No. 95-1184


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

[June 25, 1997]

Justice Souter, with whom The Chief Justice and Justice Scalia join, and with whom Justice Thomas joins except as to Part II, dissenting.

The Court today finds no First Amendment right to be free of coerced subsidization of commercial speech, for two principal reasons. First, the Court finds no discernible element of speech in the implementation of the Government's marketing orders, beyond what it sees as "germane" to the undoubtedly valid, nonspeech elements of the orders. Second, the Court in any event takes the position that a person who is neither barred from saying what he wishes, nor subject to personal attribution of speech he dislikes, has no First Amendment objection to mandatory subsidization of speech unless it is ideological or political or contains a message with which the objecting person disagrees. I part company with the Court on each of these closely related points. The legitimacy of governmental regulation does not validate coerced subsidies for speech that the government cannot show to be reasonably necessary to implement the regulation, and the very reasons for recognizing that commercial speech falls within the scope of First Amendment protection likewise justifies the protection of those who object to subsidizing it against their will. I therefore conclude that forced payment for commercial speech should be subject to the same level of judicial scrutiny as any restriction on communications in that category. Because I believe that the advertising scheme here fails that test, I respectfully dissent.

The nub of the Court's opinion is its reading of the line of cases following Abood v. Detroit Bd. of Ed., 431 U.S. 209 (1977):

"Abood, and the cases that follow it, did not announce a broad First Amendment right not to be compelled to provide financial support for any organization that conducts expressive activities. Rather, Abood merely recognized a First Amendment interest in not being compelled to contribute to an organization whose expressive activities conflict with one's `freedom of belief.' " Ante, at 13 (quoting Abood, supra, at 235).

While I certainly agree with the Court that a proper understanding of Abood is necessary for the disposition of this case (and will dwell on the scope of its holding at some length below), it seems to me that Abood appears more readily in its proper size if we begin our analysis with two more basic principles of First Amendment law: that speech as such is subject to some level of protection unless it falls within a category, such as obscenity, placing it beyond the Amendment's scope, and that protected speech may not be made the subject of coercion to speak or coercion to subsidize speech.

Even before we first recognized commercial speech protection in Virginia Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748 (1976), we had stated a basic proposition of First Amendment protection, that "[a]ll ideas having even the slightest redeeming social importance . . . have the full protection of the guaranties [of the First Amendment]," Roth v. United States, 354 U.S. 476, 484 (1957). This premise was later echoed in Virginia Bd. of Pharmacy, where we asked whether commercial speech "is so removed from any exposition of ideas, and from truth, science, morality, and arts in general, in its diffusion of liberal sentiments on the administration of Government, that it lacks all protection." 425 U. S., at 762 (citations and internal quotation marks omitted). The answer, of course, was no.

What stood against the claim of social unimportance for commercial speech was not only the consumer's interest in receiving information, id., at 763-764, but the commercial speaker's own economic interest in promoting his wares. "[W]e may assume that the advertiser's interest is a purely economic one. That hardly disqualifies him from protection under the First Amendment." Id., at 762. Indeed, so long as self interest in providing a supply is as legitimate as the self interest underlying an informed demand, the law could hardly treat the advertiser's economic stake as "utterly without redeeming social importance" and isolate the consumer's interest as the exclusive touchstone of commercial speech protection.

Nor is the advertiser's legitimate interest one dimensional. While the value of a truthful representation of the product offered is central, advertising's persuasive function is cognizable, too. Like most advertising meant to stimulate demand, the promotions for California fruit at issue here do more than merely provide objective information about a product's availability or price; they exploit all the symbolic and emotional techniques of any modern ad campaign with messages often far removed from simple proposals to sell fruit. [n.1] "Speech has the capacity to convey complex substance, yielding various insights and interpretations depending upon the identity of the listener or the reader and the context of its transmission. . . . The complex nature of expression is one reason why even so called commercial speech has become an essential part of the public discourse the First Amendment secures." Florida Bar v. Went For It, Inc., 515 U.S. 618, 636 (1995) (Kennedy, J., dissenting). Since persuasion is an essential ingredient of the competition that our public law promotes with considerable effort, the rhetoric of advertising cannot be written off as devoid of value or beyond protection, any more than can its power to inform. Of course, that value may well be of a distinctly lower order than the importance of providing accurate factual information, and the inextricable linkage between advertising and underlying commercial transaction "may give [the government] a concomitant interest in the expression itself," Edenfield v. Fane, 507 U.S. 761, 767 (1993) (citation and internal quotation marks omitted); see also 44 Liquormart, Inc. v. Rhode Island, 517 U. S. ___, ___ (1996) (slip op., at 12) (opinion of Stevens, J.). But these considerations amount to nothing more than the premise justifying a merely moderate level of scrutiny for commercial speech regulations generally: "the `common sense' distinction between speech proposing a commercial transaction, which occurs in an area traditionally subject to government regulation, and other varieties of speech." Rubin v. Coors Brewing Co., 514 U.S. 476, 482 (1995) (citations and internal quotation marks omitted).

Since commercial speech is not subject to any categorical exclusion from First Amendment protection, and indeed is protectible as a speaker's chosen medium of commercial enterprise, it becomes subject to a second First Amendment principle: that compelling cognizable speech officially is just as suspect as suppressing it, and is typically subject to the same level of scrutiny. In Riley v. National Federation of Blind of N. C., Inc., 487 U.S. 781 (1988), for example, the State argued that "the First Amendment interest in compelled speech is different [from] the interest in compelled silence," and ought therefore to merit a more "deferential test." Id., at 796. We rejected that argument out of hand: "There is certainly some difference between compelled speech and compelled silence, but in the context of protected speech, the difference is without constitutional significance, for the First Amendment guarantees `freedom of speech,' a term necessarily comprising the decision of both what to say and what not to say." Id., at 796-797; see also Hurley v. Irish American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U.S. 557, 573 (1995) ("Since all speech inherently involves choices of what to say and what to leave unsaid, one important manifestation of the principle of free speech is that one who chooses to speak may also decide what not to say") (citations and internal quotation marks omitted); Wooley v. Maynard, 430 U.S. 705, 714 (1977) ("[T]he right of freedom of thought protected by the First Amendment against state action includes both the right to speak freely and the right to refrain from speaking at all"); West Virginia Bd. of Ed. v. Barnette, 319 U.S. 624, 633 (1943) ("[I]nvoluntary affirmation c[an] be commanded only on even more immediate and urgent grounds than silence").

As a familiar corollary to the principle that what may not be suppressed may not be coerced, we have recognized (thus far, outside the context of commercial speech) that individuals have a First Amendment interest in freedom from compulsion to subsidize speech and other expressive activities undertaken by private and quasi private organizations. [n.2] We first considered this issue in Abood v. Detroit Bd. of Ed., 431 U.S. 209 (1977), in addressing the First Amendment claims of dissenting employees subject to an "agency shop" agreement between their government employer and a union. The agreement required each employee to pay the union a "service fee" equal to the dues required of union members, but limited no one's right to speak separately and obliged no employee to join the union, personally espouse unionism, or participate in the union in any other way. Id., at 212. Thus, as in this case, the sole imposition upon nonmembers was the assessment to help pay for the union's activities. And yet, purely financial as the imposition was, we held that the union's use of dissenters' service fees for expressive purposes unrelated to collective bargaining violated the First Amendment rights of those employees. In so holding, Abood drew together several lines of First Amendment doctrine; after recognizing the parallels between expression per se and associating for expressive purposes, id., at 233-234, the Court relied on compelled speech cases such as Barnette, supra, in concluding that just as the government may not (without a compelling reason) prohibit a person from contributing money to propagate ideas, neither may it force an individual to contribute money to support some group's distinctly expressive activities, id., at 234-235. We have repeatedly adhered to this reasoning in cases of compelled contributions to unions in agency shops, see, e.g., Lehnert v. Ferris Faculty Assn., 500 U.S. 507 (1991); Teachers v. Hudson, 475 U.S. 292 (1986); Ellis v. Railway Clerks, 466 U.S. 435 (1984) (statutory case); Machinists v. Street, 367 U.S. 740 (1961) (statutory case anticipating Abood), and have followed the same rationale in holding that state compelled dues to an integrated bar association may not constitutionally be used to advance political and ideological causes distinct from the core objectives of professional regulation, Keller v. State Bar of Cal., 496 U.S. 1 (1990).

The Court recognizes the centrality of the Abood line of authority for resolving today's case, but draws the wrong conclusions from it. Since Abood struck down the mandatory "service fee" only insofar as it funded the union's expression of support for "ideological causes not germane to its duties as collective bargaining representative," 431 U. S., at 235; see also id., at 232, the Court reads Abood for the proposition that the First Amendment places no limits on government's power to force one individual to pay for another's speech, except when the speech in question both is ideological or political in character and is not germane to an otherwise lawful regulatory program. Ante, at 13-15. [n.3]

The Court's first mistaken conclusion lies in treating Abood as permitting any enforced subsidy for speech that is germane to permissible economic regulation, in the sense that it relates to the subject matter of the regulation and tends to further its objectives. But Abood and its subsequent line of cases is not nearly so permissive as the Court makes out. In Abood, we recognized that even in matters directly related to collective bargaining, compulsory funding of union activities has an impact on employees' First Amendment interests, since the employees might disagree with positions taken by the union on issues such as the inclusion of abortion in a medical benefit plan, or negotiating no strike agreements, or even the desirability of unionism in general. 431 U. S., at 222. To be sure, we concluded that any interference with such interests was "constitutionally justified by the legislative assessment of the important contribution of the union shop to the system of labor relations established by Congress." Ibid.; see also Keller, supra, at 13-14 ("[T]he State's interest in regulating the legal profession and improving the quality of legal services" justifies "the compelled association [inherent in the] integrated bar"). But this was simply a way of saying that the government's objective of guaranteeing the opportunity for a union shop, the importance and legitimacy of which were already settled, see Abood, 217-232 (following Railway Employes v. Hanson, 351 U.S. 225 (1956), and Machinists v. Street, 367 U.S. 740 (1961)), could not be attained without the incidental infringements of the interests in unfettered speech and association that petitioners there claimed. Collective bargaining, and related activities such as grievance arbitration and contract administration, are part and parcel of the very economic transactions between employees and employer that Congress can regulate, and which it could not regulate without these potential impingements on the employees' First Amendment interests. Abood is thus a specific instance of the general principle that government retains its full power to regulate commercial transactions directly, despite elements of speech and association inherent in such transactions. See Ohralik v. Ohio State Bar Assn., 436 U.S. 447, 456 (1978) (commercial conduct may be regulated without offending First Amendment despite use of language); Roberts v. United States Jaycees, 468 U.S. 609, 634 (1984) (opinion of O'Connor, J., concurring in part and concurring in judgment) (in contrast to right of expressive association, "there is only minimal constitutional protection of the freedom of commercial association," because "the State is free to impose any rational regulation on the commercial transaction itself"); see also New York State Club Assn., Inc. v. City of New York, 487 U.S. 1, 13 (1988) (constitutional right of expressive association is not implicated by every instance in which individuals choose their associates); Dallas v. Stanglin, 490 U.S. 19, 25 (1989) (same); Ellis v. Railway Clerks, 466 U.S. 435, 456 (1984) (funding of union social activities, as opposed to expressive activities, has minimal connection with First Amendment rights).

Decisions postdating Abood have made clear, however, that its limited sanction for laws affecting First Amendment interests may not be expanded to cover every imposition that is in some way "germane" to a regulatory program in the sense of relating sympathetically to it. Rather, to survive scrutiny under Abood, a mandatory fee must not only be germane to some otherwise legitimate regulatory scheme; it must also be justified by vital policy interests of the government and not add significantly to the burdening of free speech inherent in achieving those interests. Lehnert v. Ferris Faculty Assn., 500 U.S. 507, 519 (1991); accord Ellis, supra, at 456.

Thus, in Lehnert eight Justices concluded that a teachers' union could not constitutionally charge objecting employees for a public relations campaign meant to raise the esteem for teachers in the public mind and so increase the public's willingness to pay for public education. See 500 U. S., at 528-529 (plurality opinion); id., at 559 (Scalia, J., concurring in judgment in part and dissenting in part). "Expression of this kind extends beyond the negotiation and grievance resolution contexts and imposes a substantially greater burden upon First Amendment rights than do [collective bargaining functions]." Id., at 528-529 (plurality opinion). The advertising campaigns here suffer from the same defect as the public relations effort to stimulate demand for the teachers' product: a local union can negotiate a particular contract for the benefit of a shop's whole labor force without globally espousing the virtues of teachers, and (in the absence of further explanation) produce markets can be directly regulated in the interest of stability and growth without espousing the virtues of fruit. They were, indeed, for a quarter century, and still are under the many agricultural marketing orders that authorize no advertising schemes. See infra, at 19-25. In each instance, the challenged burden on dissenters' First Amendment rights is substantially greater than anything inherent in regulation of the commercial transactions. Thus, the Abood line does not permit this program merely because it is germane to the marketing orders. [n.4]

The Court's second misemployment of Abood and its successors is its reliance on them for the proposition that when government neither forbids speech nor attributes it to an objector, it may compel subsidization for any objectionable message that is not political or ideological. But this, of course, is entirely at odds with the principle that speech significant enough to be protected at some level is outside the government's power to coerce or to support by mandatory subsidy without further justification. Supra, at 5-7. Since a commercial speaker (who does not mislead) may generally promote commerce as he sees fit, the government requires some justification (such as its necessity for otherwise valid regulation) before it may force him to subsidize commercial speech to which he objects. While it is perfectly true that cases like Abood and Keller did involve political or ideological speech, and the Court made reference to that character in explaining the gravity of the First Amendment interests at stake, nothing in those cases suggests that government has free rein to compel funding of nonpolitical speech (which might include art, [n.5] for example, as well as commercial advertising). While an individual's First Amendment interest in commercial speech, and thus the government's burden in justifying a regulation of it, may well be less weighty than the interest in ideological speech, Abood continues to stand for the proposition that being compelled to make expenditures for protected speech "works no less an infringement of . . . constitutional rights" than being prohibited from making such expenditures. Abood, 431 U. S., at 234. The fact that no prior case of this Court has applied this principle to commercial and nonideological speech simply reflects the fortuity that this is the first commercial speech subsidy case to come before us.

An apparent third ground for the Court's conclusion that the First Amendment is not implicated here is its assumption that respondents do not disagree with the advertisements they object to subsidizing. See ante, at 11, 13. But this assumption is doubtful and would be beside the point even if true. As the Court itself notes, ante, at 8-9, and n. 11, respondents do claim to disagree with the messages of some promotions they are being forced to fund: some of the ads promote specific varieties of plums, peaches, and nectarines marketed by respondents' competitors but not by respondents; other ads characterize California tree fruits as a generic and thus fungible commodity, whereas respondents believe that their produce is superior to most grown in California. While these points of disagreement may seem trivial to the Court, they in fact relate directly to a vendor's recognized First Amendment interest in touting his wares as he sees fit, so long as he does not mislead. Supra, at 3-4. Whether the "central message," ante, at 11, of the generic advertising is that all California peaches, plums, and nectarines are equally good, or that only the varieties and characteristics featured in the advertisements are desirable, respondents do indeed disagree with that message.

In any event, the requirement of disagreement finds no legal warrant in our compelled speech cases. In Riley, for example, we held that the free speech rights of charitable solicitors were infringed by a law compelling statements of fact with which the objectors could not, and did not profess to, disagree. See Riley v. National Federation of Blind of N. C., Inc., 487 U. S., at 797-798. See also Hurley, 515 U. S., at 573 ("[The] general rule, that the speaker has the right to tailor the speech, applies not only to expressions of value, opinion, or endorsement, but equally to statements of fact the speaker would rather avoid . . ."); Barnette, 319 U. S., at 635 (if the Free Speech Clause bars the government from making the flag salute a legal duty, nonconformist beliefs are not required to exempt one from saluting). Indeed, the Abood cases themselves protect objecting employees from being forced to subsidize ideological union activities unrelated to collective bargaining, without any requirement that the objectors declare that they disagree with the positions espoused by the union. See, e.g., Teachers v. Hudson, 475 U. S., at 301-302; Abood, 431 U. S., at 234. Requiring a profession of disagreement is likewise at odds with our holding two Terms ago that no articulable message is necessary for expression to be protected, Hurley, 515 U. S., at 569; protection of speech is not limited to clear cut propositions subject to assent or contradiction, but covers a broader sphere of expressive preference. What counts here, then, is not whether respondents fail to disagree with the generalized message of the generic ads that California fruit is good, but that they do indeed deny that the general message is as valuable and worthy of their support as more particular claims about the merits of their own brands. One need not "disagree" with an abstractionist when buying a canvas from a representational painter; one merely wishes to support a different act of expression.

The Secretary of Agriculture has a further argument for minimizing or eliminating scrutiny of this subsidization mandate, which deserves some mention even though the Court does not adopt it. The Secretary calls for lesser scrutiny of forced payments for truthful advertising and promotion than for restrictions on commercial speech, on the ground that the effect of compelled funding is to increase the sum of information to the consuming public. This argument rests, however, on the assumption that regulation of commercial speech is justified solely or largely on preservation of public access to truthful information, an assumption we have already seen to be inaccurate. Supra, at 2-5. Truth is indeed a justifiable objective of commercial speech protection, but so is nonmisleading persuasion directed to the advertiser's own choice of what to promote.

Although not cited by the Secretary, the closest pass at authority for his limited rationale of commercial speech protection is Zauderer v. Office of Disciplinary Counsel of Supreme Court of Ohio, 471 U.S. 626 (1985), our only examination of a commercial speech mandate before today. The state law there required disclosures about the method of calculating a contingent fee when legal representation on that basis was advertised. In speaking of the objecting lawyer's comparatively modest interest in challenging the state requirement, we referred to protection of commercial speech as "justified principally by the value to consumers of the information such speech provides . . . ." Id., at 651 (citation omitted); see also Virginia Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U. S., at 765, 770; Rubin v. Coors Brewing Co., 514 U. S., at 481 (1995). But this proposition will not bear the weight of the government's position. We said "principally," not exclusively, and proceeded to uphold the state requirement not because a regulation adding to public information is immune from scrutiny, but because the mandate at issue bore a reasonable relation to the "State's interest in preventing deception of consumers," 471 U. S., at 651, who might otherwise be ignorant of the real terms on which the advertiser intended to do business. Zauderer thereby reaffirmed a long standing preference for disclosure requirements over outright bans, as more narrowly tailored cures for the potential of commercial messages to mislead by saying too little. See id., at 651-652, n. 14; see also Hurley, 515 U. S., at 573; Riley v. National Federation of Blind of N. C., Inc., supra, at 796, n. 9 (1988); Central Hudson, 447 U. S., at 565; Virginia Bd. of Pharmacy, supra, at 771-772. But however long the pedigree of such mandates may be, and however broad the government's authority to impose them, Zauderer carries no authority for a mandate unrelated to the interest in avoiding misleading or incomplete commercial messages.

For the reasons discussed above, none of the Court's grounds suffices for discounting respondents' interests in expression here and treating these compelled advertising schemes as regulations of purely economic conduct instead of commercial speech. I would therefore adhere to the principle laid down in our compelled speech cases: laws requiring an individual to engage in or pay for expressive activities are reviewed under the same standard that applies to laws prohibiting one from engaging in or paying for such activities. Under the test for commercial speech, the law may be held constitutional only if (1) the interest being pursued by the government is substantial, (2) the regulation directly advances that interest and (3) is narrowly tailored to serve it. Central Hudson, supra, at 566 (1980). [n.6] The burden is on the government. Edenfield v. Fane, 507 U. S., at 770; Board of Trustees of State Univ. of N. Y. v. Fox, 492 U.S. 469, 480 (1989). In this case, the Secretary has failed to establish that the challenged advertising programs satisfy any of these three prongs of the Central Hudson test.

The express purposes of the Agriculture Marketing Agreement Act, 7 U.S.C. § 601 et seq. (AMAA or Act), including the advertising programs established under it, are to stabilize markets for covered agricultural products and maintain the prices received by farmers. 7 U.S.C. §§ 602(1), (4); see also Federal Agriculture Improvement and Reform Act of 1996 (FAIR Act) §§501(b)(1), (3), Pub. L. 104-127, 110 Stat. 888, 1030 (finding by Congress that the purpose of agricultural commodity promotion laws is to maintain and expand the market for covered commodities). [n.7] It is doubtless true that at a general level these are substantial government interests, and unless there were some reason to doubt that undue market instability or income fluctuation has in fact affected a given segment of the economy, governmental efforts to address such problems would require little to satisfy the first Central Hudson criterion that a substantial government interest be the object of the regulation. Thus, if the government were to attack these problems across an interstate market for a given agricultural commodity or group of them, the substantiality of the national interest would not be open to apparent question, and the sole issues under Central Hudson would seem to be whether the means chosen were sufficiently direct and well tailored. But when the government's program targets expression in only a narrow band of a broad spectrum of similar market activities in which its interests appear to be at stake, a question naturally does arise. For the arbitrariness or underinclusiveness of the scheme chosen by the government may well suggest that the asserted interests either are not pressing or are not the real objects animating the restriction on speech. See Rubin v. Coors Brewing Co., 514 U. S., at 489 ("[E]xemptions and inconsistencies" in alcohol labeling ban "bring into question the purpose of the . . . ban," such that it does not survive the Central Hudson test); City of Ladue v. Gilleo, 512 U.S. 43, 52-53 (1994) ("Exemptions from an otherwise legitimate regulation of a medium of speech . . . may diminish the credibility of the government's rationale for restricting speech in the first place"); Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 424-426 (1993) (same); Florida Star v. B. J. F., 491 U.S. 524, 540 (1989) ("[T]he facial underinclusiveness" of a regulation of speech "raises serious doubts about whether Florida is, in fact, serving, with this statute, the significant interests" invoked in support of it). Under such circumstances, the government's obligation to establish the empirical reality of the problems it purports to be addressing, see Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 664 (1994); Edenfield v. Fane, supra, at 770-771, requires a sensible reason for drawing the line between those instances in which the government burdens First Amendment freedom in the name of the asserted interest and those in which it does not.

Here, the AMAA's authorization of compelled advertising programs is so random and so randomly implemented, in light of the Act's stated purposes, as to unsettle any inference that the Government's asserted interest is either substantial or even real. First, the Act authorizes paid advertising programs in marketing orders for 25 listed fruits, nuts, vegetables, and eggs, but not for any other agricultural commodity. See 7 U.S.C. § 608c(6)(I). [n.8] The list includes onions but not garlic, tomatoes but not cucumbers, Tokay grapes but not other grapes, and so on. The selection is puzzling. The only thing the limited list unambiguously shows is that a need for promotional control does not go hand in hand with a need for market and economic stability, since the authorization for marketing orders bears no such narrow restriction to specific types of produce. But no general criterion for selection is stated in the text, and neither Congress nor the Secretary has so much as suggested that such a criterion exists. Instead, the legislative history shows that from time to time Congress has simply amended the Act to add particular commodities to the list at the request of interested producers or handlers, without ever explaining why compelled advertising programs were necessary for the specific produce chosen and not others. [n.9] The legislative history for the bill authorizing paid advertising programs for plums, nectarines, and several other commodities is a good case on point. The record indicates merely that "[o]ver the past several years, numerous commodity groups have come to the Congress and asked for authority to provide for [market development and advertising] activities under the terms of their agreement and it has always been granted. This bill combines several such individual requests made by various producer groups operating under marketing agreements or orders." H. R. Rep. No. 89-846, 89th Cong., 1st Sess., 2 (1965). A letter from the Acting Secretary of Agriculture appended to the cited House Report similarly accounts for the choice of covered products solely by reference to grower and handler interest. Id., at 3-4. Or, again, the legislative history of the amendment adding "California grown peaches" to the list refers only to the view of the Department of Agriculture that "any fruit or vegetable commodity group which actively supports the development of a promotion program by this means should be given an opportunity to do so." S. Rep. No. 92-295, p. 2 (1971). Nor do the proposed rulemakings for authorizing advertising programs in marketing orders carry findings that might explain why such programs might be needed for the specified commodities but not others; the announcements rely instead on a "consensus of the industry . . . that promotional activities . . . have been beneficial in increasing demand," 36 Fed. Reg. 8736 (1971) (plums); see also 41 Fed. Reg. 14376-14377 (1976) (peaches). [n.10]

Of course, when government goes no further than regulating the underlying economic activity, this sort of piecemeal legislation in answer to expressions of interest by affected parties is plainly permissible, short of something so arbitrary as to fail the rational basis test. See, e.g., Williamson v. Lee Optical of Okla., Inc., 348 U.S. 483, 487-489 (1955). But when speech is at stake, the government fails to carry its burden of showing a substantial interest when it does nothing more than refer to a "consensus" within a limited interest group that wants the regulation. Instead, the erratic pattern of regulation itself places the reality of any public or governmental interest in question, and a correlation with nothing more than the priorities of particular interest groups gives no reassuring answer. [n.11]

A second element of the arbitrary in this statutory and regulatory scheme inheres in the geographical limitations on the marketing orders that include the advertising programs challenged in this case, which apply only to peaches, plums, and nectarines grown in California, unaccompanied by counterparts for advertising the same commodities grown elsewhere. Some geographical restriction, it must be said, follows from the general provision of the AMAA limiting marketing orders to the smallest production or marketing area practicable and consistent with the policy of the Act. See 7 U.S.C. § 608c(11)(B). But this provision merely explains why a substantial governmental interest in advertising a type of produce would have to be manifested in as many orders under the AMAA as there are defined production or marketing areas; it does nothing to explain the oddity that a government interest worth vindicating should occur within such geographically select boundaries and nowhere else, or to negate the suggestion of the evidence already mentioned, that the government's asserted interest is nothing more than the preference of a local interest group.

The oddity is most pronounced in the instance of peaches, since the statute itself authorizes forced advertising only in marketing orders for "California grown peaches," not in orders for peaches grown anywhere else in the country. §608c(6)(I). Although California is the biggest peach growing State, more than 30 others also grow peaches commercially and together typically account for about half of the national crop, and roughly two thirds of the peaches sold fresh. See App. 389; U. S. Dept. of Agriculture, Agricultural Statistics, 1995-96, p. V-23 (Table 294). Yet the non California peaches are utterly ignored in the Government's promotional orders. The challenged advertising campaign for "California Summer Fruits," running in markets throughout the United States and in Canada, see App. 341-343, 477-479, does not proclaim simply that peaches or the other fruits are good things. Rather, as the Secretary tells us, the advertising program "promotes California fruit as unique." Brief for Petitioner 31. It may or may not be, but promoting a crop from one State at the expense of essentially the same thing grown in the others reveals nothing about a substantial national interest justifying the National Government in restricting speech. Without more, the most reasonable inference is not of a substantial government interest, but effective politics on the part of producers who see the chance to spread their advertising costs. Nothing more appears. [n.12]

The Secretary makes no attempt to explain how the Act's geographical scope restrictions relate to the asserted goals of the advertising programs. The general restriction of marketing orders to the smallest practicable area has been part of the Act since it became law, long before Congress permitted compelled advertising, the authorization for which was simply grafted onto the existing Act as a convenient vehicle for the funding schemes. See n. 9., supra; see also S. Rep. No. 92-295, p. 2 (letter from Department of Agriculture indicating that the AMAA "could provide the facility for" financing commodity advertising programs). Nor does any explanation appear for restricting peach advertising programs to California produce. Without some explanation, one would expect something quite different, that a compelled advertising program of the National Government intended to increase consumer demand for an agricultural commodity would apply to produce grown throughout the land. Indeed, in recently enacting the FAIR Act, which authorizes compulsory advertising programs for all agricultural commodities on a national basis (but also leaves the separate provisions of the AMAA intact, see §524, 110 Stat. 1047), Congress specifically found that "[t]he cooperative development, financing, and implementation of a coordinated national program of research, promotion, and information regarding agricultural commodities are necessary to maintain and expand existing markets and to develop new markets for these commodities." §512(a)(7), id., at 1033 (emphasis added); see also §514(a)(2), id., at 1035 ("Each order issued under this section shall be national in scope"). The AMAA, of course, actually prohibits orders of national scope. In sum, these advertising schemes come with a statutory text and regulatory history so remote from the government's asserted interests as to undermine the reality, let alone the substantiality, of the claims put forward by the Secretary in attempting to satisfy Central Hudson's first requirement.

Even if the Secretary could establish a sufficiently substantial interest, he would need also to show how the compelled advertising programs directly advance that interest, that is, how the schemes actually contribute to stabilizing agricultural markets and maintaining farm income by stimulating consumer demand. To show this required causation, the Secretary relies on cases concerning governmental bans on particular advertising content, where we have accepted the unremarkable presumption that advertising actually works to increase consumer demand, so that limiting advertising tends to soften it. See United States v. Edge Broadcasting Co., 509 U.S. 418, 428 (1993); Posadas de Puerto Rico Associates v. Tourism Co. of P. R., 478 U.S. 328, 341-342 (1986); Central Hudson, 447 U. S., at 569. This presumption is not, however, automatically convertible into support for the Secretary here. In the cases mentioned, the question has been whether some advertising (in the absence of the government's ban) would be more effective in stimulating demand than no advertising (due to the ban). Here, in contrast, the causal question of direct advancement does not involve comparing the effectiveness of something with nothing, for even without the coercive promotional schemes there would be some voluntary advertising. Thus, the question here requires a comparison of the effectiveness of advertising under the Government's program with the effectiveness of whatever advertising would likely exist without it. [n.13]

For this purpose, the Secretary correctly notes that the effectiveness of the Government's regulation must be viewed overall, considering the market behavior of growers and handlers generally, not just in its isolated application to one or a few individuals such as respondents. Edge Broadcasting, supra, at 427. The Secretary therefore argues that though respondents have voiced the desire to do more individual advertising if the system of mandatory assessments were ended, other handlers who benefit from the Government's program might well become "free riders" if promotion were to become wholly voluntary, to the point of cutting the sum total of advertising done. That might happen. It is also reasonably conceivable, though, that pure self interest would keep the level of voluntary advertising high enough that the mandatory program could only be seen as affecting the details of the ads or shifting their costs, in either event without effect on market stability or income to producers as a group. [n.14] We, of course, do not know, but these possibilities alone should be fatal to the Government here, which has the burden to establish the factual justification for ordering a subsidy for commercial speech. Mere speculation about one or another possibility does not carry the burden, see Turner Broadcasting System, 512 U. S., at 664; Edenfield v. Fane, 507 U. S., at 770-771, and the Government has to show that its mandatory scheme appreciably increases the total amount of advertising for a commodity or somehow does a better job of sparking the right level of consumer demand than a wholly voluntary system would. There is no evidence of this in the record here.

Finally, a regulation of commercial speech must be narrowly tailored to achieving the government's interests; there must be a " `fit' between the legislature's ends and the means chosen to accomplish those ends,--a fit . . . that represents not necessarily the single best disposition but one whose scope is in proportion to the interest served." Board of Trustees of State Univ. of N. Y. v. Fox, 492 U. S., at 480 (citations and internal quotation marks omitted). This sense of fitness is not precise, to be sure, but it rules out a regulation if "far less restrictive and more precise means" are available. Id., at 479 (internal quotation marks omitted). Respondents argue that the mandatory advertising schemes for California peaches, plums, and nectarines fail this narrow tailoring requirement, because they deny handlers any credit toward their assessments for some or all of their individual advertising expenditures. The point is well taken. On its face, at least, a credit system would be a far less restrictive and more precise way to achieve the government's stated interests, eliminating as it would much of the burden on respondents' speech without diminishing the total amount of advertising for a particular commodity. Indeed, the remarkable thing is that the AMAA itself provides for exactly such credits for individual advertising expenditures under marketing orders for almonds, filberts, raisins, walnuts, olives, and Florida Indian River grapefruit, but not for other commodities. 7 U.S.C. § 608c(6)(I).

The Secretary contends, however, that the purpose of individual "branded" advertising is to increase the market share of a single handler, and so is at odds with the purpose of the Government's mandatory program, which is to expand the overall size of the market through the use of "generic" advertising for a commodity generally. See also FAIR Act §§501(b)(6), (7), 110 Stat. 1030-1031 (congressional finding of same). Perhaps so, but that does not tell us what to make of the credit for, say, private raisin advertising. It would be hard to imagine more effectively "branded" advertising than promotions for Sun Maid raisins, but the statute would allow Sun Maid a credit. Why would that be consistent with the Government's generic objective, but a credit for respondents' nectarine ads not be? The Government gives us no answer. Without some further explanation, the statute on raisin advertising seems to reflect a conclusion that could reasonably be drawn after examining some of the "branded" advertising in the record before us. A consumer galvanized by respondents' depiction of "Mr. Plum," App. 542, might turn down a plum by any other name, but I doubt it. [n.15]

I acknowledge that in implementing a credit program for individual advertising in an otherwise valid compulsory program, the government would need substantial leeway in determining whether such expenditures do in fact further the goal of expanding markets generally. But where, as here, no particular evaluation has been made, and the statute dealing with other fruit apparently assumes that some private advertising does serve the common good, and everything else is left to assertion, there could be no finding that a program completely denying credits for all individual advertising expenditures is narrowly tailored to an interest in the stability or expansion of overall markets for a commodity.

* * *

Although the government's obligation is not a heavy one in Central Hudson and the cases that follow it, we have understood it to call for some showing beyond plausibility, and there has been none here. I would accordingly affirm the judgment of the Ninth Circuit.


1 Thus, commercial advertising generally and these programs in particular involve messages that go well beyond the ideal type of pure commercial speech hypothesized in Virginia Bd. of Pharmacy, which would do " `no more than propose a commercial transaction,' " 425 U. S., at 762 (quoting Pittsburgh Press Co. v. Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 385 (1973)), by communicating the idea " `I will sell you the X prescription drug at the Y price,' " 425 U. S., at 761.

2 The Secretary of Agriculture does not argue that the advertisements at issue represent so called "government speech," with respect to which the government may have greater latitude in selecting content than otherwise permissible under the First Amendment, see Keller v. State Bar of Cal., 496 U.S. 1, 10-13 (1990); Abood v. Detroit Bd. of Ed., 431 U.S. 209, 259, n. 13 (1977) (Powell, J., concurring in judgment). See Brief for Petitioner 25, n. 16 (waiving argument).

3 That is, the Court appears to hold that a compelled subsidy of speech does not implicate the First Amendment if the speech either is germane to an otherwise permissible regulatory scheme or is non ideological, so that each of these characteristics constitutes an independent, sufficient criterion for upholding the subsidy. See, e.g., ante, at 15 ("[The Abood] test is clearly satisfied in this case because (1) the generic advertising of California peaches and nectarines is unquestionably germane to the purposes of the marketing orders and, (2) in any event, the assessments are not used to fund ideological activities") (emphasis added).

4 The Court purports to find support for its more permissive reading of the Abood "germaneness test" in a separate holding of Lehnert allowing mandatory charges for portions of the union's internal newsletter, the Teachers' Voice, that concerned " `teaching and education generally, professional development, unemployment, job opportunities, award programs . . ., and other miscellaneous matters.' " Ante, at 14 (quoting Lehnert v. Ferris Faculty Assn., 500 U.S. 507, 529 (1991)). But the Lehnert Court noted that these communications, though plainly speech, were not "public in nature," 500 U. S., at 529; the Teachers' Voice was the union's means of communicating with its members, not the public at large, see Lehnert v. Ferris Faculty Assn MEA NEA, 643 F. Supp. 1306, 1328 (WD Mich. 1986), aff'd, 881 F. 2d 1388 (CA6 1989), aff'd in part and rev'd in part on other grounds, 500 U.S. 507 (1991). In upholding charges for this type of internal communication, Lehnert simply followed our earlier decision in Ellis v. Railway Clerks, 466 U.S. 435 (1984), in which we reasoned that "[t]he union must have a channel for communicating with the employees, including the objecting ones, about its activities. . . . [The union surely may] charge objecting employees for reporting to them about those activities it can charge them for doing." Id., at 450-451. In other words, this type of internal communication about chargeable activities, unlike the public advertising campaign struck down in Lehnert, was necessary to the union's role as collective bargaining agent and imposed no greater burden on the employees' First Amendment interests than their compelled association with the union in the first instance. In these respects, however, the instant advertising programs are much more like the impermissible public relations campaign than the permissible internal communications at issue in Lehnert.

5 Cf. Schad v. Mount Ephraim, 452 U.S. 61, 65 (1981) ("Entertainment, as well as political and ideological speech, is protected; motion pictures, programs broadcast by radio and television, and live entertainment, such as musical and dramatic works, fall within the First Amendment guarantee").

6 Contrary to some arguments offered by respondents, these advertising schemes are not removed from the commercial category on the grounds that they are content based, producing not mere "dissemination of `purely factual and uncontroversial information,' " Hurley v. Irish American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U.S. 557, 573 (1995) (quoting Zauderer v. Office of Disciplinary Counsel of Supreme Court of Ohio, 471 U.S. 626, 651 (1985)), but controversial and ideological messages, and even objectionable sexual imagery. Regulation of commercial speech necessarily turns on some assessment of content, Virginia Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 761 (1976), yet that fact has never been thought sufficient to require a standard of strict scrutiny. And we have consistently held that advertising does not automatically lose its character as commercial speech simply because it may do much more than propose a transaction or disseminate purely factual information. See, e.g., Board of Trustees of State Univ. of N. Y. v. Fox, 492 U.S. 469, 473-475 (1989); Bolger v. Youngs Drug Products Corp., 463 U.S. 60, 66-68 (1983). The concept of commercial speech would be reduced to a relic if the threshold for imposing strict scrutiny were reached simply because certain advertisements evoke vaguely nostalgic themes of indeterminate political import or because the hypersensitive may see the specter of sex in the film of a child eating a peach.

7 A subtitle of the FAIR Act, which was enacted on April 4, 1996, authorizes promotion and advertising orders for any agricultural commodity. Its procedural mechanisms are similar to those put in place by the AMAA, although there is one noticeable difference (other than breadth of coverage) between the two laws: orders issued under FAIR, unlike those under the AMAA, must be national in scope. FAIR Act §§511-526, 110 Stat. 1032-1048. The new Act does not, however, affect or pre-empt any other federal or state law, such as the AMAA, authorizing promotion or research relating to an agricultural commodity. §524, id., at 1047. The FAIR Act also includes new findings in support of "commodity promotion laws," including the advertising provisions of the AMAA. §501, id., at 1029.

8 Section 608c(6)(I) currently provides that marketing orders may include terms "[e]stablishing or providing for the establishment of production research, marketing research and development projects designed to assist, improve, or promote the marketing, distribution, and consumption or efficient production of any such commodity or product, the expense of such projects to be paid from funds collected pursuant to the marketing order: Provided, That with respect to orders applicable to almonds, filberts (otherwise known as hazelnuts), California grown peaches, cherries, papayas, carrots, citrus fruits, onions, Tokay grapes, pears, dates, plums, nectarines, celery, sweet corn, limes, olives, pecans, eggs, avocados, apples, raisins, walnuts, tomatoes, or Florida grown strawberries, such projects may provide for any form of marketing promotion including paid advertising and with respect to almonds, filberts (otherwise known as hazelnuts), raisins, walnuts, olives, and Florida Indian River grapefruit may provide for crediting the pro rata expense assessment obligations of a handler with all or any portion of his direct expenditures for such marketing promotion including paid advertising as may be authorized by the order . . . ."

9 The substantive terms of marketing orders under the AMAA as originally enacted were generally limited to restrictions on the total marketable quantity of the commodity, allocations among handlers, disposition of surplus quantities, and maintenance of reserve supplies. 7 U.S.C. § 608c(6) (1934 ed., Supp. III). For the first time in 1954, Congress permitted marketing orders to establish "marketing research and development projects designed to assist, improve, or promote the marketing, distribution, and consumption [of a] commodity or product, the expense of such projects to be paid from funds collected pursuant to the marketing order." 68 Stat. 907; 7 U.S.C. § 608c(6)(I). Since then, Congress has repeatedly amended the Act to authorize, but only for specified commodities, "any form of marketing promotion including paid advertising." §608c(6)(I). The first such amendment, in 1962, allowed advertising programs for cherries, Pub. L. 87-703, 76 Stat. 632; similar schemes for plums and nectarines followed in 1965, Pub. L. 89-330, 79 Stat. 1270, and for "California grown peaches" in 1971, Pub. L. 92-120, 85 Stat. 340; and today, various authorizations cover the 25 commodities listed in §608c(6)(I). The Act now also permits crediting some or all of a handler's independent expenditures for advertising against his assessment obligations with respect to 6 commodities (but not nectarines, plums, or peaches). §608c(6)(I).

10 A possible exception is the proposed rulemaking for nectarines, which refers to the relative unfamiliarity of the consuming public with nectarines, due in part to the fact that new varieties that could be marketed nationally had only recently been developed. See 31 Fed. Reg. 5635, 5636 (1966). This solitary finding does not cure the other defects of the statutory scheme, however.

11 This does not mean that taking the views of the industry into account in itself renders a program suspect. Both the AMAA and the more general authorization of compelled agricultural advertising programs recently enacted as part of the FAIR Act require orders implementing such programs to be approved by producers and/or handlers in periodic referenda. See 7 U.S.C. §§ 608c(8)(A), (B), (9)(B)(i), (16), (19); FAIR Act §518, 110 Stat. 1043-1044. Since the asserted purpose of these advertising schemes is to increase demand for the covered commodities and thereby maintain the income of producers and handlers, requiring periodic approval by those most likely to benefit if a program is working as planned may serve as an additional check on whether the purpose of the program is in fact being achieved. Contrary to what the majority implies, see ante, at 19, however, the mere vote of a majority is never enough to compel dissenters to pay for private or quasi private speech whose message they do not wish to foster; otherwise, the First Amendment would place no limitation on this type of majoritarian action.

12 While plum and nectarine production is more highly concentrated in California, see U. S. Dept. of Agriculture, Agricultural Statistics, 1995-96, pp. V-21, V-27 to V-28 (Tables 288, 304-308), the AMAA's requirement that marketing orders cover the smallest geographical area practicable still lacks any reasonable connection to the asserted purposes of the advertising programs instituted thereunder.

13 Although they do not apply the Central Hudson test, the majority does criticize the Court of Appeals' application of it as "illogical" insofar as that court enquired whether collective advertising or purely private advertising is more effective at stabilizing markets, because the Act's basic policy is to achieve its economic goals by compelling cooperation in lieu of independent, competitive decisionmaking. Ante, at 16-17. But the extent to which the Act eliminates competition varies among different marketing orders, and the spottiness of collective advertising schemes under the Act demonstrates that there is no necessary connection between some compelled economic cooperation and forced collective advertising. There is thus nothing "illogical" in comparing the effectiveness of collective and private advertising schemes in the context of the marketing order regime.

14 While even on the cost shifting scenario the Government would have reduced the "problem" of free riders referred to by the Secretary, that would not be a sufficient free standing justification for the program. "[P]rivate speech often furthers the interests of nonspeakers, and that does not alone empower the state to compel the speech to be paid for," Lehnert v. Ferris Faculty Assn., 500 U. S., at 556 (Scalia, J., concurring in judgment in part and dissenting in part). We have never sustained a restriction on speech solely because some individuals would ride free on the private speech of others, but only when the free rider problem arises in serving other substantial governmental interests.

15 The Secretary also maintains that credit programs are appropriate for market conditions specific to the almond industry, where a single producer cooperative has a 92% share of the market for direct sales to consumers, see Cal Almond, Inc. v. United States Dept. of Agriculture, 14 F. 3d 429, 438, n. 9 (CA9 1993), because in such circumstances "certain types of individual and brand advertising may accomplish the government's goals of market stability and increased consumption without creating a significant free rider problem." Brief for Petitioner 47. As with the Secretary's other proffered justifications for the seemingly arbitrary choices made in the AMAA provisions concerning advertising, this explanation rests on nothing more than an unsubstantiated assertion, here about the effects of brand advertising. Moreover, the legislative and regulatory history provides no indication that this was the reason for permitting credits for almonds, but not plums, nectarines, or California grown peaches. To the extent the record says anything, it seems to say quite the contrary of what the Secretary claims. See S. Rep. No. 91-1204, p. 2 (1970) (incorporating letter from Almond Growers Council noting that credit provision for almonds "will be model legislation for other commodities"); 37 Fed. Reg. 3983 (1972). The Secretary's explanation only leads one to wonder about filberts, for example; is their production, too, under the domination of a large cooperative? Is the grapefruit market structured in a way that renders virtually generic the brand specific advertising for the Indian River crop?