintiffs want to be able to make political robocalls to cell phones, and they have not received that relief. But the First Amendment complaint at the heart of their suit was unequal treatment. Invalidating and severing the government-debt exception fully addresses that First Amendment injury.[13] Justice Gorsuch further suggests that plaintiffs may lack standing to challenge the government-debt exception, because that exception merely favors others. See ibid. But the Court has squarely held that a plaintiff who suffers unequal treatment has standing to challenge a discriminatory exception that favors others. See Heckler v. Mathews, 465 U. S., at 737–740 (a plaintiff who suffers unequal treatment has standing to seek “withdrawal of benefits from the favored class”); see also Northeastern Fla. Chapter, Associated Gen. Contractors of America v. Jacksonville, 508 U.S. 656, 666 (1993) (“The ‘injury in fact’ in an equal protection case of this variety is the denial of equal treatment resulting from the imposition of the barrier, not the ultimate inability to obtain the benefit”).

Justice Gorsuch also objects that our decision today “harms strangers to this suit” by eliminating favorable treatment for debt collectors. Post, at 6. But that is necessarily true in many cases where a court cures unequal treatment by, for example, extending a burden or nullifying a benefit. See, e.g., Morales-Santana, 582 U. S., at ___ (slip op., at 28) (curing unequal treatment of children born to unwed U. S.-citizen fathers by extending a burden to children of unwed U. S.-citizen mothers); Orr v. Orr, 374 So. 2d 895, 896–897 (Ala. Civ. App. 1979) (extending alimony obligations to women after a male plaintiff successfully challenged Alabama’s discriminatory alimony statute in this Court).

Moreover, Justice Gorsuch’s approach to this case would not solve the problem of harming strangers to this suit; it would just create a different and much bigger problem. His proposed remedy of injunctive relief, plus stare decisis, would in effect allow all robocalls to cell phones—notwithstanding Congress’s decisive choice to prohibit most robocalls to cell phones. That is not a judicially modest approach but is more of a wolf in sheep’s clothing. That approach would disrespect the democratic process, through which the people’s representatives have made crystal clear that robocalls must be restricted. Justice Gorsuch’s remedy would end up harming a different and far larger set of strangers to this suit—the tens of millions of consumers who would be bombarded every day with nonstop robocalls notwithstanding Congress’s clear prohibition of those robocalls.

Justice Gorsuch suggests more broadly that severability doctrine may need to be reconsidered. But when and how? As the saying goes, John Marshall is not walking through that door. And this Court, in this and other recent decisions, has clarified and refined severability doctrine by emphasizing firm adherence to the text of severability clauses, and underscoring the strong presumption of severability. The doctrine as so refined is constitutionally well-rooted, see, e.g., Marbury v. Madison, 1 Cranch 137 (Marshall, C. J.), and can be predictably applied. True, there is no magic solution to severability that solves every conundrum, especially in equal-treatment cases, but the Court’s current approach as reflected in recent cases such as Free Enterprise Fund and Seila Law is constitutional, stable, predictable, and commonsensical.

* �ă* �ă*

In 1991, Congress enacted a general restriction on robocalls to cell phones. In 2015, Congress carved out an exception that allowed robocalls made to collect government debt. In doing so, Congress favored debt-collection speech over plaintiffs’ political speech. We hold that the 2015 government-debt exception added an unconstitutional exception to the law. We cure that constitutional violation by invalidating the 2015 government-debt exception and severing it from the remainder of the statute. The judgment of the U. S. Court of Appeals for the Fourth Circuit is affirmed.

It is so ordered.

Notes

1. The robocall restriction, as implemented by the Federal Communications Commission, bars both automated voice calls and automated text messages. See In re Rules and Regulations Implementing the Telephone Consumer Protection Act of 1991, 18 FCC Rcd. 14014, 14115 (2003). The robocall restriction applies to “persons,” which does not include the Government itself. See 47 U. S. C. §153(39). Congress has also authorized the FCC to promulgate regulatory exceptions to the robocall restriction. See §227(b)(2)(C). The FCC has authorized various exceptions over the years, such as exceptions for package-delivery notifications and certain healthcare-related calls. In this case, plaintiffs do not separately challenge the validity of the FCC’s regulatory exceptions.

2. After the 2015 amendment, §227(b)(1) now provides: “It shall be unlawful for any person within the United States, or any person outside the United States if the recipient is within the United States— (A) to make any call (other than a call made for emergency purposes or made with the prior express consent of the called party) using any automatic telephone dialing system or an artificial or prerecorded voice— . . . . (iii) to any telephone number assigned to a paging service, cellular telephone service, specialized mobile radio service, or other radio common carrier service, or any service for which the called party is charged for the call, unless such call is made solely to collect a debt owed to or guaranteed by the United States.” (Emphasis added.)

3. Plaintiffs have not challenged the TCPA’s separate restriction on robocalls to home phones. See 47 U. S. C. §227(b)(1)(B).

4. This opinion uses the term “debt-collection speech” and “debt-collection robocalls” as shorthand for government-debt collection speech and robocalls.

5.  In his scholarly separate opinion, Justice Breyer explains how he would apply freedom of speech principles. But the Court’s longstanding precedents, which we carefully follow here, have not adopted that approach. In essence, therefore, Justice Breyer argues for overruling several of the Court’s First Amendment cases, including the recent 2015 decision in Reed v. Town of Gilbert, 576 U.S. 155 (2015). Before overruling precedent, the Court usually requires that a party ask for overruling, or at least obtains briefing on the overruling question, and then the Court carefully evaluates the traditional stare decisis factors. Here, no party has asked for overruling, and Justice Breyer’s opinion does not analyze the usual stare decisis factors. Justice Breyer’s opinion therefore discounts both the Court’s precedent and the Court’s precedent on precedent.

6. When Congress enacts a law with a severability clause and later adds new provisions to that statute, the severability clause applies to those new provisions to the extent dictated by the text of the severability clause. Likewise, when Congress has not included a severability clause in initial legislation, Congress can subsequently enact a severability clause that applies to the existing statute to the extent dictated by the text of the later-added severability clause. In both scenarios, the text of the severability clause remains central to the severability inquiry.

7. If courts had broad license to invalidate more than just the offending provision, a reviewing court would have to consider what other provisions to invalidate: the whole section, the chapter, the statute, the public law, or something else altogether. Courts would be largely at sea in making that determination, and usually could not do it in a principled way. Here, for example, would a court invalidate all or part of the Bipartisan Budget Act of 2015 rather than all or part of the 1991 TCPA? After all, that 2015 Bipartisan Budget Act, not the 1991 TCPA, added the constitutionally problematic government-debt exception. That is the kind of free-wheeling policy question that the Court’s presumption of severability avoids.

8. The term “invalidate” is a common judicial shorthand when the Court holds that a particular provision is unlawful and therefore may not be enforced against a plaintiff. To be clear, however, when it “invalidates” a law as unconstitutional, the Court of course does not formally repeal the law from the U. S. Code or the Statutes at Large. Instead, in Chief Justice Marshall’s words, the Court recognizes that the Constitution is a “superior, paramount law,” and that “a legislative act contrary to the constitution is not law” at all. Marbury v. Madison, 1 Cranch 137, 177 (1803). The Court’s authority on this front “amounts to little more than the negative power to disregard an unconstitutional enactment.” Massachusetts v. Mellon, 262 U.S. 447, 488 (1923). Justice Thomas’s thoughtful approach to severability as outlined in Murphy v. National Collegiate Athletic Assn., 584 U. S. ___, ___–___ (2018) (slip op., at 2–6), and Seila Law LLC v. Consumer Financial Protection Bureau, ante, at 14–24, (joined by Justice Gorsuch in the latter)would simply enjoin enforcement of a law as applied to the particular plaintiffs in a case. Under either the Court’s approach or Justice Thomas’s approach, an offending provision formally remains on the statute books (at least unless Congress also formally repeals it). Under either approach, the formal remedyafforded to the plaintiff is an injunction, declaration, or damages. One difference between the two approaches is this: Under the Court’s approach, a provision is declared invalid and cannot be lawfully enforced against others. Under Justice Thomas’s approach, the Court’s ruling that a provision cannot be enforced against the plaintiff, plus executive respect in its enforcement policies for controlling decisional law, plus vertical and horizontal stare decisis in the courts, will mean that the provision will not and cannot be lawfully enforced against others. The Court and Justice Thomas take different analytical paths, but in many cases, the different paths lead to the same place.

9. On occasion, of course, it may be that a particular surrounding or connected provision is not operative in the absence of the unconstitutional provision, even though the rest of the law would be operative. That scenario may require severance of somewhat more than just the offending provision, albeit not of the entire law. Courts address that scenario as it arises.

10. A codifier’s note explains a change in wording from the original Public Law: “This chapter, referred to in text, was in the original ‘this Act’, meaning act June 19, 1934, ch. 652, 48Stat. 1064, known as the Communications Act of 1934, which is classified principally to this chapter.” Note following 47 U. S. C. §608.

11. The cases cited in the text above are pre-Erie decisions involving the constitutionality of state laws. See Erie R. Co. v. Tompkins, 304 U.S. 64 (1938). In that era, the Court often treated severability of state laws and federal laws in the same general way. In the post-Erie era, severability of state laws can potentially pose different questions than severability of federal laws. We need not address post-Erie severability of state laws. See, e.g., Ayotte v. Planned Parenthood of Northern New Eng., 546 U.S. 320, 328–331 (2006); Leavitt v. Jane L., 518 U.S. 137, 139 (1996) (per curiam) (“Severability is of course a matter of state law”).

12. As the Government acknowledges, although our decision means the end of the government-debt exception, no one should be penalized or held liable for making robocalls to collect government debt after the effective date of the 2015 government-debt exception and before the entry of final judgment by the District Court on remand in this case, or such date that the lower courts determine is appropriate. See Reply Brief 24. On the other side of the ledger, our decision today does not negate the liability of parties who made robocalls covered by the robocall restriction.

13. Plaintiffs suggest that parties will not have incentive to sue if the cure for challenging an unconstitutional exception to a speech restriction is to eliminate the exception and extend the restriction. But many individuals and organizations often have incentive to challenge unequal treatment of speech, especially when a competitor is regulated less heavily.