Federal Communications Commission v. Fox Television Stations, Inc.

556 U.S. 502 | SCOTUS | 2009

Lead Opinion

*505Justice Scalia

delivered the opinion of the Court, except as to Part III-E.

Federal law prohibits the broadcasting of “any . . . indecent . . . language,” 18 U. S. C. § 1464, which includes expletives referring to sexual or excretory activity or organs, see FCC v. Pacifica Foundation, 438 U. S. 726 (1978). This case concerns the adequacy of the Federal Communications Commission’s explanation of its decision that this sometimes forbids the broadcasting of indecent expletives even when the offensive words are not repeated.

I. Statutory and Regulatory Background

The Communications Act of 1934,48 Stat. 1064,47 U. S. C. § 151 et seq. (2000 ed. and Supp. V), established a system of *506limited-term broadcast licenses subject to various “conditions” designed “to maintain the control of the United States over all the channels of radio transmission,” § 301 (2000 ed.). Almost 28 years ago we said that “[a] licensed broadcaster is granted the free and exclusive use of a limited and valuable part of the public domain; when he accepts that franchise it is burdened by enforceable public obligations.” CBS, Inc. v. FCC, 453 U. S. 367, 395 (1981) (internal quotation marks omitted).

One of the burdens that licensees shoulder is the indecency ban — the statutory proscription against “utter[ingj any obscene, indecent, or profane language by means of radio communication,” 18 U. S. C. § 1464 — which Congress has instructed the Commission to enforce between the hours of 6 a.m. and 10 p.m. Public Telecommunications Act of 1992, § 16(a), 106 Stat. 954, note following 47 U. S. C. § 303.1 Congress has given the Commission various means of enforcing the indecency ban, including civil fines, see § 503(b)(1), and license revocations or the denial of license renewals, see §§ 309(k), 312(a)(6).

The Commission first invoked the statutory ban on indecent broadcasts in 1975, declaring a daytime broadcast of George Carlin's “Filthy Words” monologue actionably indecent. In re Citizen’s Complaint Against Pacifica Foundation Station WBAI (FM), 56 F. C. C. 2d 94. At that time, the Commission announced the definition of indecent speech that it uses to this day, prohibiting “language that describes, *507in terms patently offensive as measured by contemporary community standards for the broadcast medium, sexual or excretory activities and organs, at times of the day when there is a reasonable risk that children may be in the audience.” Id., at 98.

In FCC v. Pacifica Foundation, supra, we upheld the Commission’s order against statutory and constitutional challenge. We rejected the broadcasters’ argument that the statutory proscription applied only to speech appealing to the prurient interest, noting that “the normal definition of ‘indecent’ merely refers to nonconformance with accepted standards of morality.” Id., at 740. And we held that the First Amendment allowed Carlin’s monologue to be banned in light of the “uniquely pervasive presence” of the medium and the fact that broadcast programming is “uniquely accessible to children.” Id., at 748-749.

In the ensuing years, the Commission took a cautious, but gradually expanding, approach to enforcing the statutory prohibition against indecent broadcasts. Shortly after Pacifica, 438 U. S. 726, the Commission expressed its “intension] strictly to observe the narrowness of the Pacifica holding,” which “relied in part on the repetitive occurrence of the ‘indecent’ words” contained in Carlin’s monologue. In re Application of WGBH Educ. Foundation, 69 F. C. C. 2d 1250, 1254, ¶ 10 (1978). When the full Commission next considered its indecency standard, however, it repudiated the view that its enforcement power was limited to “deliberate, repetitive use of the seven words actually contained in the George Carlin monologue.” In re Pacifica Foundation, Inc., 2 FCC Red. 2698, 2699, ¶ 12 (1987). The Commission determined that such a “highly restricted enforcement standard . . . was unduly narrow as a matter of law and inconsistent with [the Commission’s] enforcement responsibilities under Section 1464.” In re Infinity Broadcasting Corp. of Pa., 3 FCC Red. 930, ¶ 5 (1987). The Court of Appeals for the District of Columbia Circuit upheld this *508expanded enforcement standard against constitutional and Administrative Procedure Act challenge. See Action for Children’s Television v. FCC, 852 F. 2d 1332 (1988) (R. Ginsburg, J.), superseded in part by Action for Children’s Television v. FCC, 58 F. 3d 654 (1995) (en banc).

Although the Commission had expanded its enforcement beyond the “repetitive use of specific words or phrases,” it preserved a distinction between literal and nonliteral (or “expletive”) uses of evocative language. In re Pacifica Foundation, Inc., 2 FCC Red., at 2699, ¶13. The Commission explained that each literal “description or depiction of sexual or excretory functions must be examined in context to determine whether it is patently offensive,” but that “deliberate and repetitive use ... is a requisite to a finding of indecency” when a complaint focuses solely on the use of nonliteral expletives. Ibid.

Over a decade later, the Commission emphasized that the “full context” in which particular materials appear is “critically important,” but that a few “principal” factors guide the inquiry, such as the “explicitness or graphic nature” of the material, the extent to which the material “dwells on or repeats” the offensive material, and the extent to which the material was presented to “pander,” to “titillate,” or to “shock.” In re Industry Guidance on Commission’s Case Law Interpreting 18 U. S. C. §1464 and Enforcement Policies Regarding Broadcast Indecency, 16 FCC Rcd. 7999, 8002, ¶ 9, 8003, ¶ 10 (2001) (emphasis deleted). “No single factor,” the Commission said, “generally provides the basis for an indecency finding,” but “where sexual or excretory references have been made once or have been passing or fleeting in nature, this characteristic has tended to weigh against a finding of indecency.” Id., at 8003, ¶ 10,8008, ¶ 17.

In 2004, the Commission took one step further by declaring for the first time that a nonliteral (expletive) use of the F- and S-Words could be actionably indecent, even when the word is used only once. The first order to this effect dealt *509with an NBC broadcast of the Golden Globe Awards, in which the performer Bono commented, “ ‘[T]his is really, really, f***ing brilliant.’ ” In re Complaints Against Various Broadcast Licensees Regarding Their Airing of “Golden Globe Awards” Program, 19 FCC Red. 4975, 4976, n. 4 (2004) (Golden Globes Order). Although the Commission had received numerous complaints directed at the broadcast, its enforcement bureau had concluded that the material was not indecent because “Bono did not describe, in context, sexual or excretory organs or activities and ... the utterance was fleeting and isolated.” Id., at 4975-4976, ¶ 3. The full Commission reviewed and reversed the staff ruling.

The Commission first declared that Bono’s use of the F-Word fell within its indecency definition, even though the word was used as an intensifier rather than a literal descriptor. “[G]iven the core meaning of the ‘F-Word,’” it said, “any use of that word . . . inherently has a sexual connotation.” Id., at 4978, ¶8. The Commission determined, moreover, that the broadcast was “patently offensive” because the F-Word “is one of the most vulgar, graphic and explicit descriptions of sexual activity in the English language,” because “[i]ts use invariably invokes a coarse sexual image,” and because Bono’s use of the word was entirely “shocking and gratuitous.” Id., at 4979, ¶ 9.

The Commission observed that categorically exempting such language from enforcement actions would “likely lead to more widespread use.” Ibid. Commission action was necessary to “safeguard the well-being of the nation’s children from the most objectionable, most offensive language.” Ibid. The order noted that technological advances have made it far easier to delete (“bleep out”) a “single and gratuitous use of a vulgar expletive,” without adulterating the content of a broadcast. Id., at 4980, ¶ 11.

The order acknowledged that “prior Commission and staff action [has] indicated that isolated or fleeting broadcasts of the ‘F-Word’ . . . are not indecent or would not be acted *510upon.” It explicitly ruled that “any such interpretation is no longer good law.” Ibid., ¶ 12. It “clarified]... that the mere fact that specific words or phrases are not sustained or repeated does not mandate a finding that material that is otherwise patently offensive to the broadcast medium is not indecent.” Ibid. Because, however, “existing precedent would have permitted this broadcast,” the Commission determined that “NBC and its affiliates necessarily did not have the requisite notice to justify a penalty.” Id., at 4981-4982, ¶ 15.

II. The Present Case

This case concerns utterances in two live broadcasts aired by Fox Television Stations, Inc., and its affiliates prior to the Commission’s Golden Globes Order. The first occurred during the 2002 Billboard Music Awards, when the singer Cher exclaimed, “I’ve also had critics for the last 40 years saying that I was on my way out every year. Right. So f*** 'em.” Brief for Petitioners 9. The second involved a segment of the 2008 Billboard Music Awards, during the presentation of an award by Nicole Richie and Paris Hilton, principals in a Fox television series called “The Simple Life.” Ms. Hilton began their interchange by reminding Ms. Richie to “watch the bad language,” but Ms. Richie proceeded to ask the audience, “Why do they even call it ‘The Simple Life?’ Have you ever tried to get cow s*** out of a Prada purse? It’s not so f***ing simple.” Id., at 9-10. Following each of these broadcasts, the Commission received numerous complaints from parents whose children were exposed to the language.

On March 15, 2006, the Commission released “Notices of Apparent Liability” for a number of broadcasts that the Commission deemed actionably indecent, including the two described above. In re Complaints Regarding Various Television Broadcasts Between Feb. 2,2002 and Mar. 8,2005, 21 FCC Red. 2664 (2006). Multiple parties petitioned the Court of Appeals for the Second Circuit for judicial review of *511the order, asserting a variety of constitutional and statutory challenges. Since the order had declined to impose sanctions, the Commission had not previously given the broadcasters an opportunity to respond to the indecency charges. It therefore requested and obtained from the Court of Appeals a voluntary remand so that the parties could air their objections. 489 F. 3d 444, 453 (2007). The Commission’s order on remand upheld the indecency findings for the broadcasts described above. See In re Complaints Regarding Various Television Broadcasts Between Feb. 2, 2002, and Mar. 8, 2005, 21 FCC Red. 13299 (2006) (Remand Order).

The order first explained that both broadcasts fell comfortably within the subject-matter scope of the Commission’s indecency test because the 2003 broadcast involved a literal description of excrement and both broadcasts invoked the “F-Word,” which inherently has a sexual connotation. Id., at 13304, ¶ 16,13323, ¶ 58. The order next determined that the broadcasts were patently offensive under community standards for the medium. Both broadcasts, it noted, involved entirely gratuitous uses of “one of the most vulgar, graphic, and explicit words for sexual activity in the English language.” Id., at 13305, ¶ 17, 13324, ¶ 59. It found Ms. Richie’s use of the “F-Word” and her “explicit description of the handling of excrement” to be “vulgar and shocking,” as well as to constitute “pandering,” after Ms. Hilton had playfully warned her to “‘watch the bad language.’” Id., at 13305, ¶ 17. And it found Cher’s statement patently offensive in part because she metaphorically suggested a sexual act as a means of expressing hostility to her critics. Id., at 13324, ¶60. The order relied upon the “‘critically important’ ” context of the utterances, id., at 13304, ¶ 15, noting that they were aired during prime-time awards shows “designed to draw a large nationwide audience that could be expected to include many children interested in seeing their favorite music stars,” id., at 13305, ¶ 18,13324, ¶ 59. Indeed, *512approximately 2.5 million minors witnessed each of the broadcasts. Id., at 13306, ¶ 18,13326, ¶ 65.

The order asserted that both broadcasts under review would have been actionably indecent under the staff rulings and Commission dicta in effect prior to the Golden Globes Order — the 2003 broadcast because it involved a literal description of excrement, rather than a mere expletive, because it used more than one offensive word, and because it was planned, 21 FCC Red., at 13307, ¶ 22; and the 2002 broadcast because Cher used the F-Word not as a mere intensifier, but as a description of the sexual act to express hostility to her critics, id., at 13324, ¶ 60. The order stated, however, that the pre-Golden Globes regime of immunity for isolated indecent expletives rested only upon staff rulings and Commission dicta, and that the Commission itself had never held “that the isolated use of an expletive . . . was not indecent or could not be indecent,” 21 FCC Red., at 13307, ¶ 21. In any event, the order made clear, the Golden Globes Order eliminated any doubt that fleeting expletives could be action-ably indecent, 21 FCC Red., at 13308, ¶ 23, 13325, ¶ 61, and the Commission disavowed the bureau-level decisions and its own dicta that had said otherwise, id., at 13306-13307, ¶¶ 20, 21. Under the new policy, a lack of repetition “weights] against a finding of indecency,” id., at 13325, ¶ 61, but is not a safe harbor.

The order explained that the Commission’s prior “strict dichotomy between ‘expletives’ and ‘descriptions or depictions of sexual or excretory functions’ is artificial and does not make sense in light of the fact that an ‘expletive’s’ power to offend derives from its sexual or excretory meaning.” Id., at 13308, ¶ 23. In the Commission’s view, “granting an automatic exemption for ‘isolated or fleeting’ expletives unfairly forces viewers (including children)” to take “ ‘the first blow’ ” and would allow broadcasters “to air expletives at all hours of a day so long as they did so one at a time.” Id., at 13309, ¶ 25. Although the Commission determined that Fox *513encouraged the offensive language by using suggestive scripting in the 2003 broadcast, and unreasonably failed to take adequate precautions in both broadcasts, id., at 13311-13314, ¶¶ 31-37, the order again declined to impose any forfeiture or other sanction for either of the broadcasts, id., at 13321, ¶ 53, 13326, ¶66.

Fox returned to the Second Circuit for review of the Remand Order, and various intervenors including CBS, NBC, and ABC joined the action. The Court of Appeals reversed the agency’s orders, finding the Commission’s reasoning inadequate under the Administrative Procedure Act. 489 F. 3d 444. The majority was “skeptical that the Commission [could] provide a reasoned explanation for its ‘fleeting expletive’ regime that would pass constitutional muster,” but it declined to reach the constitutional question. Id., at 462. Judge Leval dissented, id., at 467. We granted certiorari, 552 U. S. 1255 (2008).

III. Analysis

A. Governing Principles

The Administrative Procedure Act, 5 U. S. C. §551 et seq., which sets forth the full extent of judicial authority to review executive agency action for procedural correctness, see Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U. S. 519, 545-549 (1978), permits (insofar as relevant here) the setting aside of agency action that is “arbitrary” or “capricious,” 5 U. S. C. § 706(2)(A). Under what we have called this “narrow” standard of review, we insist that an agency “examine the relevant data and articulate a satisfactory explanation for its action.” Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 43 (1983). We have made clear, however, that “a court is not to substitute its judgment for that of the agency,” ibid., and should “uphold a decision of less than ideal clarity if the agency’s path may reasonably *514be discerned,” Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., 419 U. S. 281, 286 (1974).

In overturning the Commission’s judgment, the Court of Appeals here relied in part on Circuit precedent requiring a more substantial explanation for agency action that changes prior policy. The Second Circuit has interpreted the Administrative Procedure Act and our opinion in State Farm as requiring agencies to make clear “ ‘why the original reasons for adopting the [displaced] rule or policy are no longer dis-positive’ ” as well as “ ‘why the new rule effectuates the statute as well as or better than the old rule.’” 489 F. 3d, at 456-457 (quoting New York Council, Assn. of Civilian Technicians v. FLRA, 757 F. 2d 502, 508 (CA2 1985); emphasis deleted). The Court of Appeals for the District of Columbia Circuit has similarly indicated that a court’s standard of review is “heightened somewhat” when an agency reverses course. NAACP v. FCC, 682 F. 2d 993, 998 (1982).

We find no basis in the Administrative Procedure Act or in our opinions for a requirement that all agency change be subjected to more searching review. The Act mentions no such heightened standard. And our opinion in State Farm neither held nor implied that every agency action representing a policy change must be justified by reasons more substantial than those required to adopt a policy in the first instance. That case, which involved the rescission of a prior regulation, said only that such action requires “a reasoned analysis for the change beyond that which may be required when an agency does not act in the first instance.” 463 U. S., at 42 (emphasis added).2 Treating failures to act and *515rescissions of prior action differently for purposes of the standard of review makes good sense, and has basis in the text of the statute, which likewise treats the two separately. It instructs a reviewing court to “compel agency action unlawfully withheld or unreasonably delayed,” 5 U. S. C. § 706(1), and to “hold unlawful and set aside agency action, findings, and conclusions found to be [among other things]... arbitrary [or] capricious,” § 706(2)(A). The statute makes no distinction, however, between initial agency action and subsequent agency action undoing or revising that action.

To be sure, the requirement that an agency provide reasoned explanation for its action would ordinarily demand that it display awareness that it is changing position. An agency may not, for example, depart from a prior policy sub silentio or simply disregard rules that are still on the books. See United States v. Nixon, 418 U. S. 683, 696 (1974). And of course the agency must show that there are good reasons for the new policy. But it need not demonstrate to a court’s satisfaction that the reasons for the new policy are better than the reasons for the old one; it suffices that the new policy is permissible under the statute, that there are good reasons for it, and that the agency believes it to be better, which the conscious change of course adequately indicates. This means that the agency need not always provide a more detailed justification than what would suffice for a new policy created on a blank slate. Sometimes it must — when, for example, its new policy rests upon factual findings that contradict those which underlay its prior policy; or when its prior policy has engendered serious reliance interests that must be taken into account. Smiley v. Citibank (South Dakota), N. A., 517 U. S. 735, 742 (1996). It would be arbitrary or capricious to ignore such matters. In such cases it is not *516that farther justification is demanded by the mere fact of policy change; but that a reasoned explanation is needed for disregarding facts and circumstances that underlay or were engendered by the prior policy.

In this appeal from the Second Circuit’s setting aside of Commission action for failure to comply with a procedural requirement of the Administrative Procedure Act, the broadcasters’ arguments have repeatedly referred to the First Amendment. If they mean to invite us to apply a more stringent arbitrary-and-eapricious review to agency actions that implicate constitutional liberties, we reject the invitation. The so-called canon of constitutional avoidance is an interpretive tool, counseling that ambiguous statutory language be construed to avoid serious constitutional doubts. See Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Constr. Trades Council, 485 U. S. 568, 575 (1988). We know of no precedent for applying it to limit the scope of authorized executive action. In the same section authorizing courts to set aside “arbitrary [or] capricious” agency action, the Administrative Procedure Act separately provides for setting aside agency action that is “unlawful,” 5 U. S. C. § 706(2)(A), which of course includes unconstitutional action. We think that is the only context in which constitutionality bears upon judicial review of authorized agency action. If the Commission’s action here was not arbitrary or capricious in the ordinary sense, it satisfies the Administrative Procedure Act’s “arbitrary [or] capricious” standard; its lawfulness under the Constitution is a separate question to be addressed in a constitutional challenge.3

*517B. Application to This Case

Judged under the above described standards, the Commission’s new enforcement policy and its order finding the broadcasts actionably indecent were neither arbitrary nor capricious. First, the Commission forthrightly acknowledged that its recent actions have broken new ground, taking account of inconsistent “prior Commission and staff action” and explicitly disavowing them as “no longer good law.” Golden Globes Order, 19 FCC Red., at 4980, ¶ 12. To be sure, the (superfluous) explanation in its Remand Order of why the Cher broadcast would even have violated its earlier policy may not be entirely convincing. But that unnecessary detour is irrelevant. There is no doubt that the Commission knew it was making a change. That is why it declined to assess penalties; and it relied on the Golden Globes Order as removing any lingering doubt. Remand Order, 21 FCC Red., at 13308, ¶23, 13325, ¶ 61.

Moreover, the agency’s reasons for expanding the scope of its enforcement activity were entirely rational. It was certainly reasonable to determine that it made no sense to distinguish between literal and nonliteral uses of offensive words, requiring repetitive use to render only the latter indecent. As the Commission said with regard to expletive use of the F-Word, “the word’s power to insult and offend derives from its sexual meaning.” Id., at 13323, ¶ 58. And the Commission’s decision to look at the patent offensiveness of even isolated uses of sexual and excretory words fits with the context-based approach we sanctioned in Pacifica, 438 *518U. S., at 750. Even isolated utterances can be made in “pander[ing,] . . . vulgar and shocking” manners, Remand Order, 21 FCC Red., at 13305, ¶ 17, and can constitute harmful '“first blow[s]’” to children, id., at 13309, ¶25. It is surely rational (if not inescapable) to believe that a safe harbor for single words would “likely lead to more widespread use of the offensive language,” Golden Globes Order, supra, at 4979, ¶ 9.

When confronting other requests for per se rules governing its enforcement of the indecency prohibition, the Commission has declined to create safe harbors for particular types of broadcasts. See In re Pacifica Foundation, Inc., 2 FCC Red., at 2699, ¶ 12 (repudiating the view that the Commission’s enforcement power was limited to “deliberate, repetitive use of the seven words actually contained in the George Carlin monologue”); In re Infinity Broadcasting Corp. of Pa., 3 FCC Red., at 932, ¶ 17 (“rejecting] an approach that would hold that if a work has merit, it is per se not indecent”). The Commission could rationally decide it needed to step away from its old regime where nonrepetitive use of an expletive was per se nonactionable because that was “at odds with the Commission’s overall enforcement policy.” Remand Order, supra, at 13308, ¶ 23.

The fact that technological advances have made it easier for broadcasters to bleep out offending words further supports the Commission’s stepped-up enforcement policy. Golden Globes Order, supra, at 4980, ¶ 11. And the agency’s decision not to impose any forfeiture or other sanction precludes any argument that it is arbitrarily punishing parties without notice of the potential consequences of their action.

C. The Court of Appeals’ Reasoning

The Court of Appeals found the Commission’s action arbitrary and capricious on three grounds. First, the court criticized the Commission for failing to explain why it had not previously banned fleeting expletives as “harmful 'first *519blow[s].’ ” 489 F. 3d, at 458. In the majority’s view, without “evidence that suggests a fleeting expletive is harmful [and]... serious enough to warrant government regulation,” the agency could not regulate more broadly. Id., at 461. As explained above, the fact that an agency had a prior stance does not alone prevent it from changing its view or create a higher hurdle for doing so. And it is not the Commission, but Congress that has proscribed “any . . . indecent . .. language.” 18 U. S. C. § 1464.

There are some propositions for which scant empirical evidence can be marshaled, and the harmful effect of broadcast profanity on children is one of them. One cannot demand a multiyear controlled study, in which some children are intentionally exposed to indecent broadcasts (and insulated from all other indecency), and others are shielded from all indecency. It is one thing to set aside agency action under the Administrative Procedure Act because of failure to adduce empirical data that can readily be obtained. See, e. g., State Farm, 463 U. S., at 46-56 (addressing the costs and benefits of mandatory passive restraints for automobiles). It is something else to insist upon obtaining the unobtainable. Here it suffices to know that children mimic the behavior they observe — or at least the behavior that is presented to them as normal and appropriate. Programming replete with one-word indecent expletives will tend to produce children who use (at least) one-word indecent expletives. Congress has made the determination that indecent material is harmful to children, and has left enforcement of the ban to the Commission. If enforcement had to be supported by empirical data, the ban would effectively be a nullity.

The Commission had adduced no quantifiable measure of the harm caused by the language in Pacifica, and we nonetheless held that the “government’s interest in the ‘well-being of its youth’ . . . justified the regulation of otherwise protected expression.” 438 U. S., at 749 (quoting Ginsberg v. New York, 390 U. S. 629, 640, 639 (1968)). If the Constitu*520tion itself demands of agencies no more scientifically certain criteria to comply with the First Amendment, neither does the Administrative Procedure Act to comply with the requirement of reasoned decisionmaking.

The court’s second objection is that fidelity to the agency’s “first blow” theory of harm would require a categorical ban on all broadcasts of expletives; the Commission’s failure to go to this extreme thus undermined the coherence of its rationale. 489 F. 3d, at 458-459. This objection, however, is not responsive to the Commission’s actual policy under review — the decision to include patently offensive fleeting expletives within the definition of indecency. The Commission’s prior enforcement practice, unchallenged here, already drew distinctions between the offensiveness of particular words based upon the context in which they appeared. Any complaint about the Commission’s failure to ban only some fleeting expletives is better directed at the agency’s context-based system generally rather than its inclusion of isolated expletives.

More fundamentally, however, the agency’s decision to consider the patent offensiveness of isolated expletives on a case-by-case basis is not arbitrary or capricious. “Even a prime-time recitation of Geoffrey Chaucer’s Miller’s Tale,” we have explained, “would not be likely to command the attention of many children who are both old enough to understand and young enough to be adversely affected.” Pacifica, supra, at 750, n. 29. The same rationale could support the Commission’s finding that a broadcast of the film Saving Private Ryan was not indecent — a finding to which the broadcasters point as supposed evidence of the Commission’s inconsistency. The frightening suspense and the graphic violence in the movie could well dissuade the most vulnerable from watching and would put parents on notice of potentially objectionable material. See In re Complaints Against Various Television Licensees Regarding Their Broadcast on Nov. 11, 2001 of ABC Television Network's *521Presentation of Film “Saving Private Ryan,” 20 FCC Red. 4507, 4513, ¶ 15 (2005) (noting that the broadcast was not “intended as family entertainment”). The agency’s decision to retain some discretion does not render arbitrary or capricious its regulation of the deliberate and shocking uses of offensive language at the award shows under review — shows that were expected to (and did) draw the attention of millions of children.

Finally, the Court of Appeals found unconvincing the agency’s prediction (without any evidence) that a per se exemption for fleeting expletives would lead to increased use of expletives one at a time. 489 F. 3d, at 460. But even in the absence of evidence, the agency’s predictive judgment (which merits deference) makes entire sense. To predict that complete immunity for fleeting expletives, ardently desired by broadcasters, will lead to a substantial increase in fleeting expletives seems to us an exercise in logic rather than clairvoyance. The Court of Appeals was perhaps correct that the Commission’s prior policy had not yet caused broadcasters to “barrag[e] the airwaves with expletives,” ibid. That may have been because its prior permissive policy had been confirmed (save in dicta) only at the staff level. In any event, as the Golden Globes order demonstrated, it did produce more expletives than the Commission (which has the first call in this matter) deemed in conformity with the statute.

D. Respondents’ Arguments

Respondents press some arguments that the court did not adopt. They claim that the Commission failed to acknowledge its change in enforcement policy. That contention is not tenable in light of the Golden Globes Order’s specific declaration that its prior rulings were no longer good law, 19 FCC Red., at 4980, ¶ 12, and the Remand Order’s disavowal of those staff rulings and Commission dicta as “seriously flawed,” 21 FCC Red., at 13308, ¶ 23. The broadcasters also try to recharacterize the nature of the Commission’s shift, *522contending that the old policy was not actually a per se rule against liability for isolated expletives and that the new policy is a presumption of indecency for certain words. This description of the prior agency policy conflicts with the broadcasters’ own prior position in this case. See, e.g., Brief in Opposition for Respondent Fox Television Stations, Inc., et al. 4 (“For almost 30 years following Pacifica, the FCC did not consider fleeting, isolated or inadvertent expletives to be indecent”). And we find no basis for the contention that the Commission has now adopted a presumption of indecency; its repeated reliance on context refutes this claim.

The broadcasters also make much of the fact that the Commission has gone beyond the scope of authority approved in Pacifica, which it once regarded as the farthest extent of its power. But we have never held that Pacifica represented the outer limits of permissible regulation, so that fleeting expletives may not be forbidden. To the contrary, we explicitly left for another day whether “an occasional expletive” in “a telecast of an Elizabethan comedy” could be prohibited. 438 U. S., at 748-750. By using the narrowness of Pacifica!s holding to require empirical evidence of harm before the Commission regulates more broadly, the broadcasters attempt to turn the sword of Pacifica, which allowed some regulation of broadcast indecency, into an administrative-law shield preventing any regulation beyond what Pacifica sanctioned. Nothing prohibits federal agencies from moving in an incremental manner. Cf. National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U. S. 967, 1002 (2005).

Finally, the broadcasters claim that the Commission’s repeated appeal to “context” is simply a smokescreen for a standardless regime of unbridled discretion. But we have previously approved Commission regulation based “on a nuisance rationale under which context is all-important,” Pacifica, supra, at 750, and we find no basis in the Administrative Procedure Act for mandating anything different.

*523E. The Dissents’ Arguments

Justice Breyer purports to “begin with applicable law,” post, at 547, but in fact begins by stacking the deck. He claims that the FCC’s status as an “independent” agency sheltered from political oversight requires courts to be “all the more” vigilant in ensuring “that major policy decisions be based upon articulable reasons.” Ibid. Not so. The independent agencies are sheltered not from politics but from the President, and it has often been observed that their freedom from Presidential oversight (and protection) has simply been replaced by increased subservience to congressional direction. See, e. g., In re Sealed Case, 838 F. 2d 476, 507-508 (CADC) (Silbermian, J.), rev’d sub nom. Morrison v. Olson, 487 U. S. 654 (1988); Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2271, n. 93 (2001); Calabresi & Prakash, The President’s Power to Execute the Laws, 104 Yale L. J. 541, 583 (1994); Easterbrook, The State of Madison’s Vision of the State: A Public Choice Perspective, 107 Harv. L. Rev. 1328, 1341 (1994). Indeed, the precise policy change at issue here was spurred by significant political pressure from Congress.4

*524Justice Stevens apparently recognizes this political control by Congress, and indeed sees it as the manifestation of a principal-agency relationship. In his judgment, the FCC is “better viewed as an agent of Congress” than as part of the Executive. Post, at 540 (dissenting opinion). He nonetheless argues that this is a good reason for requiring the FCC to explain “why its prior policy is no longer sound before allowing it to change course.” Post, at 541. Leaving aside the unconstitutionally of a scheme giving the power to enforce laws to agents of Congress, see Bowsher v. Synar, 478 U. S. 714, 726 (1986), it seems to us that Justice Stevens’ conclusion does not follow from his premise. If the FCC is indeed an agent of Congress, it would seem an adequate explanation of its change of position that Congress *525made clear its wishes for stricter enforcement, see n. 4, supra.5 The Administrative Procedure Act, after all, does not apply to Congress and its agencies.6

Regardless, it is assuredly not “applicable law” that rule-making by independent regulatory agencies is subject to heightened scrutiny. The Administrative Procedure Act, which provides judicial review, makes no distinction between independent and other agencies, neither in its definition of agency, 5 U. S. C. § 701(b)(1), nor in the standards for reviewing agency action, § 706. Nor does any case of ours express or reflect the “heightened scrutiny” Justice Breyer and Justice Stevens would impose. Indeed, it is hard to imagine any closer scrutiny than that we have given to the Environmental Protection Agency, which is not an independent agency. See Massachusetts v. EPA, 549 U. S. 497, 533-535 (2007); Whitman v. American Trucking Assns., Inc., 531 U. S. 457, 481-486 (2001). There is no reason to magnify the separation-of-powers dilemma posed by the headless Fourth *526Branch, see Freytag v. Commissioner, 501 U. S. 868, 921 (1991) (Scalia, J., concurring in part and concurring in judgment), by letting Article III judges — like jackals stealing the lion’s kill — expropriate some of the power that Congress has wrested from the unitary Executive.

Justice Breyer and Justice Stevens rely upon two supposed omissions in the FCC’s analysis that they believe preclude a finding that the agency did not act arbitrarily. Neither of these omissions could undermine the coherence of the rationale the agency gave, but the dissenters’ evaluation of each is flawed in its own right.

First, both claim that the Commission failed adequately to explain its consideration of the constitutional issues inherent in its regulation, post, at 553-556 (opinion of Breyer, J.); post, at 542-546 (opinion of Stevens, J.). We are unaware that we have ever before reversed an executive agency, not for violating our cases, but for failure to discuss them adequately. But leave that aside. According to Justice Breyer, the agency said “next to nothing about the relation between the change it made in its prior ‘fleeting expletive’ policy and the First-Amendment-related need to avoid ‘censorship,’ ” post, at 553. The Remand Order does, however, devote four full pages of small-type, single-spaced text (over 1,300 words not counting the footnotes) to explaining why the Commission believes that its indecency-enforcement regime (which includes its change in policy) is consistent with the First Amendment — and therefore not censorship as the term is understood. More specifically, Justice Breyer faults the FCC for “not explaining] why the agency changed its mind about the line that Pacifica draws or its policy’s relation to that line,” post, at 556. But in fact (and as the Commission explained) this Court’s holding in Pacifica, 438 U. S. 726, drew no constitutional line; to the contrary, it expressly declined to express any view on the constitutionality of prohibiting isolated indecency. Justice Breyer and Justice Stevens evidently believe that when an agency has *527obtained this Court’s determination that a less restrictive rule is constitutional, its successors acquire some special burden to explain why a more restrictive rule is not '^constitutional. We know of no such principle.7

Second, Justice Breyer looks over the vast field of particular factual scenarios unaddressed by the FCC’s 35-page Remand Order and finds one that is fatal: the plight of the small local broadcaster who cannot afford the new technology that enables the screening of live broadcasts for indecent utterances. Cf. post, at 556-561. The Commission has failed to address the fate of this unfortunate, who will, he believes, be subject to sanction.

We doubt, to begin with, that small-town broadcasters run a heightened risk of liability for indecent utterances. In programming that they originate, their down-home local guests probably employ vulgarity less than big-city folks; and small-town stations generally cannot afford or cannot attract foul-mouthed glitteratae from Hollywood. Their main exposure with regard to self-originated programming is live coverage of news and public affairs. But the Remand Order went out of its way to note that the ease at hand did not involve “breaking news coverage,” and that “it may be inequitable to hold a licensee responsible for airing offensive *528speech during live coverage of a public event,” 21 FCC Red., at 13311, ¶ 33. As for the programming that small stations receive on a network “feed”: This will be cleansed by the expensive technology small stations (by Justice Breyer’s hypothesis) cannot afford.

But never mind the detail of whether small broadcasters are uniquely subject to a great risk of punishment for fleeting expletives. The fundamental fallacy of Justice Breyer’s small-broadcaster gloomy scenario is its demonstrably false assumption that the Remand Order makes no provision for the avoidance of unfairness — that the single-utterance prohibition will be invoked uniformly, in all situations. The Remand Order made very clear that this is not the case. It said that in determining “what, if any, remedy is appropriate” the Commission would consider the facts of each individual case, such as the “possibility of human error in using delay equipment,” id., at 13313, ¶ 35. Thus, the fact that the agency believed that Fox (a large broadcaster that used suggestive scripting and a deficient delay system to air a prime-time awards show aimed at millions of children) “fail[ed] to exercise ‘reasonable judgment, responsibility and sensitivity,”’ id., at 13311, ¶33, and n. 91 (quoting Pacifica Foundation, Inc., 2 FCC Red., at 2700, ¶ 18), says little about how the Commission would treat smaller broadcasters who cannot afford screening equipment. Indeed, that they would not be punished for failing to purchase equipment they cannot afford is positively suggested by the Remand Order’s statement that “[hjolding Fox responsible for airing indecent material in this case does not. .. impose undue burdens on broadcasters.” 21 FCC Red., at 13313, ¶ 36.

There was, in sum, no need for the Commission to compose a special treatise on local broadcasters.8 And Justice *529Breyer can safely defer his concern for those yeomen of the airwaves until we have before us a case that involves one.

IV. Constitutionality

The Second Circuit did not definitively rule on the constitutionality of the Commission’s orders, but respondents nonetheless ask us to decide their validity under the First Amendment. This Court, however, is one of final review, “not of first view.” Cutter v. Wilkinson, 544 U. S. 709, 718, n. 7 (2005). It is conceivable that the Commission’s orders may cause some broadcasters to avoid certain language that is beyond the Commission’s reach under the Constitution. Whether that is so, and, if so, whether it is unconstitutional, will be determined soon enough, perhaps in this very case. Meanwhile, any chilled references to excretory and sexual material “surely lie at the periphery of First Amendment concern,” Pacifica, 438 U. S., at 743 (plurality opinion of Stevens, J.). We see no reason to abandon our usual procedures in a rush to judgment without a lower court opinion. We decline to address the constitutional questions at this time.

The Second Circuit believed that children today “likely hear this language far more often from other sources than they did in the 1970s when the Commission first began sanctioning indecent speech,” and that this cuts against more stringent regulation of broadcasts. 489 F. 3d, at 461. Assuming the premise is true (for this point the Second Circuit did not demand empirical evidence) the conclusion does not necessarily follow. The Commission could reasonably conclude that the pervasiveness of foul language, and the coars*530ening of public entertainment in other media such as cable, justify more stringent regulation of broadcast programs so as to give conscientious parents a relatively safe haven for their children. In the end, the Second Circuit and the broadcasters quibble with the Commission’s policy choices and not with the explanation it has given. We decline to “substitute [our] judgment for that of the agency,” State Farm, 468 U. S., at 43, and we find the Commission’s orders neither arbitrary nor capricious.

The judgment of the United States Court of Appeals for the Second Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

The statutory prohibition applicable to commercial radio and television stations extends by its terms from 6 a.m. to 12 midnight. The Court of Appeals for the District of Columbia Circuit held, however, that because “Congress and the Commission [had] backed away from the consequences of their own reasoning,” by allowing some public broadcasters to air indecent speech after 10 p.m., the court was forced “to hold that the section is unconstitutional insofar as it bars the broadcasting of indecent speech between the hours of 10:00 p.m. and midnight.” Action for Children’s Television v. FCC, 58 F. 3d 654, 669 (1995) (en banc), cert. denied, 516 U. S. 1043 (1996).

Justice Breyer’s contention that State Farm did anything more, post, at 549-552 (dissenting opinion), rests upon his failure to observe the italicized phrase and upon a passage quoted in State Farm from a plurality opinion in Atchison, T. & S. F. R. Co. v. Wichita Bd. of Trade, 412 U. S. 800 (1973). That passage referred to “a presumption that [congressional] policies will be carried out best if the settled rule is adhered to.” Id., at 807-808 (opinion of Marshall, J.). But the Atchison plurality made this statement in the context of requiring the agency to provide some explanation for a change, “so that the reviewing court may understand the basis *515of the agency’s action and so may judge the consistency of that action with the agency’s mandate,” id., at 808. The opinion did not assert the authority of a court to demand explanation sufficient to enable it to weigh (by its own lights) the merits of the agency’s change. Nor did our opinion in State Farm.

Justice Bkeyer claims that “[t]he Court has often applied [the doctrine of constitutional avoidance] where an agency’s regulation relies on a plausible but constitutionally suspect interpretation of a statute.” Post, at 566. The cases he cites, however, set aside an agency regulation because, applying the doctrine of constitutional avoidance to the ambiguous statute under which the agency acted, the Court found the agency’s interpretation of the statute erroneous. See Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U. S. 159, 174 (2001); NLRB v. Catholic Bishop of Chicago, 440 U. S. 490, 507 (1979). But Justice *517Breyer does not urge that we issue such a holding, evidently agreeing that we should limit our review to what the Court of Appeals decided, see Part IV, infra — which included only the adequacy of the Commission’s rulemaking procedure, and not the statutory question. Rather, Justice Breyer seeks a “remand [that] would do no more than ask the agency to reconsider its policy decision in light of” constitutional concerns. Post, at 566. That strange and novel disposition would be entirely unrelated to the doctrine of constitutional avoidance, and would better be termed the doctrine of judicial arm-twisting or appellate review by the wagged finger.

A Subcommittee of the FCC’s House Oversight Committee held hearings on the FCC’s broadcast indecency enforcement on January 28, 2004. “Can You Say That on TV?”: An Examination of the FCC’s Enforcement with Respect to Broadcast Indecency, Hearing before the Subcommittee on Telecommunications and the Internet of the House Committee on Energy and Commerce, 108th Cong., 2d Sess. Members of the Subcommittee specifically “called on the full Commission to reverse [the staff ruling in the Golden Globes case]” because they perceived a “feeling amongst many Americans that some TV broadcasters are engaged in a race to the bottom, pushing the decency envelope in order to distinguish themselves in the increasingly crowded entertainment field.” Id., at 2 (statement of Rep. Upton); see also, e. g., id., at 17 (statement of Rep. Terry), 19 (statement of Rep. Pitts). They repeatedly expressed disapproval of the FCC’s enforcement policies, see, e. g., id., at 3 (statement of Rep. Upton) (“At some point, we have to ask the FCC: How much is enough? When will it revoke a license?”); id., at 4 (statement of Rep. Markey) (“Today’s hearing *524will allow us to explore the FCC’s lackluster enforcement record with respect to these violations”).

About two weeks later, on February 11, 2004, the same Subcommittee held hearings on a bill increasing the fines for indecency violations. Hearings on H. R. 3717 before the Subcommittee on Telecommunications and the Internet of the House Committee on Energy and Commerce, 108th Cong., 2d Sess. All five Commissioners were present and were grilled about enforcement shortcomings. See, e. g., id., at 124 (statement of Rep. Terry) (“Chairman Powell, ... it seems like common sense that if we had . . . more frequent enforcement instead of only a few examples of fines ... that would be a deterrent in itself”); id., at 7 (statement of Rep. Dingell) (“I see that apparently ... there is no enforcement of regulations at the FCC”). Certain statements, moreover, indicate that the political pressure applied by Congress had its desired effect. See ibid. (“I think our committee’s work has gotten the attention of FCC Chairman Powell and the Bush Administration. And I’m happy to see the FCC now being brought to a state of apparent alert on these matters”); see also id., at 124 (statement of Michael Copps, FCC Commissioner) (noting “positive” change in other Commissioners’ 'willingness to step up enforcement in light of proposed congressional action). A version of the bill ultimately became law as the Broadcast Decency Enforcement Act of 2005, 120 Stat. 491.

The FCC adopted the change that is the subject of this litigation on March 3,2004, about three weeks after this second hearing. See Golden Globes Order, 19 FCC Rcd. 4975.

Justice Stevens accuses us of equating statements made in a congressional hearing with the intent of Congress. Post, at 541-542, n. 3. In this opinion, we do not. The intent of the full Congress (or at least a majority of each House) is thought relevant to the interpretation of statutes, since they must be passed by the entire Congress. See U. S. Const., Art. I, § 7. It is quite irrelevant, however, to the extrastatutory influence Congress exerts over agencies of the Executive Branch, which is exerted by the congressional committees responsible for oversight and appropriations with respect to the relevant agency. That is a major reason why committee assignments are important, and committee chairmanships powerful. Surely Justice Stevens knows this.

The Administrative Procedure Act defines “agency” to mean “each authority of the Government of the United States,” 5 U. S. C. § 551(1), but specifically excludes “the Congress,” § 551(1)(A). The Court of Appeals for the District of Columbia Circuit has “interpreted [this] exemption for ‘the Congress’ to mean the entire legislative branch,” Washington Legal Foundation v. United States Sentencing Comm’n, 17 F. 3d 1446, 1449 (1994); see also Ethnic Employees of Library of Congress v. Boorstin, 751 F. 2d 1405, 1416, n. 15 (CADC 1985) (holding that the Library of Congress is not an “agency” under the Act).

Justice Stevens criticizes us for “assuming that Pacifica endorsed” the enforcement at issue here. Post, at 542. We do nothing of the sort. We rely on the fact that certain aspects of the agency’s decision mirror the context-based approach Pacifica approved, supra, at 517-518, but that goes to our holding on administrative law, and says nothing about constitutionality. Justice Stevens also argues that heightened deference should be due the FCC’s prior policy because the “FCC’s initial views ... reflect the views of the Congress that delegated the Commission authority to flesh out details not fully defined in the enacting statute.” Post, at 541. We do not believe that the dead hand of a departed congressional oversight Committee should constrain the discretion that the text of a statute confers — but the point is in any event irrelevant in this appeal, which concerns not whether the agency has exceeded its statutory mandate but whether the reasons for its actions are adequate.

Justice Breyer posits that the FCC would have been required to give more explanation had it used notice-and-comment rulemaking, which “should lead us to the same conclusion” in this review of the agency’s change through adjudication. Post, at 562. Even assuming the premise, *529there is no basis for incorporating all of the Administrative Procedure Act’s notice-and-comment procedural requirements into arbitrary-andeapricious review of adjudicatory decisions. Cf. Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U. S. 519, 545-549 (1978).






Concurrence Opinion

Justice Thomas,

concurring.

I join the Court’s opinion, which, as a matter of administrative law, correctly upholds the Federal Communications Commission’s (FCC) policy with respect to indecent broadcast speech under the Administrative Procedure Act. I write separately, however, to note the questionable viability of the two precedents that support the FCC’s assertion of constitutional authority to regulate the programming at issue in this case. See Red Lion Broadcasting Co. v. FCC, 395 U. S. 367 (1969); FCC v. Pacifica Foundation, 438 U. S. 726 (1978). Red Lion and Pacifica were unconvincing when they were issued, and the passage of time has only increased doubt regarding their continued validity. “The text of the First Amendment makes no distinctions among print, broadcast, and cable media, but we have done so” in these cases. Denver Area Ed. Telecommunications Consortium, Inc. v. FCC, 518 U. S. 727, 812 (1996) (Thomas, J., concurring in judgment in part and dissenting in part).

In Red Lion, this Court upheld the so-called “fairness doctrine,” a Government requirement “that discussion of public issues be presented on broadcast stations, and that each side of those issues must be given fair coverage.” 395 U. S., at *531369, 400-401. The decision relied heavily on the scarcity of available broadcast frequencies. According to the Court, because broadcast spectrum was so scarce, it “could be regulated and rationalized only by the Government. Without government control, the medium would be of little use because of the cacophony of competing voices, none of which could be clearly and predictably heard.” Id., at 376. To this end, the Court concluded that the Government should be “permitted to put restraints on licensees in favor of others whose views should be expressed on this unique medium.” Id., at 390; see also id., at 389 (concluding that “as far as the First Amendment is concerned those who are licensed stand no better than those to whom licenses are refused”). Applying this principle, the Court held that “[i]t does not violate the First Amendment to treat licensees given the privilege of using scarce radio frequencies as proxies for the entire community, obligated to give suitable time and attention to matters of great public concern.” Id., at 394.

Red Lion specifically declined to answer whether the First Amendment authorized the Government’s “refusal to permit the broadcaster to carry a particular program or to publish his own views[,]... [or] government censorship of a particular program,” id., at 396. But then in Pacifica, this Court rejected a challenge to the FCC’s authority to impose sanctions on the broadcast of indecent material. See 438 U. S., at 729-730, 750-751; id., at 742 (plurality opinion). Relying on Red Lion, the Court noted that “broadcasting . . . has received the most limited First Amendment protection.” 438 U. S., at 748. The Court also emphasized the “uniquely pervasive presence” of the broadcast media in Americans’ lives and the fact that broadcast programming was “uniquely accessible to children.” Id., at 748-749.

This deep intrusion into the First Amendment rights of broadcasters, which the Court has justified based only on the nature of the medium, is problematic on two levels. First, instead of looking to first principles to evaluate the constitu*532tional question, the Court relied on a set of transitory facts, e. g., the “scarcity of radio frequencies,” Red Lion, supra, at 390, to determine the applicable First Amendment standard. But the original meaning of the Constitution cannot turn on modern necessity: “Constitutional rights are enshrined with the scope they were understood to have when the people adopted them, whether or not future legislatures or (yes) even future judges think that scope too broad.” District of Columbia v. Heller, 554 U.S. 570, 634-635 (2008). In breaching this principle, Red Lion adopted, and Pacifica reaffirmed, a legal rule that lacks any textual basis in the Constitution. Denver Area, supra, at 813 (Thomas, J., concurring in judgment in part and dissenting in part) (“First Amendment distinctions between media [have been] dubious from their infancy”). Indeed, the logical weakness of Red Lion and Pacifica has been apparent for some time: “It is certainly true that broadcast frequencies are scarce but it is unclear why that fact justifies content regulation of broadcasting in a way that would be intolerable if applied to the editorial process of the print media.” Telecommunications Research & Action Center v. FCC, 801 F. 2d 501, 508 (CADC 1986) (Bork, J.).

Highlighting the doctrinal incoherence of Red Lion and Pacifica, the Court has declined to apply the lesser standard of First Amendment scrutiny imposed on broadcast speech to federal regulation of telephone dial-in services, see Sable Communications of Cal, Inc. v. FCC, 492 U. S. 115, 127-128 (1989), cable television programming, see Turner Broadcasting System, Inc. v. FCC, 512 U. S. 622, 637 (1994), and the Internet, see Reno v. American Civil Liberties Union, 521 U. S. 844, 867-868 (1997). “There is no justification for this apparent dichotomy in First Amendment jurisprudence. Whatever the merits of Pacifica when it was issued[,]... it makes no sense now.” Action for Children’s Television v. FCC, 58 F. 3d 654, 673 (CADC 1995) (Edwards, C. J., dissenting). The justifications relied on by the Court in Red Lion *533and Pacifica — “spectrum scarcity, intrusiveness, and accessibility to children — neither distinguish broadcast from cable, nor explain the relaxed application of the principles of the First Amendment to broadcast.” 58 F. 3d, at 673; see also In re Industry Guidance on Commission’s Case Law Interpreting 18 US. C. §1464 and Enforcement Policies Regarding Broadcast Indecency, 16 FCC Red. 7999, 8021, n. 11 (2001) (statement of Commissioner Furchtgott-Roth) (“It is ironic that streaming video or audio content from a television or radio station would likely receive more constitutional protection, see Reno [v. American Civil Liberties Union, 521 U. S. 844 (1997)], than would the same exact content broadcast over-the-air”).

Second, even if this Court’s disfavored treatment of broadcasters under the First Amendment could have been justified at the time of Red Lion and Pacifica, dramatic technological advances have eviscerated the factual assumptions underlying those decisions. Broadcast spectrum is significantly less scarce than it was 40 years ago. See Brief for Respondent NBC Universal et al. 37-38 (hereinafter NBC Brief). As NBC notes, the number of over-the-air broadcast stations grew from 7,411 in 1969, when Red Lion was issued, to 15,273 by the end of 2004. See NBC Brief 37-38; see also FCC Media Bureau Staff Research Paper, J. Berresford, The Scarcity Rationale for Regulating Traditional Broadcasting: An Idea Whose Time Has Passed 12-13 (Mar. 2005) (No. 2005-2). And the trend should continue with broadcast television’s imminent switch from analog to digital transmission, which will allow the FCC to “stack broadcast channels right beside one another along the spectrum, and ultimately utilize significantly less than the 400 MHz of spectrum the analog system absorbs today.” Consumer Electronics Assn. v. FCC, 347 F. 3d 291, 294 (CADC 2003).

Moreover, traditional broadcast television and radio are no longer the “uniquely pervasive” media forms they once were. For most consumers, traditional broadcast media program*534ming is now bundled with cable or satellite services. See App. to Pet. for Cert. 106a-107a. Broadcast and other video programming is also widely available over the Internet. See Stelter, Serving Up Television Without the TV Set, N. Y. Times, Mar. 10, 2008, p. Cl. And like radio and television broadcasts, Internet access is now often freely available over the airwaves and can be accessed by portable computer, cell phones, and other wireless devices. See May, Charting a New Constitutional Jurisprudence for the Digital Age, 3 Charleston L. Rev. 373, 375 (2009). The extant facts that drove this Court to subject broadcasters to unique disfavor under the First Amendment simply do not exist today. See In re Industry Guidance, supra, at 8020 (statement of Commissioner Furchtgott-Roth) (“If rules regulating broadcast content were ever a justifiable infringement of speech, it was because of the relative dominance of that medium in the communications marketplace of the past. As the Commission has long recognized, the facts underlying this justification are no longer true” (footnote omitted)).*

These dramatic changes in factual circumstances might well support a departure from precedent under the prevailing approach to stare decisis. See Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 855 (1992) (asking “whether facts have so changed, or come to be seen so differently, as to have robbed the old rule of significant application or justification”); see also American Trucking Assns., Inc. v. Scheiner, 483 U. S. 266, 302 (1987) (O’Connor, J., dissenting) (“Significantly changed circumstances can make an older rule, defensible when formulated, inappropriate .. . ”). “In *535cases involving constitutional issues” that turn on a particular set of factual assumptions, “this Court must, in order to reach sound conclusions, feel free to bring its opinions into agreement with experience and with facts newly ascertained.” Burnet v. Coronado Oil & Gas Co., 285 U. S. 393, 412 (1932) (Brandeis, J., dissenting). For all these reasons, I am open to reconsideration of Red Lion and Pacifica in the proper case.

With respect to reliance by FCC v. Pacifica Foundation, 438 U. S. 726 (1978), on the ease with which children could be exposed to indecent television programming, technology has provided innovative solutions to assist adults in screening their children from unsuitable programming — even when that programming appears on broadcast channels. See NBC Brief 43-47 (discussing V-chip technology, which allows targeted blocking of television programs based on content).






Concurrence in Part

Justice Kennedy,

concurring in part and concurring in the judgment.

I join Parts I, II, III-A through III-D, and IV of the opinion of the Court and agree that the judgment must be reversed. This separate writing is to underscore certain background principles for the conclusion that an agency’s decision to change course may be arbitrary and capricious if the agency sets a new course that reverses an earlier determination but does not provide a reasoned explanation for doing so. In those circumstances I agree with the dissenting opinion of Justice Breyer that the agency must explain why “it now reject[s] the considerations that led it to adopt that initial policy.” Post, at 550.

The question whether a change in policy requires an agency to provide a more reasoned explanation than when the original policy was first announced is not susceptible, in my view, to an answer that applies in all cases. There may be instances when it becomes apparent to an agency that the reasons for a longstanding policy have been altered by discoveries in science, advances in technology, or by any of the other forces at work in a dynamic society. If an agency seeks to respond to new circumstances by modifying its earlier policy, the agency may have a substantial body of data and experience that can shape and inform the new rule. In other cases the altered circumstances may be so new that the agency must make predictive judgments that are as difficult now as when the agency’s earlier policy was first an*536nounced. Reliance interests in the prior policy may also have weight in the analysis.

The question in each case is whether the agency’s reasons for the change, when viewed in light of the data available to it, and when informed by the experience and expertise of the agency, suffice to demonstrate that the new policy rests upon principles that are rational, neutral, and in accord with the agency’s proper understanding of its authority. That showing may be required if the agency is to demonstrate that its action is not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U. S. C. § 706(2)(A). And, of course, the agency action must not be “in excess of statutory jurisdiction, authority, or limitations, or short of statutory right.” §706(2)(C).

These requirements stem from the administrative agency’s unique constitutional position. The dynamics of the three branches of Government are well understood as a general matter. But the role and position of the agency, and the exact locus of its powers, present questions that are delicate, subtle, and complex. The Federal Government could not perform its duties in a responsible and effective way without administrative agencies. Yet the amorphous character of the administrative agency in the constitutional system escapes simple explanation.

If agencies were permitted unbridled discretion, their actions might violate important constitutional principles of separation of powers and checks and balances. To that end the Constitution requires that Congress’ delegation of lawmaking power to an agency must be “specific and detailed.” Mistretta, v. United States, 488 U. S. 361, 374 (1989). Congress must “‘clearly delineat[e] the general policy’” an agency is to achieve and must specify the “‘boundaries of [the] delegated authority.’” Id., at 372-373. Congress must “ ‘lay down by legislative act an intelligible principle,’ ” and the agency must follow it. Id., at 372 (quoting J. W. *537Hampton, Jr., & Co. v. United States, 276 U. S. 394, 409 (1928)).

Congress passed the Administrative Procedure Act (APA) to ensure that agencies follow constraints even as they exercise their powers. One of these constraints is the duty of agencies to find and formulate policies that can be justified by neutral principles and a reasoned explanation. To achieve that end, Congress confined agencies’ discretion and subjected their decisions to judicial review. See Stewart & Sunstein, Public Programs and Private Rights, 95 Harv. L. Rev. 1193, 1248 (1982) (the APA was a “working compromise, in which broad delegations of discretion were tolerated as long as they were checked by extensive procedural safeguards”). If an agency takes action not based on neutral and rational principles, the APA grants federal courts power to set aside the agency’s action as “arbitrary” or “capricious.” 5 U. S. C. § 706(2)(A); Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U. S. 402, 416 (1971). For these reasons, agencies under the APA are subject to a “searching and careful” review by the courts. Ibid.

Where there is a policy change the record may be much more developed because the agency based its prior policy on factual findings. In that instance, an agency’s decision to change course may be arbitrary and capricious if the agency ignores or countermands its earlier factual findings without reasoned explanation for doing so. An agency cannot simply disregard contrary or inconvenient factual determinations that it made in the past, any more than it can ignore inconvenient facts when it writes on a blank slate.

This is the principle followed in the Court’s opinion in Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29 (1983). There, Congress directed the agency to issue regulations that would “ ‘meet the need for motor vehicle safety.’ ” Id., at 33. The agency promulgated a regulation requiring cars to have passive-restraint systems — either airbags or automatic seat-*538belts. Id., at 37. The agency based this regulation on its factual finding that these systems save lives. Id., at 35.

Following a change in Presidential administration, however, the agency reversed course and rescinded the regulation. In doing so, the agency did not address its prior finding that airbags save lives. Id., at 47-48. Indeed, “[n]ot one sentence” of the agency’s “rulemaking statement” in support of rescinding the regulation discussed the benefits of airbags. Id., at 48. This Court found the agency’s rescission arbitrary and capricious because the agency did not address its prior factual findings. See id., at 49-51.

The present case does not raise the concerns addressed in State Farm. Rather than base its prior policy on its knowledge of the broadcast industry and its audience, the FCC instead based its policy on what it considered to be our holding in FCC v. Pacifica Foundation, 438 U. S. 726 (1978). See In re Application of WGBH Educ. Foundation, 69 F. C. C. 2d 1250, 1254, ¶ 10 (1978) (“We intend strictly to observe the narrowness of the Pacifica holding”). The FCC did not base its prior policy on factual findings.

The FCC's Remand Order explains that the agency has changed its reading of Pacifica. The reasons the agency announces for this change are not so precise, detailed, or elaborate as to be a model for agency explanation. But, as the opinion for the Court well explains, the FCC’s reasons for its action were the sort of reasons an agency may consider and act upon. The Court’s careful and complete analysis — both with respect to the procedural history of the FCC’s indecency policies, and the reasons the agency has given to support them — is quite sufficient to sustain the FCC’s change of course against respondents’ claim that the agency acted in an arbitrary or capricious fashion.

The holding of the Court of Appeals turned on its conclusion that the agency’s explanation for its change of policy was insufficient, and that is the only question presented here. I agree with the Court that as this case comes to us from the *539Court of Appeals we must reserve judgment on the question whether the agency’s action is consistent with the guarantees of the Constitution.






Dissenting Opinion

Justice Stevens,

dissenting.

While I join Justice Breyer’s cogent dissent, I think it important to emphasize two flaws in the Court’s reasoning. Apparently assuming that the Federal Communications Commission’s (FCC or Commission) rulemaking authority is a species of executive power, the Court espouses the novel proposition that the Commission need not explain its decision to discard a longstanding rule in favor of a dramatically different approach to regulation. See ante, at 514-515. Moreover, the Court incorrectly assumes that our decision in FCC v. Pacifica Foundation, 438 U. S. 726 (1978), decided that the word “indecent,” as used in 18 U. S. C. § 1464,1 permits the FCC to punish the broadcast of any expletive that has a sexual or excretory origin. Pacifica was not so sweeping, and the Commission’s changed view of its statutory mandate certainly would have been rejected if presented to the Court at the time.

I

“The structure of our Government as conceived by the Framers of our Constitution disperses the federal power among the three branches — the Legislative, the Executive, and the Judicial — placing both substantive and procedural limitations on each.” Metropolitan Washington Airports Authority v. Citizens for Abatement of Aircraft Noise, Inc., 501 U. S. 252, 272 (1991). The distinction among the branches is not always sharp, see Bowsher v. Synar, 478 U. S. 714, 749 (1986) (Stevens, J., concurring in judgment) (citing cases), a consequence of the fact that the “great ordinances of the Constitution do not establish and divide fields of black *540and white,” Springer v. Philippine Islands, 277 U. S. 189, 209 (1928) (Holmes, J., dissenting). Strict lines of authority are particularly elusive when Congress and the President both exert a measure of control over an agency. As a landmark decision involving the Federal Trade Commission (FTC) made clear, however, when Congress grants rule-making and adjudicative authority to an expert agency composed of commissioners selected through a bipartisan procedure and appointed for fixed terms, it substantially insulates the agency from executive control. See Humphrey’s Executor v. United States, 295 U. S. 602, 623-628 (1935).

With the view that broadcast regulation “should be as free from political influence or arbitrary control as possible,” S. Rep. No. 772, 69th Cong., 1st Sess., 2 (1926), Congress established the FCC with the same measure of independence from the Executive that it had provided the FTC. Just as the FCC’s Commissioners do not serve at the will of the President, see 47 U. S. C. § 154(c) (2000 ed.), its regulations are not subject to change at the President’s will. And when the Commission fashions rules that govern the airwaves, it exercises legislative power delegated to it by Congress. See Whitman v. American Trucking Assns., Inc., 531 U. S. 457, 489-490 (2001) (Stevens, J., concurring in part and concurring in judgment); Bowsher, 478 U. S., at 752 (opinion of Stevens, J.). Consequently, the FCC “cannot in any proper sense be characterized as an arm or an eye of the executive” and is better viewed as an agent of Congress established “to carry into effect legislative policies embodied in the statute in accordance with the legislative standard therein prescribed, and to perform other specified duties as a legislative . . . aid.” Humphrey’s Executor, 295 U. S., at 628.2

*541The FCC, like all agencies, may revise its regulations from time to time, just as Congress amends its statutes as circumstances warrant. But the FCC is constrained by its congressional mandate. There should be a strong presumption that the FCC’s initial views, reflecting the informed judgment of independent Commissioners with expertise in the regulated area, also reflect the views of the Congress that delegated the Commission authority to flesh out details not fully defined in the enacting statute. The rules adopted after Pacifica, 438 U. S. 726, have been in effect for decades and have not proved unworkable in the intervening years. As Justice Breyer’s opinion explains, broadcasters have a substantial interest in regulatory stability; the threat of crippling financial penalties looms large over these entities. See post, at 556-561. The FCC’s shifting and impermissibly vague indecency policy only imperils these broadcasters and muddles the regulatory landscape. It therefore makes eminent sense to require the Commission to justify why its prior policy is no longer sound before allowing it to change course.3 *542The FCC’s congressional charter, 47 U. S. C. § 151 et seq., the Administrative Procedure Act, 5 U. S. C. § 706(2)(A) (2006 ed.) (instructing courts to “hold unlawful and set aside . . . arbitrary [or] capricious” agency action), and the rule of law all favor stability over administrative whim.

II

The Court commits a second critical error by assuming that Pacifica endorsed a construction of the term “indecent,” as used in 18 U. S. C. § 1464, that would include any expletive that has a sexual or excretory origin. Neither the opinion of the Court, nor Justice Powell’s concurring opinion, adopted such a far-reaching interpretation. Our holding was narrow in two critical respects. First, we concluded, over the dissent of four Justices, that the statutory term “indecent” was not limited to material that had prurient appeal and instead included material that was in “nonconformance with accepted standards of morality.” Pacifica, 438 U. S., at 740. Second, we upheld the FCC’s adjudication that a 12-minute, expletive-filled monologue by satiric humorist George Carlin was indecent “as broadcast.” Id., at 735. We did not decide whether an isolated expletive could qualify as indecent. Id., at 750; id., at 760-761 (Powell, J., concurring in part and concurring in judgment). And we certainly did not hold that any word with a sexual or scatological origin, however used, was indecent.

The narrow treatment of the term “indecent” in Pacifica defined the outer boundaries of the enforcement policies adopted by the FCC in the ensuing years. The Commission originally explained that “under the legal standards set forth in Pacifica, deliberate and repetitive use [of expletives] in a *543patently offensive manner is a requisite to a finding of indecency.” In re Pacifica Foundation, 2 FCC Red. 2698,2699, ¶ 13 (1987). While the “repetitive use” issue has received the most attention in this case, it should not be forgotten that Pacifica permitted the Commission to regulate only those words that describe sex or excrement. See 438 U. S., at 743 (plurality opinion) (“[T]he Commission’s definition of indecency will deter only the broadcasting of patently offensive references to excretory and sexual organs and activities” (emphasis added)). The FCC minimizes the strength of this limitation by now claiming that any use of the words at issue in this case, in any context and in any form, necessarily describes sex or excrement. See In re Complaints Regarding Various Television Broadcasts Between Feb. 2, 2002 and Mar. 8,2005,21 FCC Red. 13299,13308, ¶ 23 (2006) (Remand Order) (“[A]ny strict dichotomy between expletives and descriptions or depictions of sexual or excretory functions is artificial and does not make sense in light of the fact that an expletive’s power to offend derives from its sexual or excretory meaning” (internal quotation marks omitted)). The customs of speech refute this claim: There is a critical distinction between the use of an expletive to describe a sexual or excretory function and the use of such a word for an entirely different purpose, such as to express an emotion. One rests at the core of indecency; the other stands miles apart. As any golfer who has watched his partner shank a short approach knows, it would be absurd to accept the suggestion that the resultant four-letter word uttered on the golf course describes sex or excrement and is therefore indecent. But that is the absurdity the FCC has embraced in its new approach to indecency.4 See In re Complaints Against Vari*544ous Broadcast Licensees Regarding Their Airing of the “Golden Globe Awards” Program, 19 FCC Rcd. 4975, 4978-4979, ¶¶8-9 (2004) (declaring that even the use of an expletive to emphasize happiness “invariably invokes a coarse sexual image”).

Even if the words that concern the Court in this case sometimes retain their sexual or excretory meaning, there are surely countless instances in which they are used in a manner unrelated to their origin. These words may not be polite, but that does not mean they are necessarily “indecent” under §1464. By improperly equating the two, the Commission has adopted an interpretation of “indecency” that bears no resemblance to what Pacifica contemplated.5 Most distressingly, the Commission appears to be entirely unaware of this fact, see Remand Order, 21 FCC Red., at 13308 (erroneously referencing Pacifica in support of its new policy), and today's majority seems untroubled by this significant oversight, see ante, at 508-510, 517-518. Because the FCC has failed to demonstrate an awareness that it has ventured far beyond Pacifica's reading of § 1464, its policy choice must be declared arbitrary and set aside as unlawful. See Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U. S. 402, 416 (1971).

Ill

For these reasons and those stated in Justice Breyer’s dissenting opinion, I would affirm the judgment of the Court of Appeals.

Section 1464 provides: “Whoever utters any obscene, indecent, or profane language by means of radio communication shall be fined under this title or imprisoned not more than two years, or both.”

Justice Scalia erroneously concludes that treating the FCC’s rule-making authority as an exercise of legislative power would somehow be unconstitutional. See ante, at 524 (citing Bowsher v. Synar, 478 U. S. 714, *541726 (1986)). But that is the nature of rulemaking: Rules promulgated by agencies (independent or not) carry the force of law precisely because they are exercises of such legislative authority. This may offend Justice Scalia’s theory of the “unitary Executive,” ante, at 526, but it does not offend the Constitution. Indeed, “the Framers vested ‘All legislative Powers’ in the Congress, Art. I, § 1, just as in Article II they vested the ‘executive Power’ in the President, Art. II, § 1. Those provisions do not purport to limit the authority of either recipient of power to delegate authority to others.” Whitman v. American Trucking Assns., Inc., 531 U. S. 457, 489 (2001) (Stevens, J., concurring in part and concurring in judgment).

It appears that Justice Scalia has come to the view that isolated statements by members of a congressional oversight subcommittee are sufficient evidence of Congress’ intent. See ante, at 523-524, n. 4. Delving into the details of how various lawmakers “grilled” the foil slate of FCC Commissioners, Justice Scalia concludes, quite remarkably, that this encounter “made clear [Congress’] wishes for stricter enforcement” and “would seem an adequate explanation of [the FCC’s] change of posi*542tion.” Ante, at 524-525. Putting to the side the question whether congressional outrage is the kind of evidence sufficient to explain the Commission's decision to adopt a thinly reasoned and unconstitutional policy, Justice Scaua’s treatment of these proceedings as evidencing the intent of Congress would make even the most ardent student of legislative history blush.

It is ironic, to say the least, that while the FCC patrols the airwaves for words that have a tenuous relationship with sex or excrement, commercials broadcast during prime-time hours frequently ask viewers whether they too are battling erectile dysfunction or are having trouble going to the bathroom.

While Justice Thomas and I disagree about the continued wisdom of Pacifica, see ante, p. 530 (concurring opinion), the changes in technology and the availability of broadcast spectrum he identifies certainly counsel a restrained approach to indecency regulation, not the wildly expansive path the FCC has chosen.






Dissenting Opinion

Justice Ginsburg,

dissenting.

The mainspring of this case is a Government restriction on spoken words. This appeal, I recognize, arises under the *545Administrative Procedure Act (APA or Act).* Justice Breyer’s dissenting opinion, which I join, cogently describes the infirmities of the Federal Communications Commission’s (FCC or Commission) policy switch under that Act. The Commission’s bold stride beyond the bounds of FCC v. Pacifica Foundation, 438 U. S. 726 (1978), I agree, exemplified “arbitrary” and “capricious” decisionmaking. I write separately only to note that there is no way to hide the long shadow the First Amendment casts over what the Commission has done. Today’s decision does nothing to diminish that shadow.

More than 30 years ago, a sharply divided Court allowed the FCC to sanction a midafternoon radio broadcast of comedian George Carlin’s 12-minute “Filthy Words” monologue. Ibid. Carlin satirized the “original” seven dirty words and repeated them relentlessly in a variety of colloquialisms. The monologue was aired as part of a program on contemporary attitudes toward the use of language. In re Citizen’s Complaint Against Pacifica Foundation Station WBAI (FM), 56 F. C. C. 2d 94, 95 (1975). In rejecting the First Amendment challenge, the Court “emphasize[d] the narrowness of [its] holding.” Pacifica, 438 U. S., at 750. See also ante, at 539 (Stevens, J., dissenting). In this regard, the majority stressed that the Carlin monologue deliberately repeated the dirty words “over and over again.” 438 U. S., at 729, 751-755 (appendix). Justice Powell, concurring, described Carlin’s speech as “verbal shock treatment.” Id., at 757 (concurring in part and concurring in judgment).

*546In contrast, the unscripted fleeting expletives at issue here are neither deliberate nor relentlessly repetitive. Nor does the Commission’s policy home in on expressions used to describe sexual or excretory activities or organs. Spontaneous utterances used simply to convey an emotion or intensify a statement fall within the order’s compass. Cf. Cohen v. California, 403 U. S. 15, 26 (1971) (“[W]ords are often chosen as much for their emotive as their cognitive force. We cannot sanction the view that the Constitution, while solicitous of the cognitive content of individual speech, has little or no regard for that emotive function which, practically speaking, may often be the more important element of the overall message sought to be communicated.”); Denver Area Ed. Telecommunications Consortium, Inc. v. FCC, 518 U. S. 727, 805 (1996) (Kennedy, J., concurring in part, concurring in judgment in part, and dissenting in part) (a word categorized as indecent “often is inseparable from the ideas and viewpoints conveyed, or separable only with loss of truth or expressive power”).

The Pacifica decision, however it might fare on reassessment, see ante, at 535 (Thomas, J., concurring), was tightly cabined, and for good reason. In dissent, Justice Brennan observed that the Government should take care before enjoining the broadcast of words or expressions spoken by many “in our land of cultural pluralism.” 438 U. S., at 775. That comment, fitting in the 1970’s, is even more potent today. If the reserved constitutional question reaches this Court, see ante, at 529 (majority opinion), we should be mindful that words unpalatable to some may be “commonplace” for others, “the stuff of everyday conversations,” 438 U. S., at 776 (Brennan, J., dissenting).

Justice Breyer, with whom Justice Stevens, Justice Souter, and Justice Ginsburg join, dissenting.

In my view, the Federal Communications Commission failed adequately to explain why it changed its indecency pol*547icy from a policy permitting a single “fleeting use” of an expletive, to a policy that made no such exception. Its explanation fails to discuss two critical factors, at least one of which directly underlay its original policy decision. Its explanation instead discussed several factors well known to it the first time around, which by themselves provide no significant justification for a change of policy. Consequently, the FCC decision is “arbitrary, capricious, an abuse of discretion.” 5 U. S. C. §706(2)(A); Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U. S. 29, 41-43 (1983); Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U. S. 402, 420-421 (1971). And I would affirm the Second Circuit’s similar determination.

I

I begin with applicable law. That law grants those in charge of independent administrative agencies broad authority to determine relevant policy. But it does not permit them to make policy choices for purely political reasons nor to rest them primarily upon unexplained policy preferences. Federal Communications Commissioners have fixed terms of office; they are not directly responsible to the voters; and they enjoy an independence expressly designed to insulate them, to a degree, from “‘the exercise of political oversight.’” Freytag v. Commissioner, 501 U. S. 868, 916 (1991) (Scalia, J., concurring in part and concurring in judgment); see also Morrison v. Olson, 487 U. S. 654, 691, n. 30 (1988). That insulation helps to secure important governmental objectives, such as the constitutionally related objective of maintaining broadcast regulation that does not bend too readily before the political winds. But that agency’s comparative freedom from ballot-box control makes it all the more important that courts review its decisionmaking to assure compliance with applicable provisions of the law — including law requiring that major policy decisions be based upon articulable reasons.

*548The statutory provision applicable here is the Administrative Procedure Act’s (APA) prohibition of agency action that is “arbitrary, capricious, [or] an abuse of discretion,” 5 U. S. C. § 706(2)(A). This legal requirement helps assure agency decisionmaking based upon more than the personal preferences of the decisionmakers. Courts have applied the provision sparingly, granting agencies broad policymaking leeway. But they have also made clear that agency discretion is not “ ‘unbounded.’ ” Burlington Truck Lines, Inc. v. United States, 371 U. S. 156, 167-168 (1962). In so holding, American courts have followed a venerable legal tradition, stretching back at least to the days of Sir Edward Coke and the draining of the English fens. See Rooke’s Case, 77 Eng. Rep. 209, 210, 5 Coke Rep. 99b, 100a (C. P. 1598) (Coke, J.) (members of sewer commission with authority to act according “to their discretio[n]” are nonetheless “limited and bound with the rule of reason and law... and [cannot act] according to their wills and private affections” (quoted in Jaffe, Judicial Review: Constitutional and Jurisdictional Fact, 70 Harv. L. Rev. 953, 954 (1957))).

The law has also recognized that it is not so much a particular set of substantive commands but rather it is a process, a process of learning through reasoned argument, that is the antithesis of the “arbitrary.” This means agencies must follow a “logical and rational” decisionmaking “process.” Allentown Mack Sales & Service, Inc. v. NLRB, 522 U. S. 359, 374 (1998). An agency’s policy decisions must reflect the reasoned exercise of expert judgment. See Burlington Truck Lines, supra, at 167 (decision must reflect basis on which agency “exercised its expert discretion”); see also Humphrey’s Executor v. United States, 295 U. S. 602, 624 (1935) (independent agencies “exercise . . . trained judgment . . . ‘informed by experience’”). And, as this Court has specified, in determining whether an agency’s policy choice was “arbitrary,” a reviewing court “must consider whether the decision was based on a consideration of the *549relevant factors and whether there has been a clear error of judgment.” Overton Park, supra, at 416.

Moreover, an agency must act consistently. The agency must follow its own rules. Arizona Grocery Co. v. Atchison, T. & S. F. R. Co., 284 U. S. 370, 389-390 (1932). And when an agency seeks to change those rules, it must focus on the fact of change and explain the basis for that change. See, e. g., National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U. S. 967, 981 (2005) (“Unexplained inconsistency is” a “reason for holding an interpretation to be an arbitrary and capricious change from agency practice” (emphasis added)).

To explain a change requires more than setting forth reasons why the new policy is a good one. It also requires the agency to answer the question, “Why did you change?” And a rational answer to this question typically requires a more complete explanation than would prove satisfactory were change itself not at issue. An (imaginary) administrator explaining why he chose a policy that requires driving on the right side, rather than the left side, of the road might say, “Well, one side seemed as good as the other, so I flipped a coin.” But even assuming the rationality of that explanation for an initial choice, that explanation is not at all rational if offered to explain why the administrator changed driving practice, from right side to left side, 25 years later.

In State Farm, a unanimous Court applied these commonsense requirements to an agency decision that rescinded an earlier agency policy. The Court wrote that an agency must provide an explanation for the agency’s “revocation” of a prior action that is more thorough than the explanation necessary when it does not act in the first instance. The Court defined “revocation,” not simply as rescinding an earlier policy, cf. ante, at 514-515, but as “a reversal of the agency’s former views as to the proper course,” State Farm, 463 U. S., at 41 (emphasis added). See also Verizon Communications Inc. v. FCC, 535 U. S. 467, 502, n. 20 (2002) (portion of Court’s *550opinion joined by Scalia, Kennedy, and Thomas, JJ.) (noting State Farm “may be read as prescribing more searching judicial review” when “an agency [is] ‘changing its course’ as to the interpretation of a statute”); Thomas Jefferson Univ. v. Shalala, 512 U. S. 504, 524, n. 3 (1994) (Thomas, J., dissenting) (similar).

At the same time, the Court described the need for explanation in terms that apply, not simply to pure rescissions of earlier rules, but rather to changes of policy as it more broadly defined them. But see ante, at 514-515. It said that the law required an explanation for such a change because the earlier policy, representing a “‘settled course of behaviorf,] embodies the agency’s informed judgment that, by pursuing that course, it will carry out the policies ... best if the settled rule is adhered to.’” State Farm, supra, at 41-42. Thus, the agency must explain why it has come to the conclusion that it should now change direction. Why does it now reject the considerations that led it to adopt that initial policy? What has changed in the world that offers justification for the change? What other good reasons are there for departing from the earlier policy?

Contrary to the majority’s characterization of this dissent, it would not (and State Farm does not) require a “heightened standard” of review. Ante, at 514 (emphasis added). Rather, the law requires application of the same standard of review to different circumstances, namely, circumstances characterized by the fact that change is at issue. It requires the agency to focus upon the fact of change where change is relevant, just as it must focus upon any other relevant circumstance. It requires the agency here to focus upon the reasons that led the agency to adopt the initial policy, and to explain why it now comes to a new judgment.

I recognize that sometimes the ultimate explanation for a change may have to be, “We now weigh the relevant considerations differently.” But at other times, an agency can and should say more. "Where, for example, the agency rested its *551previous policy on particular factual findings, see ante, at 537-538 (Kennedy, J., concurring in part and concurring in judgment); or where an agency rested its prior policy on its view of the governing law, see infra, at 553-556; or where an agency rested its previous policy on, say, a special need to coordinate with another agency, one would normally expect the agency to focus upon those earlier views of fact, of law, or of policy and explain why they are no longer controlling. Regardless, to say that the agency here must answer the question “why change” is not to require the agency to provide a justification that is “better than the reasons for the old [policy].” Ante, at 515 (majority opinion). It is only to recognize the obvious fact that change is sometimes (not always) a relevant background feature that sometimes (not always) requires focus (upon prior justifications) and explanation lest the adoption of the new policy (in that circumstance) be “arbitrary, capricious, an abuse of discretion.”

That is certainly how courts of appeals, the courts that review agency decisions, have always treated the matter in practice. See, e. g., Pennsylvania Federation of Sportsmen’s Clubs, Inc. v. Kempthorne, 497 F. 3d 337, 351 (CA3 2007); Yale-New Haven Hosp. v. Leavitt, 470 F. 3d 71, 79 (CA2 2006); Citizens Awareness Network, Inc. v. United States, 391 F. 3d 338, 352 (CA1 2004). But see NAACP v. FCC, 682 F. 2d 993, 998 (CADC 1982) (using word “heightened”). The majority’s holding could in this respect significantly change judicial review in practice, and not in a healthy direction. But see ante, at 535-539 (Kennedy, J., concurring in part and concurring in judgment). After all, if it is always legally sufficient for the agency to reply to the question “why change?” with the answer “we prefer the new policy” (even when the agency has not considered the major factors that led it to adopt its old policy), then why bother asking the agency to focus on the fact of change? More to the point, why would the law exempt this and no other aspect of an agency decision from “arbitrary, capricious” re*552view? Where does, and why would, the APA grant agencies the freedom to change major policies on the basis of nothing more than political considerations or even personal whim?

Avoiding the application of any heightened standard of review, the Court in State Farm recognized that the APA’s “nonarbitrary” requirement affords agencies generous leeway when they set policy. 463 U. S., at 42. But it also recognized that this leeway is not absolute. The Court described its boundaries by then listing considerations that help determine whether an explanation is adequate. Mirroring and elaborating upon its statement in Overton Park, 401 U. S. 402, the Court said that a reviewing court should take into account whether the agency had “relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” State Farm, supra, at 43; see also Overton Park, supra, at 416.

II

We here must apply the general standards set forth in State Farm and Overton Park to an agency decision that changes a 25-year-old “fleeting expletive” policy from (1) the old policy that would normally permit broadcasters to transmit a single, fleeting use of an expletive to (2) a new policy that would threaten broadcasters with large fines for transmitting even a single use (including its use by a member of the public) of such an expletive, alone with nothing more. The question is whether that decision satisfies the minimal standards necessary to assure a reviewing court that such a change of policy is not “arbitrary, capricious, [or] an abuse of discretion,” 5 U. S. C. § 706(2)(A), particularly as set forth in, e. g., State Farm and Overton Park, supra, at 548-551 and this page. The decision, in my view, does not satisfy those standards.

*553Consider the requirement that an agency at least minimally “consider . . . important aspect[s] of the problem.” State Farm, supra, at 43. The FCC failed to satisfy this requirement, for it failed to consider two critically important aspects of the problem that underlay its initial policy judgment (one of which directly, the other of which indirectly). First, the FCC said next to nothing about the relation between the change it made in its prior “fleeting expletive” policy and the First-Amendment-related need to avoid “censorship,” a matter as closely related to broadcasting regulation as is health to that of the environment. The reason that discussion of the matter is particularly important here is that the FCC had explicitly rested its prior policy in large part upon the need to avoid treading too close to the constitutional line.

Thirty years ago, the Court considered the location of that constitutional line. In FCC v. Pacifica Foundation, 438 U. S. 726 (1978), the Court reviewed an FCC decision forbidding the broadcast of a monologue that deliberately and repeatedly uttered the expletives here at issue more than 100 times in one hour at a time of day when children were likely to hear the broadcast. Id., at 739. The Court held that the FCC’s prohibition did not violate the First Amendment. But the Court divided 5 to 4. And two Members of the majority, Justices Powell and Blackmun, explicitly noted that the Court “does not speak to cases involving the isolated use of a potentially offensive word ... as distinguished from the verbal shock treatment administered by respondent here.” Id., at 760-761 (Powell, J., concurring in part and concurring in judgment) (emphasis added). This statement by two Members of the majority suggested that they could reach a different result, finding an FCC prohibition unconstitutional, were that prohibition aimed at the fleeting or single use of an expletive.

The FCC subsequently made clear that it thought that Justice Powell’s concurrence set forth a constitutional line *554that its indecency policy should embody. In 1978, the Commission wrote that the First Amendment “severely limit[s]” the Commission’s role in regulating indecency. It added that the Court, in Pacifica, had “relied ... on the repetitive occurrence of the ‘indecent’ words in question.” And it said that, in setting policy, it “intended] strictly to observe the narrowness of the Pacifica holding.” In re Application of WGBH Educ. Foundation, 69 F. C. C. 2d 1250, 1254, ¶ 10.

In 1983, the Commission again wrote that it understood the Court’s decision in Pacifica to rest on the “‘repetitive occurrence of the “indecent” words in question.’” And, again, the Commission explained that its regulation of fleeting or isolated offensive words would reflect Justice Powell’s understanding of the First Amendment’s scope. In re Application of Pacifica Foundation, 95 F. C. C. 2d 750, 760, ¶¶ 17-18. In 1987, the Commission once more explained that its “fleeting expletives” policy reflected the Court’s decision in Pacifica. It said that, under its policy, “speech that is indecent must involve more than an isolated use of an offensive word,” adding that “we believe that under the legal standards set forth in Pacifica, deliberate and repetitive use in a patently offensive manner is a requisite to a finding of indecency.” In re Pacifica Foundation, 2 FCC Red. 2698, 2699, ¶ 13 (emphasis added). In another order that same year, the Commission stated that “the First Amendment dictated] a careful and restrained approach with regard to review of matters involving broadcast programming”; it then explained, citing Pacifica, that “[s]peech that is indecent must involve more than the isolated use of an offensive word.” In re Infinity Broadcasting, 2 FCC Red. 2705, 2705, ¶¶ 6-7 (1987) (emphasis added). And in 2001, in giving the industry guidance, the FCC once again said in respect to its regulation of indecent speech that it “must both identify a compelling interest for any regulation . . . and choose the least restrictive means to further that interest.” In re Industry Guidance on Commission’s Case Law Interpreting *55518 U. S. C. § 1464 and Enforcement Policies Regarding Broadcast Indecency, 16 FCC Red. 7999, 8000-8001, ¶¶ 3-5.

The FCC thus repeatedly made clear that it based its “fleeting expletive” policy upon the need to avoid treading too close to the constitutional line as set forth in Justice Powell’s Pacifica concurrence. What then did it say, when it changed its policy, about why it abandoned this Constitution-based reasoning? The FCC devoted “four full pages of small-type, single-spaced text,” ante, at 526 (majority opinion), responding to industry arguments that, e.g., changes in the nature of the broadcast industry made all indecency regulation, i. e., 18 U. S. C. § 1464, unconstitutional. In doing so it repeatedly reaffirmed its view that Pacifica remains good law. In re Complaints Regarding Various Television Broadcasts Between Feb. 2, 2002, and Mar. 8, 2005, 21 FCC Red. 13299, 13317-13321, ¶¶ 43-52 (2006) (Remand Order). All the more surprising then that, in respect to why it abandoned its prior view about the critical relation between its prior fleeting expletive policy and Justice Powell’s Pacifica concurrence, it says no more than the following: “[0]ur decision is not inconsistent with the Supreme Court ruling in Pacifica. The Court explicitly left open the issue of whether an occasional expletive could be considered indecent.” In re Complaints Against Various Broadcast Licensees Regarding Their Airing of the “Golden Globe Awards” Program, 19 FCC Red. 4975, 4982, ¶ 16 (2004) (Golden Globe Order). And (repeating what it already had said), “[Pacifica] specifically reserved the question of ‘an occasional expletive’ and noted that it addressed only the ‘particular broadcast’ at issue in that case.” Remand Order, supra, at 13308-13309, ¶ 24.

These two sentences are not a summary of the FCC’s discussion about why it abandoned its prior understanding of Pacifica. They are the discussion. These 28 words (repeated in two opinions) do not acknowledge that an entirely different understanding of Pacifica underlay the FCC’s ear*556lier policy; they do not explain why the agency changed its mind about the line that Pacifica draws or its policy’s relation to that line; and they tell us nothing at all about what happened to the FCC’s earlier determination to search for “compelling interests” and “less restrictive alternatives.” They do not explain the transformation of what the FCC had long thought an insurmountable obstacle into an open door. The result is not simply Hamlet without the prince, but Hamlet with a prince who, in midplay and without explanation, just disappears.

I have found one other related reference to Pacifica, but that reference occurs in an opinion written by a dissenting Commissioner. That dissenter said that the FCC had “ Tail[ed] to address the many serious [constitutional] concerns raised’ ” by the new policy, while adding that the new policy was “not the restrained enforcement policy encouraged by the Supreme Court in Pacifica.” Remand Order, supra, at 13331, 13334. Neither that Commissioner in his dissent, nor I in this dissent, claim that agencies must always take account of possible constitutional issues when they formulate policy. Cf. ante, at 516 (majority opinion). But the FCC works in the shadow of the First Amendment, and its view of the application of that Amendment to “fleeting expletives” directly informed its initial policy choice. Under these circumstances, the FCC’s failure to address this “aspect” of the problem calls for a remand to the agency. Overton Park, 401 U. S., at 420-421.

Second, the FCC failed to consider the potential impact of its new policy upon local broadcasting coverage. This “aspect of the problem” is particularly important because the FCC explicitly took account of potential broadcasting impact. Golden Globe Order, supra, at 4980, ¶ 11 (“The ease with which broadcasters today can block even fleeting words in a live broadcast is an element in our decision”). Indeed, in setting forth “bleeping” technology changes (presumably lowering bleeping costs) as justifying the policy change, it *557implicitly reasoned that lower costs, making it easier for broadcasters to install bleeping equipment, made it less likely that the new policy would lead broadcasters to reduce coverage, say, by canceling coverage of public events. Ibid. (“ [Technological advances have made it possible ... to prevent the broadcast of a single offending word or action without blocking or disproportionately disrupting the message of the speaker or performer”).

What then did the FCC say about the likelihood that smaller independent broadcasters, including many public service broadcasters, still would not be able to afford “bleeping” technology and, as a consequence, would reduce local coverage, indeed cancel coverage, of many public events? It said nothing at all.

The FCC cannot claim that local coverage lacks special importance. To the contrary, “the concept of localism has been a cornerstone of broadcast regulation for decades.” In re Broadcast Localism, 23 FCC Red. 1324, 1326, 1327, ¶ ¶ 3, 5 (2008). That policy seeks to provide “viewers and listeners . . . access to locally responsive programming including, but not limited to, local news and public affairs matter” id., at 1326, ¶ 3, and to ensure “diversity in what is seen and heard over the airwaves,” ibid. That policy has long favored local broadcasting, both as a means to increase coverage of local events and, insofar as it increases the number of broadcast voices, as an end in itself. See, e. g., In re Reexamination of Comparative Standards for Noncommercial Educ. Applicants, 15 FCC Red. 7386, 7399, ¶29 (2000) (adopting a system for selecting applicants for broadcast channels that “would foster our goal of broadcast diversity by enabling the local public to be served by differing . . . licensees”); In re 2002 Biennial Regulatory Review, 18 FCC Red. 13620, 13644, ¶¶77, 79 (2003) (“We remain firmly committed to the policy of promoting localism among broadcast outlets. . . . A . . . measure of localism is the quantity and quality of local news and public affairs programming”).

*558Neither can the FCC now claim that the impact of its new policy on local broadcasting is insignificant and obviously so. Broadcasters tell us, as they told the FCC, the contrary. See Brief for Former FCC Commissioners et al. as Amici Curiae 17-19; App. 235-237; Joint Comments of Fox Television Stations, Inc., et al., In re Remand of Section IH.B of Commission’s Mar. 15, 2006 Omnibus Order Resolving Numerous Broadcast Television Indecency Complaints 14-15, http://www.fcc.gov/BA06-1739/joint-networks.pdf (all Internet materials as visited Apr. 7, 2009, and available in Clerk of Court’s case file). They told the FCC, for example, that the costs of bleeping/delay systems, up to $100,000 for installation and annual operation, place that technology beyond the financial reach of many smaller independent local stations. See id., at 14 (“The significant equipment and personnel costs associated with installing, maintaining, and operating delay equipment sufficient to cover all live news, sports, and entertainment programs could conceivably exceed the net profits of a small local station for an entire year”); id., at App. XI. And they ask what the FCC thinks will happen when a small local station without bleeping equipment wants to cover, say, a local city council meeting, a high school football game, a dance contest at a community center, or a Fourth of July parade.

Relevant literature supports the broadcasters’ financial claims. See, e. g., Ho, Taking No Chances, Austin American-Statesman, June 18, 2006, p. Jl; Dotinga, Dirty-Word Filters Prove Costly, Wired.com, July 9, 2004, http://www.wired.com/entertainment/music/news/2004/ 07/64127; Stations, Cable Networks Finding Indecency Rules Expensive, Public Broadcasting Report, Aug. 4, 2006. It also indicates that the networks with which some small stations are affiliated are not liable for the stations’ local transmissions (unless the networks own them). Ho, supra, at Jl; Public Stations Fear Indecency Fine Jump Means Premium Hikes, Public Broadcasting Report, July 7,2006. The result *559is that smaller stations, fearing “fleeting expletive” fines of up to $325,000, may simply cut back on their coverage. See Romano, Reporting Live. Very Carefully, Broadcasting & Cable, July 4, 2005, p. 8; see also ibid. (“Afraid to take chances” of getting fined under the FCC’s new policy, “local broadcasters are responding by altering — or halting altogether — the one asset that makes local stations so valuable to their communities: live TV”); Daneman, WRUR Drops Its Live Radio Programs, Rochester Democrat and Chronicle, May 27, 2004, p. IB (reporting that a local broadcast station ceased broadcasting all local live programming altogether in response to the Commission’s policy change). And there are many such smaller stations. See, e. g., Corporation for Public Broadcasting, Frequently Asked Questions, available at http://www.cpb.org/aboutpb/faq/stations.html (noting there are over 350 local public television stations and nearly 700 local public radio stations that receive support from the Corporation for Public Broadcasting).

As one local station manager told the FCC:

“To lessen the risk posed by the new legal framework ... I have directed [the station’s] news staff that [our station] may no longer provide live, direct-to-air coverage” of “live events where crowds are present. .. unless they affect matters of public safety or convenience. Thus, news coverage by [my station] of live events where crowds are present essentially will be limited to civil emergencies.” App. 236-237 (declaration of Dennis Fisher).

What did the FCC say in response to this claim? What did it say about the likely impact of the new policy on the coverage that its new policy is most likely to affect, coverage of local live events — city council meetings, local sports events, community arts productions, and the like? It said nothing at all.

*560The plurality acknowledges that the Commission entirely failed to discuss this aspect of the regulatory problem. But it sees “no need” for discussion in light of its, i. e., the plurality’s, own “doubt[s]” that “small-town broadcasters run a heightened risk of liability for indecent utterances” as a result of the change of policy. Ante, at 527. The plurality’s “doubt[sj” rest upon its views (1) that vulgar expression is less prevalent (at least among broadcast guests) in smaller towns, ibid.; (2) that the greatest risk the new policy poses for “small-town broadcasters” arises when they broadcast local “news and public affairs,” ibid.; and (3) that the Remand Order says “little about how the Commission would treat smaller broadcasters who cannot afford screening equipment,” while also pointing out that the new policy “ ‘does not... impose undue burdens on broadcasters’ ” and emphasizing that the case before it did not involve “ ‘breaking news,’ ” ante, at 528, 527.

As to the first point, about the prevalence of vulgarity in small towns, I confess ignorance. But I do know that there are independent stations in many large and medium sized cities. See Television & Cable Factbook, Directory of Television Stations in Operation 2008. As to the second point, I too believe that coverage of local public events, if not news, lies at the heart of the problem.

I cannot agree with the plurality, however, about the critical third point, namely, that the new policy obviously provides smaller independent broadcasters with adequate assurance that they will not be fined. The new policy removes the “fleeting expletive” exception, an exception that assured smaller independent stations that they would not be fined should someone swear at a public event. In its place, it puts a policy that places all broadcasters at risk when they broadcast fleeting expletives, including expletives uttered at public events. The Remand Order says that there “is no outright news exemption from our indecency rules.” 21 FCC Red., at 13327, ¶ 71 (emphasis added). The best it can pro*561vide by way of assurance is to say that “it may be inequitable to hold a licensee responsible for airing offensive speech during live coverage of a public event under some circumstances.” Id., at 13311, ¶33 (emphasis added). It does list those circumstances as including the “possibility of human error in using delay equipment.” Id., at 13313, ¶ 35. But it says nothing about a station’s inability to afford delay equipment (a matter that in individual cases could itself prove debatable). All the FCC had to do was to consider this matter and either grant an exemption or explain why it did not grant an exemption. But it did not. And the result is a rule that may well chill coverage — the kind of consequence that the law has considered important for decades, to which the broadcasters pointed in their arguments before the FCC, and which the FCC nowhere discusses. See, e. g., Dombrowski v. Pfister, 380 U. S. 479, 494 (1965) (“So long as the statute remains available to the State the threat of prosecutions of protected expression is a real and substantial one. Even the prospect of ultimate failure of such prosecutions by no means dispels their chilling effect on protected expression”); see also Ashcroft v. Free Speech Coalition, 535 U. S. 234, 244 (2002); Gibson v. Florida Legislative Investigation Comm., 372 U. S. 539, 556-557 (1963); Wieman v. Updegraff 344 U. S. 183, 195 (1952) (Frankfurter, J., concurring).

Had the FCC used traditional administrative notice-and-comment procedures, 5 U. S. C. § 553, the two failures I have just discussed would clearly require a court to vacate the resulting agency decision. See ACLU v. FCC, 823 F. 2d 1554, 1581 (CADC 1987) (per curiam) (“Notice and comment rulemaking procedures obligate the FCC to respond to all significant comments, for the opportunity to comment is meaningless unless the agency responds to significant points raised by the public” (emphasis added; internal quotation marks omitted)). Here the agency did not make new policy through the medium of notice-and-comment proceedings. *562But the same failures here — where the policy is important, the significance of the issues clear, the failures near complete — should lead us to the same conclusion. The agency’s failure to discuss these two “important aspect[s] of the problem” means that the resulting decision is “ ‘arbitrary, capricious, an abuse of discretion’” requiring us to remand the matter to the agency. State Farm, 463 U. S., at 43; Overton Park, 401 U. S., at 416.

Ill

The three reasons the FCC did set forth in support of its change of policy cannot make up for the failures I have discussed. Consider each of them. First, as I have pointed out, the FCC based its decision in part upon the fact that “bleeping/delay systems” technology has advanced. I have already set forth my reasons for believing that that fact, without more, cannot provide a sufficient justification for its policy change. Supra, at 556-561 and this page.

Second, the FCC says that the expletives here in question always invoke a coarse excretory or sexual image; hence it makes no sense to distinguish between whether one uses the relevant terms as an expletive or as a literal description. The problem with this answer is that it does not help to justify the change in policy. The FCC was aware of the coarseness of the “image” the first time around. See, e. g., Remand Order, supra, at 13308, ¶23 (asserting that FCC has always understood the words as coarse and indecent). And it explained the first time around why it nonetheless distinguished between their literal use and their use as fleeting expletives. See, e.g., In re Application of WGBH Educ. Foundation, 69 F. C. C. 2d, at 1254-1255, ¶¶ 10-11 (discussing First Amendment considerations and related need to avoid reduced broadcast coverage). Simply to announce that the words, whether used descriptively or as expletives, call forth similar “images” is not to address those reasons.

*563Third, the FCC said that “perhaps” its “most importan[t]” justification for the new policy lay in the fact that its new “contextual” approach to fleeting expletives is better and more “[consistent with” the agency’s “general approach to indecency” than was its previous “categorica[l]” approach, which offered broadcasters virtual immunity for the broadcast of fleeting expletives. Remand Order, 21 FCC Red., at 13308, ¶ 23. This justification, however, offers no support for the change without an understanding of why, i. e., in what way, the FCC considered the new approach better or more consistent with the agency’s general approach.

The Solicitor General sets forth one way in which the new policy might be more consistent with statutory policy. The indecency statute prohibits the broadcast of “any . . . indecent. .. language.” 18 U. S. C. § 1464. The very point of the statute, he says, is to eliminate nuisance; and the use of expletives, even once, can constitute such a nuisance. The Solicitor General adds that the statutory word “any” indicates that Congress did not intend a safe harbor for a fleeting use of that language. Brief for Petitioners 24-25. The fatal flaw in this argument, however, lies in the fact that the Solicitor General and not the agency has made it. We must consider the lawfulness of an agency’s decision on the basis of the reasons the agency gave, not on the basis of those it might have given. SEC v. Chenery Corp., 332 U. S. 194, 196-197 (1947); State Farm, supra, at 50. And the FCC did not make this claim. Hence, we cannot take it into account and need not evaluate its merits.

In fact, the FCC found that the new policy was better in part because, in its view, the new policy better protects children against what it described as “ ‘the first blow’ ” of broadcast indecency that results from the “ ‘pervasive’ ” nature of broadcast media. It wrote that its former policy of “granting an automatic exemption for ‘isolated or fleeting’ expletives unfairly forces viewers (including children) to take ‘the first blow.’ ” Remand Order, supra, at 13309, ¶ 25.

*564The difficulty with this argument, however, is that it does not explain the change. The FCC has long used the theory of the “first blow” to justify its regulation of broadcast indecency. See, e. g., In re Enforcement of Prohibitions Against Broadcast Indecency in 18 U. S. C. §1464, 5 FCC Red. 5297, 5301-5302, ¶¶ 34-35 (1990). Yet the FCC has also long followed its original “fleeting expletives” policy. Nor was the FCC ever unaware of the fact to which the majority points, namely, that children’s surroundings influence their behavior. See, e. g., In re Enforcement of Prohibitions Against Broadcast Indecency in 18 U ,S. C. § 1464,, 8 FCC Rcd. 704, 705-706, ¶ 11 (1993). So, to repeat the question: What, in respect to the “first blow,” has changed?

The FCC points to no empirical (or other) evidence to demonstrate that it previously understated the importance of avoiding the “first blow.” Like the majority, I do not believe that an agency must always conduct full empirical studies of such matters. Ante, at 519-520. But the FCC could have referred to, and explained, relevant empirical studies that suggest the contrary. One review of the empirical evidence, for example, reports that “[i]t is doubtful that children under the age of 12 understand sexual language and innuendo; therefore it is unlikely that vulgarities have any negative effects.” Kaye & Sapolsky, Watch Your Mouth! An Analysis of Profanity Uttered by Children on Prime-Time Television, 2004 Mass Communication & Soc’y 429, 433 (Vol. 7) (citing two studies). The Commission need not have accepted this conclusion. But its failure to discuss this or any other such evidence, while providing no empirical evidence at all that favors its position, must weaken the logical force of its conclusion. See State Farm, supra, at 43 (explaining that an agency’s failure to “examine the relevant data” is a factor in determining whether the decision is “arbitrary”).

The FCC also found the new policy better because it believed that its prior policy “would as a matter of logic permit broadcasters to air expletives at all hours of a day so long as *565they did so one at a time.” Remand Order, 21 FCC Red., at 13309, ¶ 25. This statement, however, raises an obvious question: Did that happen? The FCC’s initial “fleeting expletives” policy was in effect for 25 years. Had broadcasters during those 25 years aired a series of expletives “one at a time”? If so, it should not be difficult to find evidence of that fact. But the FCC refers to none. Indeed, the FCC did not even claim that a change had taken place in this respect. It spoke only of the pure “logic” of the initial policy “permitting” such a practice. That logic would have been apparent to anyone, including the FCC, in 1978 when the FCC set forth its initial policy.

Finally, the FCC made certain statements that suggest it did not believe it was changing prior policy in any major way. It referred to that prior policy as based on “staff letters and dicta” and it said that at least one of the instances before it (namely, the Cher broadcast) would have been actionably indecent under that prior policy. Id., at 13306-13307, 13324, ¶¶ 20-21, 60. As we all agree, however, in fact the FCC did change its policy in a major way. See ante, at 517 (majority opinion). To the extent that the FCC minimized that fact when considering the change, it did not fully focus on the fact of change. And any such failure would make its decision still less supportable. See National Cable, 545 U. S., at 981.

IV

Were the question a closer one, the doctrine of constitutional avoidance would nonetheless lead me to remand the ease. See United States v. Jin Fuey Moy, 241 U. S. 394, 401 (1916) (“A statute must be construed, if fairly possible, so as to avoid not only the conclusion that it is unconstitutional but also grave doubts upon that score” (emphasis added)). That doctrine seéks to avoid unnecessary judicial consideration of constitutional questions, assumes that Congress, no less than the Judicial Branch, seeks to act within constitutional bounds, and thereby diminishes the friction between *566the branches that judicial holdings of unconstitutionally might otherwise generate. See Almendarez-Torres v. United States, 523 U. S. 224, 237-238 (1998); see also Solid Waste Agency of Northern Cook Cty. v. Army Corps of Engineers, 531 U. S. 159, 172-173 (2001); Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Constr. Trades Council, 485 U. S. 568, 575 (1988); Rescue Army v. Municipal Court of Los Angeles, 331 U. S. 549, 571 (1947); Ashwander v. TVA, 297 U. S. 288, 345-348 (1936) (Brandeis, J., concurring). The doctrine assumes that Congress would prefer a less-than-optimal interpretation of its statute to the grave risk of a constitutional holding that would set the statute entirely aside. See Almendarez-Torres, supra, at 238 (construction of statute that avoids invalidation best reflects congressional will); cf. United States v. Booker, 543 U. S. 220, 249, 267 (2005).

Unlike the majority, I can find no convincing reason for refusing to apply a similar doctrine here. The Court has often applied that doctrine where an agency’s regulation relies on a plausible but constitutionally suspect interpretation of a statute. See, e. g., Solid Waste Agency, supra, at 172-174; NLRB v. Catholic Bishop of Chicago, 440 U. S. 490, 506-507 (1979). The values the doctrine serves apply whether the agency’s decision does, or does not, rest upon a constitutionally suspect interpretation of a statute. And a remand here would do no more than ask the agency to reconsider its policy decision in light of the concerns raised in a judicial opinion. Cf. Fullilove v. Klutznick, 448 U. S. 448, 551 (1980) (Stevens, J., dissenting) (a holding that a congressional action implicating the Equal Protection Clause “was not adequately preceded by a consideration of less drastic alternatives or adequately explained by a statement of legislative purpose would be far less intrusive than a final determination that the substance of” that action was unconstitutional). I would not now foreclose, as the majority forecloses, our further consideration of this matter. (Of course, nothing in *567the Court's decision today prevents the Commission from reconsidering its current policy in light of potential constitutional considerations or for other reasons.)

V

In sum, the FCC’s explanation of its change leaves out two critically important matters underlying its earlier policy, namely, Pacifica and local broadcasting coverage. Its explanation rests upon three considerations previously known to the agency (“coarseness,” the “first blow,” and running single expletives all day, one at a time). With one exception, it provides no empirical or other information explaining why those considerations, which did not justify its new policy before, justify it now. Its discussion of the one exception (technological advances in bleeping/delay systems), failing to take account of local broadcast coverage, is seriously incomplete.

I need not decide whether one or two of these features, standing alone, would require us to remand the case. Here all come together. And taken together they suggest that the FCC’s answer to the question, “Why change?” is, “We like the new policy better.” This kind of answer, might be perfectly satisfactory were it given by an elected official. But when given by an agency, in respect to a major change of an important policy where much more might be said, it is not sufficient. State Farm, 463 U. S., at 41-42.

For these reasons I would find the FCC’s decision “arbitrary, capricious, an abuse of discretion,” 5 U. S. C. § 706(2)(A), requiring remand of this case to the FCC. And I would affirm the Second Circuit’s similar determination.

With respect, I dissent.

The Second Circuit, presented with both constitutional and statutory challenges, vacated the remand order on APA grounds. The court therefore “refrain[ed] from deciding” the “constitutional questions.” 489 F. 3d 444, 462 (2007) (quoting Lyng v. Northwest Indian Cemetery Protective Assn., 485 U. S. 439, 445 (1988)). The majority, however, stated and explained why it was “skeptical” that the Commission’s policy could “pass constitutional muster.” 489 F. 3d, at 462.