Concept
The Central Hudson Test
Regulation of commercial speech
The Central Hudson Test
Nearly every modern commercial speech dispute is decided under Central Hudson Gas & Electric Corp. v. Public Service Commission, in which Justice Powell asked whether New York could completely ban promotional advertising by electric utilities. During a fuel shortage, the state Public Service Commission had ordered the utilities to stop promoting the use of electricity. Three years later, with the shortage eased, it asked for public comment on continuing the ban. Central Hudson opposed the ban on First Amendment grounds, and the Commission extended it in a Policy Statement issued on February 25, 1977.1 The New York courts, up through the Court of Appeals, upheld the order; the Supreme Court reversed.2
Virginia Pharmacy established that commercial speech is constitutionally protected and explained why. Central Hudson then organized the law that had grown up under that decision into a test for measuring restrictions on commercial speech.3
The four parts fall into two stages. The first is a gate: a message that is false, deceptive, or related to illegal activity has no informational value, so the government may regulate or ban it without satisfying the rest of the test.5 The last three steps apply only to truthful speech about lawful activity, and they ask whether the government has a weighty reason and a close means. The means must work, because a regulation that gives only ineffective or remote support for the interest cannot stand. Its scope must also be proportionate, because a restriction that a more limited one could replace cannot survive.6
Together the last three steps make the test a form of intermediate scrutiny, a label the Court itself has used for its review of commercial speech restrictions.7 The government carries the burden of showing a substantial goal and of calculating the cost to speech with care.8 The test demands more than rational-basis review, which the Court has said it does not equate with this standard,9 and less than strict scrutiny, because commercial speech is hardy and its claims are verifiable, as the previous section explained.
In Central Hudson itself, the ban on promotional advertising passed the first three steps as to energy conservation and failed the fourth.10 The decision also drew separate writings. Justice Blackmun concurred only in the judgment because he thought the four-part test gives truthful, nonmisleading, noncoercive commercial speech too little protection. Justice Stevens, also concurring in the judgment, doubted that the case involved commercial speech at all. Justice Rehnquist dissented and would have upheld the ban.11 Those views return later in this chapter, because they explain why a test that still governs remains contested. The four steps now follow in order, with Central Hudson as the running example, beginning with the threshold inquiry into lawful activity and misleading speech.
Lawful and Not Misleading
The first step decides whether a commercial message is protected at all. Justice Powell framed the threshold in one sentence: for commercial speech to come within the First Amendment, “it at least must concern lawful activity and not be misleading.”12 A message that fails this gate receives no protection, and the three steps that follow never come into play.13
The gate follows from the reason the Amendment reaches advertising at all. Its concern for commercial speech rests on the informational function of advertising, so the Constitution does not object to suppressing commercial messages that fail to inform the public accurately about lawful activity.14 Virginia Pharmacy described that function: advertising disseminates information about who is producing and selling what product, for what reason, and at what price, and the free flow of that information is indispensable to the private economic decisions that allocate resources in a free enterprise system.15 Because the protection attaches to the information, it runs to the audience as well as the seller. Drug consumers, not a pharmacist, brought that case, and the Court held that if there is a right to advertise, there is a reciprocal right to receive the advertising.16
For the government, the consequences of failing the gate are large. When it is trying to deter or punish false or misleading advertising, a court applies no overbreadth analysis, and the government need not show that its law is no more extensive than necessary.17 A law forbidding newspapers to run advertisements for prostitution, where prostitution is illegal, would therefore be upheld without any First Amendment analysis.18
Central Hudson shows the passing side of the line. The utility’s promotional advertising was lawful commercial speech,19 yet the New York Court of Appeals questioned whether it deserved protection at all. Because the utility held a monopoly over electricity in its service area, the court reasoned, advertising in a noncompetitive market could not improve consumers’ decisions, so the order restricted no commercial speech of any worth.20 The Supreme Court disagreed. A monopoly over one product offers no shelter from competition with substitutes, and electric utilities compete with suppliers of fuel oil and natural gas. Even in a true monopoly market, a consumer may need information to decide whether to use the service at all or how much to buy, and a business that pays for promotion presumably believes consumers are interested.21
The threshold therefore asks two narrow questions, accuracy and lawfulness. Whether a message conveys much useful information is not one of them, because the remaining steps exist to weigh the government’s reasons against what the speech offers. The utility’s advertisements cleared the gate, and the Court went on to the other three steps. Lawyers’ targeted letters to accident victims cleared it too in Florida Bar v. Went For It, Inc.: the letters were not misleading and concerned a lawful service, so the Court asked whether the Bar’s asserted interest was substantial and whether the ban advanced it.22
Advertising that proposes an illegal transaction fails the gate for the same reason. In Pittsburgh Press Co. v. Pittsburgh Commission on Human Relations, the commission had concluded that sex-designated help-wanted columns were illegal sex discrimination and barred newspapers from using them. The Court held that the prohibition did not violate the First Amendment, because commercial expression connected with illegal activity enjoys no protection.23 Central Hudson later cited the decision for the rule that the government may ban commercial speech related to illegal activity.24
Legality is measured at the transaction the message proposes, so a legislature’s dislike of a lawful product does not defeat this step. The Court has rejected the argument that a power to ban a product includes the lesser power to ban its advertising, even for harmful but lawful products,25 and it has refused to give legislatures broader latitude over speech promoting socially harmful activities.26 A complete ban also carries an inner political check, such as the revenue the banned activity produces, that a ban on advertising alone lacks.27 Advertising for a lawful but harmful product therefore clears the gate and must meet the remaining three steps, a point the later section on lawful but harmful products develops.
False and deceptive messages fail the gate too. Virginia Pharmacy observed that much commercial speech is not provably false but only deceptive or misleading, and it foresaw no obstacle to a state’s dealing effectively with the problem: the First Amendment does not stop a state from insuring that the stream of commercial information “flows cleanly as well as freely.”28 Because commercial speech is hardy, regulating the false kind is unlikely to deter truthful commercial speech by accident.29 The category reaches beyond outright falsehood, since the government may ban forms of communication more likely to deceive the public than to inform it.30 Friedman v. Rogers, which upheld a Texas ban on practicing optometry under a trade name, is the standard example. The same logic supports the rule of Ohralik v. Ohio State Bar Association that the government may prohibit lawyers from soliciting prospective clients in person for profit: such speech inherently risks becoming deceptive, so even truthful solicitations can be forbidden.31
Inherent risk is a narrow category. In Peel v. Attorney Registration & Disciplinary Commission, a plurality found that a lawyer’s letterhead truthfully stating that a private organization had certified him as a civil trial specialist was neither actually, inherently, nor potentially misleading. It added that a state fearing such statements might mislead should prefer a disclaimer to a total ban. A message that might possibly mislead is therefore not automatically outside the First Amendment.
Run the first step before any balancing. When the facts show a false claim, a proposal of an illegal transaction, or a type of message that inherently invites deception, the analysis ends there. When the message is accurate and concerns lawful activity, the first step is satisfied even if the product is harmful, the speaker is a monopoly, or the information seems of little use, and the remaining three steps decide the case.
Substantial Governmental Interest
The second step asks whether the interest the government asserts is substantial, and it is the easiest step to pass. Central Hudson shows how easily. The Commission offered two interests, energy conservation and a fair utility rate structure, and the Court deemed both substantial.33 The rate argument ran this way: because the utility’s rates did not reflect marginal cost, added demand at peak periods would push the true cost of expanding production onto all consumers through higher overall rates, and without promotional advertising, the Commission said, that inequity would be less likely. The Court called the concern that rates be fair and efficient “a clear and substantial governmental interest.”34
The bar is low on purpose. Any nontrivial governmental interest will do, and laws designed to protect the health, safety, or welfare of residents usually pass without difficulty.35 The step does invite a value judgment about the importance of the interest, but the real testing falls on the third and fourth steps, which ask whether the speech limit serves the interest and fits it closely.
Florida Bar v. Went For It, Inc. shows the same easy passage. The Bar defended its rule by invoking the privacy of accident victims, and the Court had “little trouble crediting the Bar’s interest as substantial.” States have broad power to regulate the professions practiced within their borders, and the Court’s precedents leave no room for doubt that protecting potential clients’ privacy is a substantial state interest.36 The Court had also described the state’s interest in the well-being, tranquility, and privacy of the home as “of the highest order in a free and civilized society.”37
A court tests the regulation against the interest the government actually asserted. Went For It, quoting Edenfield v. Fane, explained that “[u]nlike rational basis review, the Central Hudson standard does not permit us to supplant the precise interests put forward by the State with other suppositions.”38 Under rational-basis review, a court may uphold a law by imagining any legitimate purpose it could serve. A court that rescued a speech restriction with a purpose the state never offered would leave the speaker unable to answer the real justification. The choice of interest also controls the later steps, because the third and fourth ask whether the regulation advances and fits the asserted interest and no other. In Central Hudson, the rate-fairness interest was substantial, yet the link between the ban and the utility’s rate structure was “at most, tenuous” and the effect of promotion on rate equity “highly speculative,” so that interest could not justify silencing the utility.39
An interest that surfaces only in litigation, or that the record does not support, fares worse. If the Court deems the asserted interest illusory or a post hoc rationalization, this step is not satisfied.40
Some justifications fail here for a different reason: they depend on the premise that people are better off knowing less. In Linmark Associates, Inc. v. Township of Willingboro, a township banned “For Sale” and “Sold” signs in order to promote stable, racially integrated housing.41 The Court held the ordinance unconstitutional because the First Amendment disables the state from achieving its goal by restricting the free flow of truthful information. Its primary infirmity was that the township suppressed truthful information on the belief that people would be better off with less speech and knowledge.42
Bates v. State Bar of Arizona applied the same idea to lawyers. It was no argument for the ban that the advertisements gave potential clients an incomplete basis for choosing a lawyer. The Court rejected the premise “that the public is best kept in ignorance than trusted with correct but incomplete information,” and with it any justification “based on the benefits of public ignorance.”43 Virginia Pharmacy had rejected an advertising ban as a paternalistic means of securing results the Court conceded were desirable, because the ban took away the consumer’s ability to choose by depriving him of the information needed to choose intelligently.44 Such a ban also burdens the audience’s reciprocal right to receive truthful advertising.45 The reasoning follows from how these bans work. A ban on truthful price or service information reaches its goal only if consumers who lack the information choose as the government prefers, so the goal depends on public ignorance.
Professional dignity fares no better. In Zauderer v. Office of Disciplinary Counsel, Ohio reprimanded a lawyer in part for an illustration in his advertisement and defended the restriction as protecting the dignity of lawyers’ communications. The Court was “unsure that the State’s desire that attorneys maintain their dignity in their communications with the public is an interest substantial enough to justify the abridgment of their First Amendment rights,” and the restriction on illustrations fell. The contrast with Went For It lies in the interest underneath. Dignity in how an advertisement looks is a matter of taste that injures no one, while the privacy interest credited in Went For It protects people in the aftermath of an accident.
Passing the second step settles only which interest the rest of the analysis will use. In Central Hudson, the energy interest carried the ban through the third step and the ban failed at the fourth, while the rate-fairness interest dropped out at the third. In Went For It, the privacy interest set the terms for the evidence the Bar had to produce and the fit it had to show. The next step takes up what the government must prove about the harm it recites and about the effect of its rule on that harm.
Direct Advancement
The third step tests the connection between the asserted interest and the speech restriction chosen to serve it, and Central Hudson shows the step working in both directions. The Commission offered two interests, and the Court gave each a different answer. On energy conservation the order passed, because there is “an immediate connection between advertising and demand for electricity,” and Central Hudson would not have contested the ban unless it believed promotion would increase its sales.47 On rate fairness the link was “at most, tenuous” and the effect “highly speculative.” Advertising to increase off-peak usage would have to increase peak usage while every other factor bearing on the fairness and efficiency of the utility’s rates stayed constant, and such “conditional and remote eventualities simply cannot justify silencing” the utility’s promotional advertising.48
The difference lies in how many steps separate the speech from the harm. Advertising raises demand and demand raises energy use, a single step that the utility concedes by fighting the ban. Rate fairness required a sequence of contingencies, each of which had to hold, and every added contingency is another place where the chain can break. The closeness requirement matters because of what a speech ban costs: a legislature that could defend a ban with a chain of maybes would take away information in exchange for a benefit that may never arrive. For the same reason the Court declines to uphold regulations that advance an interest only indirectly, and it has held that an advertising ban cannot be justified as protecting the ethical or performance standards of a profession, as in Bates and Virginia Pharmacy.49
Florida Bar v. Went For It states the government’s burden in the form later cases repeat. The government must show that its regulation advances the interest “in a direct and material way,” a burden that “is not satisfied by mere speculation or conjecture; rather, a governmental body seeking to sustain a restriction on commercial speech must demonstrate that the harms it recites are real and that its restriction will in fact alleviate them to a material degree.”50 That formulation contains two showings, and each guards against a different failure. Proof that the harm is real blocks a restriction aimed at a problem that exists only in the government’s assumptions. Proof of material relief blocks a restriction aimed at a real problem it cannot touch. A state can pass one showing and fail the other. The step has no mathematical measure, but a regulation fails when it offers only ineffective or remote support for the interest it invokes.51
A record that fails the step appears in Edenfield v. Fane, where the Court invalidated a Florida ban on in-person solicitation by certified public accountants. The state board claimed to fear fraud, overreaching, and compromised independence, yet it offered no studies suggesting that personal solicitation creates those dangers and no anecdotal evidence from any state, and the only evidence in the record tended to contradict its claims.52 Went For It applied the same standard to a rule barring personal injury lawyers from sending targeted direct mail to accident victims for thirty days, and it found none of Edenfield’s infirmities. The Bar had filed a 106-page summary of a two-year study of lawyer advertising and solicitation, with statistical and anecdotal data supporting its contention that the Florida public views direct mail in the immediate wake of accidents as an intrusion on privacy that reflects poorly on the profession. Lawyers were mailing about 700,000 direct solicitations in Florida each year, 40 percent of them aimed at accident victims or their survivors.53 The standard and the kind of claim were the same, and the outcomes differed because one record held no studies and no anecdotes while the other held both. The Justices split over whether the Bar’s proof sufficed, a split taken up in the Went For It sections below.
Linmark also shows a court demanding proof of the link. The Court questioned whether there was sufficient evidence that panic selling was likely, or that banning “For Sale” signs would prevent it.54 Critics of Posadas, which upheld Puerto Rico’s ban on casino advertising, argued that its link was remote at best. There is no way to know whether the ban affects resident casino gambling, much less the crimes associated with it, and because casino gambling will be promoted to tourists the ban may be wholly ineffective. A majority of the Court later disapproved the Posadas analysis in 44 Liquormart, Inc. v. Rhode Island, and a federal ban on casino advertising fell in Greater New Orleans Broadcasting Ass’n v. United States on similar reasoning.55
Internal consistency tests the link as well as evidence does. In Rubin v. Coors Brewing Co., federal law barred beer labels from stating alcohol content, yet barred alcohol content in advertising only in the 18 states that prohibited it, leaving brewers free to advertise strength in much of the country. Curbing advertising of alcohol content seemed a more potent weapon against “strength wars” than regulating labels, so the scheme was too irrational to advance its objective directly and materially.56 Exceptions of this kind matter for a reason beyond arithmetic. A government that leaves open the other routes by which a harm arrives cannot claim that closing one route relieves the harm to a material degree.
The third step is easily confused with the fourth, and the confusion costs points. Direct advancement asks whether the restriction works at all, while the fourth step asks whether it reaches farther than necessary or whether a narrower rule would serve the interest as well.57 A single regulation can fail both, as the label ban did in Rubin, where the irrational scheme failed directness and the availability of alternatives failed fit.58 An answer should assign each defect to its own step.
Fact patterns on this step usually supply a recited harm and a restriction, and the answer depends on what evidence the facts attribute to the government and what exceptions they attach to the rule. Facts that mention a survey, a study, or documented complaints point toward the Went For It result. Facts that mention a bare assertion, a contradicting record, or exceptions that let the same speech flow through other channels point toward Edenfield, Rubin, and Greater New Orleans.
Reasonable Fit
The fourth step asks whether a restriction on commercial speech is no more extensive than necessary to serve the government’s interest. In Central Hudson the Court called it “the critical inquiry in this case”: whether the Commission’s complete suppression of speech ordinarily protected by the First Amendment went further than necessary to further energy conservation.60 By this point the speech is truthful and lawful, the interest is credited, and the link has been tested, so what remains is proportion. A total ban on a category of advertising is the heaviest tool a regulator has, which is why the step does its most visible work on such bans.
The Commission’s order failed on two separate grounds. It reached all promotional advertising regardless of the effect of the advertised service on overall energy use, and the conservation rationale, however important, could not justify suppressing information about electric devices or services that would cause no net increase in total energy use. The utility said that but for the ban it would promote the heat pump, which both parties acknowledged to be a major improvement in electric heating, and electric heat as a backup to solar and other heat sources. Neither the Commission’s Policy Statement nor its order denying rehearing made findings on whether electric heating is efficient, so the Court credited as within the realm of possibility that it can be efficient in some circumstances.61 The second ground was the Commission’s failure to show that more limited regulation could not protect its interest adequately. It might have restricted the format and content of the advertising, for example by requiring advertisements to state the relative efficiency and expense of the offered service.62
Those two grounds describe what a fit analysis looks for. A regulatory technique may extend only as far as the interest it serves, so the government cannot regulate speech that poses no danger to its interest and cannot completely suppress information when narrower restrictions would serve that interest as well.63 Each limit guards against a different waste. A ban that sweeps in harmless speech pays for its benefit with messages that contribute nothing to the goal, which suggests the regulator never sorted the harmful from the harmless. A ban that forecloses a disclosure alternative replaces more information with less when added information could have done the job.
The words “no more extensive than necessary” invite a reading the Court has rejected. Read strictly, they would impose a least-restrictive-means test, but that test does not apply to commercial speech.64 The correction came in Board of Trustees of the State University of New York v. Fox, where a company offering Tupperware products was asked to leave a campus dormitory because it was not among the services authorized to operate there. The Court rejected the First Amendment challenge, accepting as substantial the university’s interests in promoting an educational rather than commercial atmosphere, safety and security, preventing commercial exploitation of students, and preserving residential tranquility.65 Went For It later restated the standard in Fox’s own words. The fourth step requires “a fit that is not necessarily perfect, but reasonable,” one whose scope is “in proportion to the interest served,” that employs “not necessarily the least restrictive means but . . . a means narrowly tailored to achieve the desired objective.”66
A strict least-restrictive-means rule would let any imaginable alternative defeat a regulation. Fox instead leaves the choice among reasonable methods with the government: so long as the means are narrowly tailored, it is for the government decisionmakers to judge what manner of regulation may best be employed.67 Reasonable fit nonetheless remains more demanding than rational-basis review. Went For It said so expressly, and it added that “numerous and obvious less-burdensome alternatives” are “certainly a relevant consideration in determining whether the ‘fit’ between ends and means is reasonable.”68 The government must also show that the cost to speech has been carefully calculated.69 The standard therefore sits between the least-restrictive-means test and rational-basis review.
Obvious alternatives count as evidence that a fit is unreasonable, but they do not become a separate rule demanding the narrowest option. In Rubin v. Coors Brewing Co., the government could have limited the alcohol content of beer directly, and the label ban was not sufficiently tailored to its goal, leaving an insufficient fit between means and ends.70 Critics of Posadas made the same point: Puerto Rico could have advanced its goals through less speech-restrictive means, among them an outright ban on casino gambling by residents.71
A second line of cases tests fit by looking at what a regulation leaves untouched. In City of Cincinnati v. Discovery Network, Inc., the city barred 62 newsracks dispensing free publications that it treated as commercial handbills, while between 1,500 and 2,000 newspaper racks stayed on the same public property. The city invoked safety and appearance, but its line between commercial handbills and newspapers had no bearing on those interests: each rack is equally unattractive whatever it contains, and removing 62 racks from a total near 2,000 would produce only marginal gain. The Court found the case for a close fit so clearly lacking that the decision signified no new protection for commercial speech.72 The decision left cities able to regulate newsracks if they do their homework and refine their methods.73
Two decisions show tailoring that passes. In United States v. Edge Broadcasting Co., relying on Central Hudson and Fox, the Court upheld federal legislation regulating broadcast lottery advertising.74 The approved part of San Diego’s billboard ordinance in Metromedia, Inc. v. City of San Diego distinguished between types of commercial speech. Such distinctions apparently face only low-level scrutiny, while Cincinnati’s line between commercial and noncommercial speech drew intermediate scrutiny.75 The same ordinance failed as to noncommercial speech, because it allowed on-site commercial messages but not on-site noncommercial messages that would be no more distracting or unattractive, and so amounted to impermissible content-based regulation.76
Exam facts usually signal which pattern is in play. A total ban on a category of advertising, paired with a plausible disclosure alternative, points toward the result in Central Hudson. A regulation of a narrow setting such as a campus dormitory points toward the deference of Fox. Exemptions that leave equally harmful speech untouched, especially where the only difference is the speech’s commercial character, point toward Discovery Network.