OPINION

PAEZ, Circuit Judge:

In this appeal, we consider Plaintiff-Appellant Retail Digital Network, LLC's ("RDN") First Amendment challenge to California Business and Professions Code § 25503(f)-(h). Section 25503(f)-(h) prohibits alcohol manufacturers and wholesalers from providing anything of value to retailers in exchange for advertising their alcohol products. As a result of Section 25503(f)-(h), alcohol manufacturers and wholesalers refused to enter into advertising agreements with RDN — which placed advertisements in wine and spirit retail stores — and RDN filed suit for declaratory and injunctive relief against Defendant-Appellee Ramona Prieto ("Prieto") in her official capacity as Acting Director of the California Department of Alcoholic Beverage Control (the "ABC").1

This is not the first time we have considered such a challenge to Section 25503(h).2 Thirty years ago, in Actmedia, Inc. v. Stroh, 830 F.2d 957 (9th Cir. 1986), we rejected a First Amendment challenge to that provision. In rejecting the challenge, we applied the four-part test established by Central Hudson Gas & Electric Corp. v. Public Service Commission of New York, 447 U.S. 557, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980), for evaluating restrictions on commercial speech.

RDN argues that Actmedia is no longer good law because the Supreme Court's decision in Sorrell v. IMS Health Inc., 564 U.S. 552, 131 S.Ct. 2653, 180 L.Ed.2d 544 (2011), fundamentally altered the Central Hudson test by adopting a more demanding standard for assessing restrictions on commercial speech. We disagree. Reviewing de novo, we hold that Sorrell did not modify the Central Hudson standard. We reaffirm Actmedia's core holding, but we disapprove of Actmedia's reliance on California's interest in promoting temperance as a justification for Section 25503(h). We therefore affirm the district court's order granting summary judgment to Prieto, which correctly relied on Actmedia.

I.

A.

RDN installed and operated seven-foot digital screen displays in one-hundred wine and spirit retail stores throughout Southern California. On its screens, RDN ran advertisements on a two-minute loop comprised of fifteen-second advertisements. RDN sold advertising slots to various companies, and, in turn, agreed to share a portion of its revenue with the retail stores.

RDN agreed to run advertisements for two alcohol manufacturers, St-Germain and Moët Hennessy. Those agreements, however, were short-lived because St-Germain and Moët Hennessy feared that the ABC would enforce Section 25503(f)-(h) against them.3 Several other alcohol manufacturers and wholesalers, including Anheuser-Busch, Beam Global, Diageo, Jack Daniel's Tennessee Whiskey, MillerCoors, and Skyy refused to contract with RDN because of the same concern.

As a result of its inability to secure advertisement placements from alcohol manufacturers and wholesalers, RDN filed this lawsuit against Prieto, seeking a declaration that Section 25503(f)-(h) is unconstitutional under the First Amendment because it impermissibly restricts commercial speech, and an injunction enjoining enforcement of those subsections.

Prieto moved for summary judgment, arguing that RDN lacked standing and that even if standing existed, she was entitled to judgment under Actmedia, which rejected a similar First Amendment challenge. The district court concluded that RDN had standing, but that Prieto nonetheless was entitled to summary judgment because Actmedia was not clearly irreconcilable with Sorrell or other subsequent Supreme Court cases.

RDN timely appealed. A three-judge panel of this court reversed and remanded for further proceedings.4 Retail Digital Network, LLC v. Appelsmith, 810 F.3d 638, 642 (9th Cir. 2016). The panel held that "Sorrell requires heightened judicial scrutiny of content-based restrictions on non-misleading commercial speech regarding lawful products, rather than the intermediate scrutiny applied to [Section 25503(h)] in Actmedia." Id. After the panel issued its opinion, a majority of nonrecused active judges voted to rehear this case en banc. Retail Digital Network, LLC v. Gorsuch, 842 F.3d 1092, 1092 (9th Cir. 2016).

B.

To understand the purpose of Section 25503(f)-(h), some historical context is necessary. Section 25503 is part of California's Alcoholic Beverage Control Act, California Business and Professions Code §§ 23000, et seq., and was adopted to prevent the resurgence of tied-houses following repeal of the Eighteenth Amendment. See Cal. Beer Wholesalers Ass'n v. Alcoholic Beverage Control Appeals Bd., 5 Cal.3d 402, 96 Cal.Rptr. 297,487 P.2d 745, 748 (1971). The term "tied-houses" refers to retailers and saloons that are controlled by "larger manufacturing or wholesale interests."5 Id. The California Supreme Court has recognized that tied-houses pose two particular dangers: "the ability and potentiality of large firms to dominate local markets through vertical and horizontal integration and the excessive sales of alcoholic beverages produced by the overly aggressive marketing techniques of larger alcoholic beverage concerns." Id. (citations omitted).

To prevent the formation of tied-houses after Prohibition, California enacted laws, including Section 25503, that established a triple-tiered alcohol distribution scheme, pursuant to which "[m]anufacturing interests were to be separated from wholesale interests; [and] wholesale interests were to be segregated from retail interests." Id.; see also Cal. Bus. & Prof. Code §§ 25500-25512. "In short, business endeavors engaged in the production, handling, and final sale of alcoholic beverages were to be kept `distinct and apart.'" Cal. Beer Wholesalers Ass'n, 96 Cal.Rptr. 297, 487 P.2d at 748 (quoting 25 Op. Cal. Att'y Gen. 288, 289 (1955)). In 2015, the California legislature reaffirmed its interest in preserving a triple-tiered distribution scheme by amending another tied-house provision, California Business and Professions Code § 25500.1, to provide:

2015 Cal. Stat. Ch. 408.

As we observed in Actmedia, Section 25503(h) addresses the California legislature's specific "concern that advertising payments could be used to conceal illegal payoffs to alcoholic beverage retailers," thereby undermining the triple-tiered distribution scheme. 830 F.2d at 967. That concern "appears to have been widely held at the time of [Section 25503(h)'s] enactment...." Id.

II.

A.

As noted above, in Actmedia we rejected a First Amendment challenge to Section 25503(h), concluding that it passed constitutional muster under Central Hudson. 830 F.2d at 968. Like RDN, the plaintiff in Actmedia was an advertising middle-man: Actmedia leased advertising space on shopping carts, and placed other companies' advertisements on the shopping carts. Id. at 958. Actmedia entered into an agreement with an alcohol manufacturer, the Adolph Coors Company ("Coors"), to advertise Coors beer on supermarket shopping carts. Id. at 961. The ABC, however, determined that Coors had violated Section 25503(h) by engaging in such conduct, and initiated an administrative proceeding, threatening to revoke Coors' California beer and wine license. Id. After remedying the violation, Coors terminated its use of Actmedia's services. Id.

In response to the ABC's action against Coors, Actmedia filed a lawsuit seeking a declaration that Coors' conduct did not violate Section 25503(h), and even if it did, the provision impermissibly restricted commercial speech under the First Amendment. Id. The ABC prevailed in the district court, and Actmedia appealed. Id. at 958.

We affirmed, and in doing so, relied on Central Hudson to hold that Section 25503(h) was constitutional under the First Amendment. See id. at 965-68. In Central Hudson, the Supreme Court reiterated that the First Amendment accords some protection to commercial speech,6 but that "[t]he Constitution [ ] accords a lesser protection to commercial speech than to other constitutionally guaranteed expression." 447 U.S. at 562-63, 100 S.Ct. 2343 (emphasis added). Central Hudson therefore established a four-part framework for assessing restrictions on commercial speech:

Id. at 566, 100 S.Ct. 2343. The Central Hudson analysis is commonly referred to as "intermediate scrutiny." See Fla. Bar v. Went For It, Inc., 515 U.S. 618, 623, 115 S.Ct. 2371, 132 L.Ed.2d 541 (1995) ("Mindful of these concerns, we engage in `intermediate' scrutiny of restrictions on commercial speech, analyzing them under the framework set forth in Central Hudson....."); see also Ashutosh Bhagwat, The Test that Ate Everything: Intermediate Scrutiny in First Amendment Jurisprudence, 2007 U. Ill. L. Rev. 783, 784-85 (explaining that the Court has created three tiers — rational basis review, intermediate scrutiny, and strict scrutiny — "to structure constitutional analysis" in free speech cases).

In applying the Central Hudson factors in Actmedia, we noted there was no dispute that the Coors advertisements displayed by Actmedia concerned lawful activity and were not misleading, thus satisfying the first factor. 830 F.2d at 965. As to the second factor, we observed that "there is little question that California has a `substantial' interest in exercising its twenty-first amendment powers and regulating the structure of the alcoholic beverage industry in California...." Id. at 965-66. The third and fourth factors, however, required a more complex analysis. See id. at 966-68.

Regarding the third factor — whether Section 25503(h) materially and directly advances California's interests — we noted that Section 25503(h) "is primarily designed to prevent or limit a specific evil: the achievement of dominance or undue influence by alcoholic beverage manufacturers and wholesalers over retail establishments." Id. at 966. We held that Section 25503(h) directly and materially advances that interest by eliminating any danger of alcohol manufacturers and wholesalers circumventing the triple-tiered distribution scheme by using advertising payments to conceal illegal payoffs. Id. at 967. "By flatly proscribing such payments, California minimized the possibility that alcoholic-beverage manufacturers and wholesalers will obtain undue influence over retail establishments, resulting in increased vertical and horizontal integration of California's liquor industry." Id.

We also held that Section 25503(h) both indirectly and directly advanced California's additional interest in promoting temperance. Id. The provision indirectly advanced the state's interest in promoting temperance "[b]y reducing the possibility that [vertical and horizontal] integration will occur, along with such side effects as a proliferation of retail liquor establishments and the emergence of quotas for individual outlets...." Id. "Moreover," we held, "in reducing the quantity of advertising that is seen in retail establishments selling alcoholic beverages, the provision also directly furthers California's interest in promoting temperance." Id.

Regarding the fourth factor, we acknowledged that, arguably, California could prevent illegal payments "by careful policing of any advertising agreements made between retailers and manufacturers or wholesalers," but that "it would be impossible for an agency like [the] ABC to determine whether an advertising company had paid an inordinately high price to a retailer or to investigate whether a retailer had in fact provided more than merely advertising services." Id. In addition, we noted that such policing could interfere in the advertising process itself, and thereby create other First Amendment problems. Id. On balance, we held that Section 25503(h) was "as narrowly drawn as possible to effectuate" California's purpose of preventing the illegal payoffs that would undermine its three-tiered alcohol distribution system. Id. We also held that Section 25503(h) was narrowly drawn to achieve California's interest in promoting temperance because "to the extent that the California legislature has determined that point-of-purchase advertising is a direct cause of excessive alcohol consumption, limiting that advertising is `obviously the most direct and perhaps the only effective approach' available." Id. (quoting Metromedia, Inc. v. City of San Diego, 453 U.S. 490, 508, 101 S.Ct. 2882, 69 L.Ed.2d 800 (1981)). As to both interests, we emphasized Section 25503(h)'s limited scope, in that "[i]t prohibits only paid advertising in retail stores, not unpaid advertising in those stores or paid advertising anywhere else." Id. at 968.

B.

Actmedia was decided only six years after Central Hudson. In the years that have followed, the Supreme Court has engaged in considerable debate about the contours of First Amendment protection for commercial speech, and whether Central Hudson provides a sufficient standard. See, e.g., Greater New Orleans Broad. Ass'n v. United States, 527 U.S. 173, 197, 119 S.Ct. 1923, 144 L.Ed.2d 161 (1999) (Thomas, J., concurring in the judgment); 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 517, 116 S.Ct. 1495, 134 L.Ed.2d 711 (1996) (Scalia, J., concurring in part and concurring in the judgment); Rubin v. Coors Brewing Co., 514 U.S. 476, 493-97, 115 S.Ct. 1585, 131 L.Ed.2d 532 (1995) (Stevens, J., concurring in the judgment); City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 431, 113 S.Ct. 1505, 123 L.Ed.2d 99 (1993) (Blackmun, J., concurring). The Court nonetheless has continued to follow the Central Hudson framework.7 See, e.g., Thompson v. W. States Med. Ctr., 535 U.S. 357, 367-77, 122 S.Ct. 1497, 152 L.Ed.2d 563 (2002); Greater New Orleans Broad., 527 U.S. at 183-96, 119 S.Ct. 1923; Edenfield v. Fane, 507 U.S. 761, 767-73, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993); Discovery Network, 507 U.S. at 416-28, 113 S.Ct. 1505. The Court has also clarified Central Hudson's fourth factor by making clear that it does not require satisfaction of a "least-restrictive-means standard," but rather requires "a fit between the legislature's ends and the means chosen to accomplish those ends, [ ]a fit that is not necessarily perfect, but reasonable[,] ... a means narrowly tailored to achieve the desired objective." Bd. of Trs. of State Univ. of N.Y. v. Fox, 492 U.S. 469, 477, 480, 109 S.Ct. 3028, 106 L.Ed.2d 388 (1989) (citations and internal quotation marks omitted).

What the Supreme Court repeatedly has declined to do, however, is to fundamentally alter Central Hudson's intermediate scrutiny standard. See, e.g., Thompson, 535 U.S. at 367-68, 122 S.Ct. 1497 ("Although several Members of the Court have expressed doubts about the Central Hudson analysis and whether it should apply in particular cases, there is no need in this case to break new ground.") (citations omitted); see also Lorillard Tobacco Co. v. Reilly, 533 U.S. 525, 554-55, 121 S.Ct. 2404, 150 L.Ed.2d 532 (2001). RDN argues that Sorrell imposed such a change.

Indeed, RDN argues that for content- or speaker-based regulations of commercial speech, Sorrell requires courts to apply a greater level of scrutiny than Central Hudson previously required. For support, RDN points to Sorrell's references to "heightened scrutiny," which it interprets to mean scrutiny greater than intermediate scrutiny. RDN further relies on the Sorrell Court's statement that "the out-come is the same whether a special commercial speech inquiry or a stricter form of judicial scrutiny is applied," 564 U.S. at 571, 131 S.Ct. 2653, to argue that the Court applied a heightened form of scrutiny without elaborating as to its form.

RDN reads Sorrell too expansively. Contrary to RDN's argument, Sorrell did not mark a fundamental departure from Central Hudson's four-factor test, and Central Hudson continues to apply.8

In Sorrell, the Supreme Court addressed a First Amendment challenge to a Vermont statute, Vt. Stat. Ann. tit. 18, § 4631(d), that restricted the sale, disclosure, and use of pharmacy records for marketing purposes. 564 U.S. at 557, 131 S.Ct. 2653. At the outset of its legal discussion, the Court addressed Vermont's argument that the law did not regulate speech. See id. at 563-71, 131 S.Ct. 2653. After concluding that the law at issue regulated speech, the Court turned to whether the law satisfied the Central Hudson factors, holding that it failed at both the third and fourth stages. See id. at 572, 131 S.Ct. 2653.

Notably, the Sorrell Court referred to "heightened scrutiny" within the context of deciding whether Section 4631(d) regulated speech whatsoever. See id. at 563-71, 131 S.Ct. 2653. That discussion was responsive to Vermont's argument that the restriction was directed at commerce, conduct, and access to information, rather than speech, and was therefore subject to rational basis review. See id. at 566-71, 131 S.Ct. 2653. It thus follows that the "heightened" scrutiny to which the Court referred is the scrutiny that courts apply to speech regulations — as opposed to the rational basis review that courts apply to non-speech regulations of commerce and non-expressive conduct. There is nothing novel in Sorrell's use of the term "heightened scrutiny" to distinguish from rational basis review. See Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 640-41, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) (rejecting, in the First Amendment context, the government's contention that rational basis review applies and concluding that "some measure of heightened First Amendment scrutiny is demanded"); see also Minn. State Bd. for Cmty. Colls. v. Knight, 465 U.S. 271, 291, 104 S.Ct. 1058, 79 L.Ed.2d 299 (1984) ("There being no other reason to invoke heightened scrutiny, the challenged state action need only rationally further a legitimate state purpose to be valid under the Equal Protection Clause.") (internal quotation marks omitted). Nor is Sorrell the first time the Court has referred to intermediate scrutiny as "heightened" scrutiny. See, e.g., Clark v. Jeter, 486 U.S. 456, 463, 465, 108 S.Ct. 1910, 100 L.Ed.2d 465 (1988) (referring to intermediate scrutiny as "heightened scrutiny" in the equal protection context, which similarly distinguishes between three levels of scrutiny — rational basis, intermediate, and strict).

To the extent that Sorrell addressed whether the law at issue imposed a content- or speaker-based burden, that discussion also appeared in the Court's consideration of whether Section 4631(d) was a speech regulation in the first instance. See 564 U.S. at 563-71, 131 S.Ct. 2653. Indeed, the Court relied on the content- and speaker-based burden imposed by Vermont's statute to conclude that the statute results in "more than an incidental burden on protected expression," and thus implicates the First Amendment, which requires scrutiny greater than rational basis review. Id. at 567, 131 S.Ct. 2653. Accordingly, the content- and speaker-based distinction is relevant only insofar as parties dispute whether the law regulates speech.9 See also United States v. Caronia, 703 F.3d 149, 180 (2d Cir. 2012) (Livingston, J., dissenting) ("Every commercial speech case, by its very nature, involves both content- and speaker-based speech restrictions.").

Further underscoring our conclusion, the Court cited Discovery Network, 507 U.S. at 418, 113 S.Ct. 1505, for the proposition that "heightened judicial scrutiny" applies to content-based burdens on protected expression. Sorrell, 564 U.S. at 565, 131 S.Ct. 2653. But Discovery Network did not adopt a new heightened standard for commercial speech; rather, the Court applied Central Hudson's four-part framework for assessing restrictions on commercial speech. See Discovery Network, 507 U.S. at 416, 113 S.Ct. 1505. In other words, it applied intermediate scrutiny. We are therefore not persuaded by RDN's first argument that Sorrell's references to "heightened scrutiny" mean something greater than intermediate scrutiny applies in commercial speech cases.

RDN's second argument — that the Court implied a greater level of scrutiny applies by stating, "the outcome is the same whether a special commercial speech inquiry or a stricter form of judicial scrutiny is applied" — fares no better. See Sorrell, 564 U.S. at 571, 131 S.Ct. 2653. In Sorrell, the Court entertained the potential application of a "stricter form of judicial scrutiny," but ultimately applied Central Hudson, deeming it unnecessary to determine whether a stricter form of scrutiny would be appropriate because Section 4631(d) failed to satisfy Central Hudson intermediate scrutiny. See id.

Turning to the Central Hudson factors, the Court focused on the third and fourth factors, and described them thusly: "the State must show at least that the statute directly advances a substantial governmental interest and that the measure is drawn to achieve that interest."10 Id. at 572, 131 S.Ct. 2653 (citing Fox, 492 U.S. at 480-81, 109 S.Ct. 3028; Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343). In analyzing those factors, the Court framed its discussion of Section 4631(d) in language that reflects classic Central Hudson considerations.11 Moreover, in its analysis of the third and fourth factors, the Court relied on cases that applied Central Hudson. Sorrell, 564 U.S. at 573-79, 131 S.Ct. 2653 (citing, e.g., Thompson, 535 U.S at 373-74, 122 S.Ct. 1497 (applying Central Hudson, concluding that the regulation at issue failed to "directly advance" the government's asserted interests, and rejecting as improper the government's interest in suppressing "the dissemination of truthful commercial information in order to prevent members of the public from making bad decisions with the information"); Greater New Orleans Broad., 527 U.S. at 195, 119 S.Ct. 1923 (applying Central Hudson and striking down an incoherent statute that failed to directly and materially advance the government's asserted interest in limiting consumer demand for casino gambling); Edenfield, 507 U.S. at 767, 775, 113 S.Ct. 1792 (observing that commercial speech is entitled to First Amendment protection, and articulating Central Hudson's four-factor test); Discovery Network, 507 U.S. at 426, 428, 113 S.Ct. 1505 (noting that "commercial speech can be subject to greater governmental regulation than noncommercial speech" but concluding that the city failed to establish Central Hudson's fourth factor requiring a "fit" between the city's goals and its chosen means)).

There is one other consideration that the Supreme Court acknowledged in Sorrell that persuades us that Central Hudson continues to set the standard for assessing restrictions on commercial speech. That consideration centers on one of the core principles that animates the Court's approach to commercial speech — that commercial speech may be subject to greater regulation than non-commercial speech. See Sorrell, 564 U.S. at 579, 131 S.Ct. 2653 ("[T]he government's legitimate interest in protecting consumers from `commercial harms' explains `why commercial speech can be subject to greater governmental regulation than noncommercial speech.'") (quoting Discovery Network, 507 U.S. at 426, 113 S.Ct. 1505). Requiring greater-than-intermediate yet lesser-than-strict scrutiny would both diminish that principle and impose an inscrutable standard. See Fox, 492 U.S. at 477, 109 S.Ct. 3028 ("The ample scope of regulatory authority ... would be illusory if it were subject to a least-restrictive-means requirement, which imposes a heavy burden on the State.").

In any event, because Sorrell applied Central Hudson, there is no need for us to "craft an exception to the Central Hudson standard." Rubin, 514 U.S. at 482 n.2, 115 S.Ct. 1585. "If a precedent of [the Supreme Court] has direct application in a case, ... the Court of Appeals should follow the case which directly controls, leaving to [the Supreme Court] the prerogative of overruling its own decisions." Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989).

We are not alone in arriving at this conclusion. In commercial speech cases post-Sorrell, the Second, Fourth, Sixth, and Eighth Circuits similarly have, at bottom, continued to apply Central Hudson.12 See 1-800-411-Pain Referral Serv., LLC v. Otto, 744 F.3d 1045, 1055 (8th Cir. 2014) ("The upshot is that when a court determines commercial speech restrictions are content- or speaker-based, it should then assess their constitutionality under Central Hudson."); Educ. Media Co. at Va. Tech, Inc. v. Insley, 731 F.3d 291, 298 (4th Cir. 2013) ("However, like the Court in Sorrell, we need not determine whether strict scrutiny is applicable here, given that, as detailed below, we too hold that the challenged regulation fails under intermediate scrutiny set forth in Central Hudson."); Caronia, 703 F.3d at 164 ("[W]e conclude the government cannot justify a criminal prohibition of off-label promotion even under Central Hudson's less rigorous intermediate test."); Discount Tobacco City & Lottery, Inc. v. United States, 674 F.3d 509, 533 (6th Cir. 2012) ("Thus, this Court is left with Plaintiffs' `action directed to consumers through the media or otherwise... respecting the product,' 21 U.S.C. § 387k(b)(2)(A), which we find to be commercial speech properly analyzed under Central Hudson.").

C.

RDN contends that even if Sorrell did not fundamentally change the Central Hudson test, Actmedia incorrectly concluded that Section 25503(h) satisfies the third and fourth factors. We disagree. Actmedia correctly held that Section 25503(h) survives scrutiny under Central Hudson, and therefore the district court properly granted Prieto's motion for summary judgment. We disapprove, however, of Actmedia's reliance on California's interest in promoting temperance as a justification for Section 25503(h).

To the extent that Actmedia upheld Section 25503(h) on the basis that it directly and materially advances the State's interest in maintaining a triple-tiered distribution scheme, we agree with the court's sound analysis. Furthermore, we concur that Section 25503(h) is sufficiently tailored to advance that interest. Section 25503(h) serves the important and narrowly tailored function of preventing manufacturers and wholesalers from exerting undue and undetectable influence over retailers. Without such a provision, retailers and wholesalers could side-step the triple-tiered distribution scheme by concealing illicit payments under the guise of "advertising" payments. Although RDN argues that the numerous exceptions to Section 25503(f)-(h) undermine its purpose,13 RDN fails to recognize that the exceptions do not apply to the vast majority of retailers, and they therefore have a minimal effect on the overall scheme. This stands in stark contrast to cases in which conflicting regulations have rendered the regulatory scheme "irrational[]," Rubin, 514 U.S. at 488, 115 S.Ct. 1585, or where the regulatory scheme is "so pierced by exemptions and inconsistencies" that it lacks "coheren[ce]," Greater New Orleans Broad., 527 U.S. at 190, 195, 119 S.Ct. 1923.

We do not, however, reach the same conclusion with respect to Actmedia's holding that Section 25503(h) directly and materially advances California's interest in promoting temperance. Even assuming that promoting temperance is a substantial interest, Actmedia erroneously concluded that Section 25503(h) directly and materially advances that interest by "reducing the quantity of advertising that is seen in retail establishments selling alcoholic beverages." 830 F.2d at 967. Section 25503(h) applies solely to advertising in retail establishments, which comprises a small portion of the alcohol advertising visible to consumers. In addition, it prohibits only paid advertisements, and therefore, by its terms, does not reduce the quantity of advertisements whatsoever. See, e.g., Rubin, 514 U.S. at 488, 115 S.Ct. 1585 ("The failure to prohibit the disclosure of alcohol content in advertising, which would seem to constitute a more [effective approach than prohibiting the same information on labels], makes no rational sense...."). If California sincerely wanted to materially reduce the quantity of alcohol advertisements viewed by consumers, surely it could have devised a more direct method for doing so. See, e.g., Greater New Orleans Broad., 527 U.S. at 192, 119 S.Ct. 1923 ("There surely are practical and nonspeech-related forms of regulation ... that could more directly and effectively alleviate some of the social costs of casino gambling."). Actmedia otherwise concluded that Section 25503(h) only indirectly advances California's interest in promoting temperance by preventing tied-houses. 830 F.2d at 967. We agree with that conclusion, but indirect advancement fails to satisfy Central Hudson's third factor. See Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343. We therefore disapprove of Actmedia's reliance on promoting temperance as a justification for Section 25503(h).14 See id.

Nonetheless, as we already have stated, disapproving the portion of Actmedia that relies on temperance does not negate the sound and well-reasoned conclusion that Section 25503(h) withstands First Amendment scrutiny because it directly and materially advances the State's interest in maintaining a triple-tiered market system, and because there is a sufficient fit between that interest and the legislative scheme. Actmedia thus forecloses RDN's First Amendment challenge to Section 25503(f)-(h).

AFFIRMED.

THOMAS, Chief Judge, dissenting:

I respectfully dissent. The key issue in this case is whether Actmedia Inc. v. Stroh, 830 F.2d 957 (9th Cir. 1986), is compatible with subsequent Supreme Court authority addressing commercial speech, culminating with Sorrell v. IMS Health, Inc., 564 U.S. 552, 131 S.Ct. 2653, 180 L.Ed.2d 544 (2011). Sorrell held that content-based restrictions on truthful, non-misleading commercial speech advertising legal goods and services require "heightened judicial scrutiny," rather than traditional intermediate scrutiny under Central Hudson Gas & Electric Corp. v. Public Service Commission of New York, 447 U.S. 557, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980). Sorrell, 564 U.S. at 565-66, 131 S.Ct. 2653. Of course, the ultimate determination as to whether Sorrell altered the Central Hudson test is entirely up to the Supreme Court. However, I think the most reasonable reading of Sorrell is that it did.1 Therefore, I would hold that Actmedia is irreconcilable with Sorrell, expressly overrule Actmedia, and remand to allow the district court to conduct a purposive inquiry into the restricting statute, as required by Sorrell.

In previous decisions, the Supreme Court considered post-hoc rationalizations proffered by the government when assessing challenged restrictions. See, e.g., Edenfield v. Fane, 507 U.S. 761, 768, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993) ("Neither will we turn away if it appears that the stated interests are not the actual interests served by the restriction."). Sorrell departs from this practice and requires a purposive inquiry into the restricting statute.2

Sorrell not only employed the phrase "heightened scrutiny" repeatedly in the opinion but also significantly added "drawn to achieve" language to the fourth prong of Central Hudson,3 as noted by Justice Breyer's interpretation of the majority opinion. See, e.g., Sorrell, 564 U.S. at 588, 131 S.Ct. 2653 (Breyer, J., dissenting) ("Nor has this Court ever previously applied any form of `heightened' scrutiny in any even roughly similar case.... The Court (suggesting a standard yet stricter than Central Hudson) says that we must give content-based restrictions that burden speech `heightened' scrutiny.") (both emphases in original). A contrary interpretation would effectively write out the "drawn to achieve" language, which Sorrell added to the Central Hudson test, and render the use of "heightened scrutiny" in the opinion entirely superfluous. Further, Sorrell's two-step test would make no sense if intermediate scrutiny applies equally to content- and speaker-based regulations and non-content- and non-speaker-based regulations.

I therefore suggest that, under Sorrell, the purposive inquiry should be applied within the framework of the four-factor Central Hudson test, heightening the scrutiny historically applied at Central Hudson's fourth factor. Accordingly, in my view, with respect to the fourth factor, the government must establish that the challenged statute "directly advances a substantial governmental interest and that the measure is drawn to achieve that interest," Sorrell, 564 U.S. at 572, 131 S.Ct. 2653 (emphasis added). Therefore, a court must determine whether the government's asserted substantial interests for the restriction are a pretext for the government's desire "to suppress a disfavored message." Id. If so, the law is invalid.

The inquiry that I suggest does not test virtue — rather, it tests congruence. Thus, the first step is to discern the government's purpose for regulating, and the second step is to determine whether there is "a fit between the legislature's ends and the means chosen to accomplish those ends." Sorrell, 564 U.S. at 572, 131 S.Ct. 2653 (citing Bd. of Tr. of State Univ. of N.Y. v. Fox, 492 U.S. 469, 480, 109 S.Ct. 3028, 106 L.Ed.2d 388 (1989)). A pretextual regulation could conceivably survive Central Hudson analysis (if it suppresses no more speech than necessary to vindicate an asserted substantial government interest), but fail Sorrell review (if the regulation's underlying purpose was not to vindicate that substantial interest but rather to suppress a message disfavored by the government). In addition, in my view, a purposive inquiry means that the government cannot rely on post-hoc rationalizations to justify the statute.

Because the district court did not analyze the statute under the heightened scrutiny required by Sorrell, I would reverse and remand. Therefore, I respectfully dissent.