Harvard Law Review Harvard Law Review Harvard Law Review

Government Speech

Walker v. Texas Division, Sons of Confederate Veterans, Inc.

The Supreme Court’s government speech doctrine offers a constitutional escape hatch — a means by which government and courts may disregard the boundaries that the Free Speech Clause of the First Amendment would otherwise impose. So long as the government comports with other constitutional requirements, it may regulate its own speech however it wishes, even when that speech involves the expression of private individuals. Courts and scholars have justified this exemption on the ground that government must be able to communicate and therefore must be able to control the content of its communications.1×1. See, e.g., Andy G. Olree, Identifying Government Speech, 42 Conn. L. Rev. 365, 367–68 (2009). But the Supreme Court’s decision last Term in Walker v. Texas Division, Sons of Confederate Veterans, Inc.2×2. 135 S. Ct. 2239 (2015). stretches this rationale to its limits. In holding that hundreds of specialty license plate designs, created at the request (and expense) of private individuals and groups, are government speech and thus unprotected by the Free Speech Clause,3×3. See id. at 2246. the decision reveals a misalignment between the test the Court applies to identify government speech and the purported rationale for the doctrine.

Texas gives its citizens a choice: a driver can either display an ordinary Texas license plate on her vehicle, choose from an assortment of specialty plate designs created by the State, or, if none of these fits her taste, design her own license plate.4×4. Id. at 2244–45. These privately sponsored designs are subject to approval by the Texas Department of Motor Vehicles Board.5×5. Id. Under the specialty plate statutory scheme, the Board “may refuse to create a new specialty license plate if the design might be offensive to any member of the public . . . or for any other reason established by rule.”6×6. Tex. Transp. Code Ann. § 504.801(c) (West 2015). The Board has approved the majority of applications, with over 350 designs having been authorized for Texas roads.7×7. See Walker, 135 S. Ct. at 2255, 2260 (Alito, J., dissenting). Existing specialty plates feature the logos and mottos of, among many other things, the Florida Gators, the Girl Scouts, and Mothers Against Drunk Driving.8×8. Id. at 2245 (majority opinion); id. at 2257 (Alito, J., dissenting).

A symbol that does not appear on Texas plates, however, is the Confederate flag. In 2009, the Texas Division of Sons of Confederate Veterans (SCV) applied for a specialty license plate featuring its logo, a square Confederate battle flag framed by the words “Sons of Confederate Veterans 1896.”9×9. Id. at 2245 (majority opinion). As with all specialty license plates, the proposed plate design also included a license plate number and the State’s name and silhouette.10×10. Id. Following a period of public comment, the Board unanimously voted against issuing the plate.11×11. Id. It justified its decision by noting that “many members of the general public find the design offensive, and . . . such comments are reasonable. . . . [A] significant portion of the public associate the confederate flag with organizations advocating expressions of hate directed toward people or groups that [are] demeaning to those people or groups.”12×12. Id. (quoting Joint Appendix, Walker, 135 S. Ct. 2239 (No. 14-144), 2014 WL 7498018, at *64–65). In response, the SCV and two of its officers brought suit against the chairman and members of the Board, claiming that the denial contravened the Free Speech Clause of the First Amendment and seeking an injunction requiring the Board to approve their proposed design.13×13. Id.

The district court found for the Board.14×14. See Tex. Div., Sons of Confederate Veterans, Inc. v. Vandergriff, No. A-11-CA-1049, 2013 WL 1562758, at *26 (W.D. Tex. Apr. 12, 2013). It determined that although the specialty license plates were private speech, the Texas program provided a nonpublic forum, and the Board’s rejection of the SCV’s design was a reasonable content-based, rather than viewpoint-based, restriction, which was thus permissible under the First Amendment.15×15. See id. at *11–24.

On appeal, a divided panel of the Fifth Circuit reversed.16×16. Tex. Div., Sons of Confederate Veterans, Inc. v. Vandergriff, 759 F.3d 388, 390 (5th Cir. 2014). Judge Prado wrote the majority opinion and was joined by Judge Elrod. That court too held that the Texas specialty plates constitute private speech, but found that by refusing to approve the SCV’s design, Texas had engaged in unconstitutional viewpoint discrimination.17×17. Id. at 395, 400. Judge Smith argued in dissent that the plate designs were in fact government speech and fell outside the bounds of the First Amendment.18×18. Id. at 401 (Smith, J., dissenting).

The Supreme Court reversed, holding that the license plates were government speech and that Texas was thus free to control their content.19×19. Walker, 135 S. Ct. at 2253. Writing for the Court, Justice Breyer20×20. Justice Breyer was joined by Justices Thomas, Ginsburg, Sotomayor, and Kagan. began his analysis by noting that “[w]hen government speaks, it is not barred by the Free Speech Clause from determining the content of what it says.”21×21. Walker, 135 S. Ct. at 2245 (citing Pleasant Grove City v. Summum, 555 U.S. 460, 467–68 (2009)). That exemption is justified, he explained, in part by “the fact that it is the democratic electoral process that first and foremost provides a check on government speech,”22×22. Id. and because under an alternative interpretation of the clause, “government would not work.”23×23. Id. at 2246. Absent the ability to express those messages it chooses while excluding those it does not, he claimed, “[i]t is not easy to imagine how government could function.”24×24. Id. (quoting Summum, 555 U.S. at 468). Accordingly, the government is “entitled to promote a program, to espouse a policy, or to take a position,” and in doing so, it may discriminate on the basis of viewpoint without running afoul of the First Amendment.25×25. Id.

Having explained the doctrine’s rationale and implications, the Court then set forth a framework for identifying government speech grounded in Pleasant Grove City v. Summum.26×26. 555 U.S. 460. There, the Court had upheld the city’s rejection of a religious organization’s request to erect a monument in a city park, despite its having accepted other private groups’ proposals. Basing its determination on the historical use of the medium,27×27. See id. at 470–71. the fact that property owners rarely permit installation of monuments conveying unwanted messages,28×28. Id. at 471. and the city’s exercise of “selective receptivity” over which monuments to display,29×29. Id.; see also id. at 476. the Court found that the city had engaged in its own expressive conduct and that the selection of monuments was thus “best viewed as a form of government speech.”30×30. Id. at 464.

Applying Summum’s framework to the Texas license plate program, Justice Breyer found that the plates constituted government speech. First, he detailed the history of license plates, the evolution of their inclusion of state slogans, and the eventual introduction of specialty plates commissioned by the government to promote programs like “Read to Succeed” and “150 Years of Statehood,” concluding that “state speech has appeared on Texas plates for decades.”31×31. Walker, 135 S. Ct. at 2248.

Second, Justice Breyer claimed that Texans are likely to attribute the messages to the State because license plates “are often closely identified in the public mind with the [State].”32×32. Id. (alteration in original) (quoting Summum, 555 U.S. at 472). In support of that claim, he pointed out that the plates serve the official government purposes of registration and identification, that the state name appears on every plate, that car owners are required to display them, and that Texas owns both the plates and the designs themselves.33×33. Id. Because Texas license plates are “essentially[] government IDs,”34×34. Id. at 2249. persons who observe the IDs “routinely — and reasonably — interpret them as conveying some message on the [issuer’s] behalf.”35×35. Id. (alteration in original) (quoting Summum, 555 U.S. at 471). Moreover, this link to the state is likely the entire point, Justice Breyer claimed, of displaying one’s message on a license plate at all.36×36. Id. Texas specialty license plates are desirable precisely because they “may well . . . convey government agreement with the message displayed.”37×37. Id.

Finally, the Court found that, as in Summum, the Texas government maintained direct control over the messages displayed on license plates.38×38. Id. Board approval is required, and having resulted in the rejection of at least a dozen plate designs, that review constitutes “effective[] control[]” of the messages conveyed.39×39. Id. (quoting Summum, 555 U.S. at 473) (citing 43 Tex. Admin. Code §§ 217.45(i)(7)–(8), 217.52(b); Reply Brief for the Petitioners, Walker, 135 S. Ct. 2239 (No. 14-144), 2015 WL 1088969, at *10; Transcript of Oral Argument at 49–51, Walker, 135 S. Ct. 2239 (No. 14-144)). Together these factors showed that the license plates are “similar enough to the monuments in Summum” to fall into the same category — government speech.40×40. Id.

In closing, Justice Breyer claimed that the Court’s determination that license plates constitute government speech did not mean that they do not also implicate the free speech rights of individuals.41×41. Id. at 2252. Still, he maintained, those individuals cannot force the State to communicate on their behalf.42×42. Id. at 2253.

Justice Alito wrote in dissent.43×43. Justice Alito was joined by Chief Justice Roberts and Justices Scalia and Kennedy. In his view, the license plate designs were private speech that the Court’s decision wrongly “strip[ped] . . . of all First Amendment protection.”44×44. Walker, 135 S. Ct. at 2255 (Alito, J., dissenting). Though he appeared to agree with the Court’s premise that attribution was the key factor in identifying government speech, he disagreed with the Court’s determination that the messages on specialty license plates were likely to be interpreted as speech by the State.45×45. Id. at 2255–56. Relying on a thought experiment to demonstrate that no reasonable observer would believe a “Rather Be Golfing” plate reveals a state policy to promote golf over tennis or bowling,46×46. See id. at 2255. he argued that the majority had expanded the government speech doctrine to include not only speech by the government, but also government “blessing” of private speech.47×47. Id. at 2261. This type of communication, Justice Alito argued, implicated not the government speech doctrine, but limited public forum analysis.48×48. Id. On this view, Texas merely opened space up on its specialty plates for the messages of private speakers. These “little billboards,”49×49. Id. at 2256. he claimed, constituted a limited public forum in which the government is prohibited from engaging in viewpoint discrimination.50×50. Id. at 2262 (citing Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 829 (1995)). And the rejection of some designs based on the controversial nature of the views they express blatantly violated that rule.51×51. Id. at 2263.

Taking issue with each element of the Court’s application of Summum,52×52. See id. at 2259–60 (history); id. at 2260–61 (“selective receptivity”); id. at 2261 (limited space and permanence). Justice Alito claimed that the license plate designs could not plausibly be conceived as promoting the government’s views, and instead represented an attempt to raise revenue by giving individual speakers an opportunity to express private ones.53×53. See id. at 2255, 2261–62. The regulation of license plates may seem benign, Justice Alito warned, but allowing government to construct forums and then discriminate among the viewpoints that can be communicated within them forebodes more significant curtailment of speech.54×54. See id. at 2255–56.

Although the majority and dissent disagreed about the proper framework for determining whether viewers will assume the State or rather private actors are speaking,55×55. While Justice Breyer focused on application of the Summum factors, emphasizing the ties between each element and the likelihood of attribution to the State, see id. at 2247–50 (majority opinion), Justice Alito would have required something like a reasonable-observer assessment, noting that mere implied endorsement by the State is not enough, see id. at 2255, 2261 (Alito, J., dissenting). at core both opinions set forth attribution-based methodologies for identifying government speech. Yet neither approach aligns with the purported justification for the exemption that regulation of government speech enjoys from the strictures of the First Amendment — namely, that government must have the ability to communicate in support of its policies or to take a position, and that it must be able to discriminate among viewpoints when it speaks in order to do so. Identifying messages that might be fairly attributed to the government does not, on its own, speak to whether such an association would foil the government’s ability to disseminate its own messages. By contrast, in the compelled private speech context, the Court has synchronized justification and doctrine: the Constitution forbids compelled speech because it impermissibly interferes with one’s constitutional right of expression, so the core doctrinal inquiry asks whether such an interference has occurred. To bring analogous coherence to the government speech doctrine, the Court should similarly align the doctrine’s test with its justification by moving beyond a wholly attribution-based inquiry and instead requiring government to prove interference with its intended message.

The Court has conceived of an individual’s right to free speech as consisting of a “sphere of intellect and spirit which it is the purpose of the First Amendment to our Constitution to reserve from all official control.”56×56. W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943). This negative privilege to be free from government interference in the realm of private speech exists as a positive power in the realm of government speech — because government is comprised of individuals who have been elected through the democratic process, it should be entitled to voice its views on its constituents’ behalf. In both contexts, the Court has recognized that the compulsion of speech may interfere with these prerogatives.

For that reason, the government speech doctrine has, from its start, relied upon the notion that government must be free from the First Amendment’s prohibition of viewpoint discrimination in order to effectively communicate.57×57. See Rust v. Sullivan, 500 U.S. 173, 193 (1991). Justice Breyer took this principle a step further, arguing that without this freedom, government “would not work.”58×58. Walker, 135 S. Ct. at 2246. The imagined alternative is a government that must send its constituents a confusing set of pamphlets arguing both for and against a healthy diet or the availability of abortion.59×59. See id. In every proffered hypothetical, the concern is that the compulsion of government speech will impede the government’s messages. If the government must say things it does not want to say in order to say the things it does wish to say, the argument goes, it won’t be able to say anything at all.

These concerns resemble those presented in another license plate case decided thirty-eight years ago that implicated compelled private speech. In Wooley v. Maynard,60×60. 430 U.S. 705 (1977). the Supreme Court held that New Hampshire’s law requiring vehicle owners to display license plates stamped with the State’s motto “Live Free or Die” unconstitutionally compelled the speech of the plaintiff.61×61. Id. at 717. The Court there found that “[a] system which secures the right to proselytize religious, political, and ideological causes must also guarantee the concomitant right to decline to foster such concepts.”62×62. Id. at 714. Since then, the compelled speech doctrine has recognized that compelling one to communicate a message she does not wish to perpetuate can interfere with her free speech rights just as would restricting a message that she does wish to foster.63×63. See, e.g., Rumsfeld v. Forum for Acad. & Institutional Rights, Inc. (FAIR), 547 U.S. 47, 61 (2006).

The concerns motivating the government speech and compelled speech doctrines thus mirror one another. Each seeks to protect the prerogative — either the government power or the individual right — to speak by preventing the compulsion of unwanted speech, which is understood to interfere with the speaker’s ability to express his own intended messages. Justice Breyer invoked this symmetry in the closing lines of his opinion in Walker, writing that “just as [under Wooley] Texas cannot require SCV to convey ‘the State’s ideological message,’ SCV cannot force Texas to include a Confederate battle flag on its specialty license plates.”64×64. Walker, 135 S. Ct. at 2253 (citation omitted) (quoting Wooley, 430 U.S. at 715). Yet in spite of the congruent concerns underlying the government speech and compelled speech doctrines, the test applied in Walker to determine whether an impermissible interference has occurred diverges from the Court’s test in compelled speech cases. And because the Walker test fails to target only the types of compulsions that create the need for the government speech exemption, its approach lacks the logical coherence of the compelled speech doctrine.

Although the Court has recognized that the compulsion of speech can inherently invade individual free speech rights, it has not held every instance of compelled communication to violate the First Amendment. In some instances, the Court has found that the compelled accommodation of unwanted speech did not constitute speech at all.65×65. See, e.g., FAIR, 547 U.S. at 61–64. In others, it has held that the unlikelihood of actual association of the message with the involuntary speaker,66×66. See, e.g., id. at 65; Bd. of Educ. v. Mergens ex rel. Mergens, 496 U.S. 226, 250 (1990) (plurality opinion). or the ability of the speaker to effectively disassociate from the message,67×67. See PruneYard Shopping Ctr. v. Robins, 447 U.S. 74, 87 (1980). remedied the First Amendment concern. And in still others, the government’s interest in the speech was so strong as to justify its compulsion.68×68. See Branzburg v. Hayes, 408 U.S. 665, 700 (1972). Yet in the government speech context, the Court has treated the government’s prerogative to avoid compelled speech as absolute. As soon as the scale set to measure whether the speech is more that of the government than that of the citizen tips in favor of the State, the speech simply may not be compelled.69×69. See Walker, 135 S. Ct. at 2245 (“When government speaks, it is not barred by the Free Speech Clause from determining the content of what it says.”). This divergence is particularly striking because personal liberty and governmental efficacy are on the line in both circumstances. When speech implicates both government and private voices, the Free Speech Clause is always implicated. If compulsion is permitted, government must sacrifice some ability to communicate, and if it is forbidden, the individual must make that sacrifice. In the compelled private speech context, the dilemma is resolved through a balancing of interests. But with government speech, the Court has applied a categorical rule, based solely on attribution, under which private expression will always lose. To avoid this result, the Court should instead adopt a test similar to that used in its compelled speech cases, one designed to measure whether the risks against which the government speech doctrine was designed to protect— that is, that government will not be able to communicate in support of its policies, projects, or positions — are actually implicated by the case.

Compelled speech jurisprudence consists of cases involving the government either requiring a private actor to speak the government’s message,70×70. See, e.g., Wooley v. Maynard, 430 U.S. 705, 707 (1977). or requiring that actor to host or accommodate another speaker’s message.71×71. See, e.g., Hurley v. Irish-Am. Gay, Lesbian & Bisexual Grp. of Bos., Inc., 515 U.S. 557, 573–74 (1995). While the former type of compulsion is thought to inherently interfere with free speech rights and is presumptively invalid (subject to strict scrutiny), the latter triggers a more complex analysis. The Court has found compelled accommodation of speech to infringe on free speech rights only where it actually interferes with the speaker’s intended message. For an interference to so qualify, the message impeded must be sufficiently expressive to merit First Amendment protection.72×72. See id. at 568–69. The likelihood that the accommodated message will be attributed to the speaker is one aspect of this determination,73×73. See Rumsfeld v. Forum for Acad. & Institutional Rights, Inc., 547 U.S. 47, 65 (2006) (“[S]tudents can appreciate the difference between speech a school sponsors and speech the school permits because legally required to do so . . . .”). as is whether the speaker could easily disassociate from the message.74×74. See id. at 60, 64–65 (citing PruneYard Shopping Ctr. v. Robins, 447 U.S. 74 (1980)).

By aligning the doctrinal test with the purported motivation for the compelled speech doctrine, the Court has prohibited compulsion only where, under its analysis, the compulsion would in fact invade the speech rights of the objecting individual, either because it forces her to foster a message she does not believe,75×75. See, e.g., Wooley, 430 U.S. at 714–17; Agency for Int’l Dev. v. All. for Open Soc’y Int’l, Inc., 133 S. Ct. 2321, 2330 (2013). or because it interferes with her ability to communicate her own message.76×76. See Hurley, 515 U.S. at 566; Pac. Gas & Elec. Co. v. Pub. Utils. Comm’n of Cal., 475 U.S. 1, 20–21 (1986) (plurality opinion); Miami Herald Publ’g Co. v. Tornillo, 418 U.S. 241, 258 (1974); see also FAIR, 547 U.S. at 63–64 (“The compelled-speech violation in each of our prior cases . . . resulted from the fact that the complaining speaker’s own message was affected by the speech it was forced to accommodate.” Id. at 63.). Accordingly, the doctrine’s application, and the sacrifice of government efficacy, is limited to only those instances where protection is necessary to preserve the speaker’s ability to control her own speech.

In contrast, the test the Court applies to identify government speech asks only whether the speech at issue may be more fairly described as that of the government or of a private party.77×77. See Note, The Curious Relationship Between the Compelled Speech and Government Speech Doctrines, 117 Harv. L. Rev. 2411, 2417 (2004) (“[I]dentifying who is speaking is central to the analysis.”). Indeed, courts have devoted many pages to assessing the difficulty and wisdom of making that determination where the speech involves the participation of both private entities and the government. See, e.g., Sons of Confederate Veterans, Inc. v. Comm’r of the Va. Dep’t of Motor Vehicles, 305 F.3d 241, 244–45 (4th Cir. 2002) (Luttig, J., respecting the denial of rehearing en banc) (“[Circuit courts] have struggled . . . because they have assumed, in oversimplification, that all speech must be either that of a private individual or that of the government, and that a speech event cannot be both private and governmental at the same time.”). To be sure, the Summum factors — history of the medium, attribution, and selective control — may assist in answering that question. But these factors do nothing to identify whether the risks justifying the government speech doctrine’s existence are present in a given case.78×78. Some scholars have concluded that in many of the Court’s government speech cases, they were not. See, e.g., Steven G. Gey, Why Should the First Amendment Protect Government Speech When the Government Has Nothing to Say?, 95 Iowa L. Rev. 1259, 1259 (2010) (“[T]he Court has resorted to its new government speech doctrine [in] situations in which the government’s ability to communicate with the public would not have been inhibited in any way if such a doctrine did not exist. The Court has even relied on its new government speech doctrine in several cases in which the government was communicating either ambiguously or not at all.”). A more logically consistent test, like the one employed in the compelled speech context, would better cohere with the doctrine’s fundamental rationale. In particular, it would ask whether the government’s forced acceptance of the proposed message would in fact inhibit its ability to express itself and, in turn, “to work.”

Applied in Walker, this test would have homed in on whether approving the SCV’s design would have actually prevented Texas from communicating a message it intended to send. Such a test might, like the one applied in the compelled private speech context, include assessment of whether there in fact exists an intended “articulable” message to be interfered with in the first place, whether the speech at issue is likely to be attributed to the government, and whether the government is able to effectively disassociate from the speech in question, thus avoiding any interference with its own message.

Here, Texas might have claimed its intended message was one of equality and unity among all its citizens, and that accommodation of the Confederate flag on its license plates would have undermined that expression. If the Court were to have found that message sufficiently expressive, it would have moved on to analyzing whether the compelled accommodation would in fact have impeded the State’s communication. Only then would the Justices have reached their debate on the proper means of answering the attribution inquiry, as well as the question of whether, by placing a small disclaimer on the bottom of each plate, Texas could have effectively disassociated from the SCV’s design.79×79. See FAIR, 547 U.S. at 64–65. It is far from clear that the result of these analyses would be consistent for every message the government asserts it intended, or that every type of plate design would fare similarly on the question of attribution80×80. See Larry Tribe, The Constitution Writ Large, Part One, Balkinization (July 13, 2015, 1:40 PM), h t t p : / / b a l k i n . b l o g s p o t . c o m / 2 0 1 5 / 0 7 / t h e – c o n s t i t u t i o n – w r i t – l a r g e – p a r t – o n e . h t m l [ h t t p : / / perma.cc/XWA3-T8MZ]. — while a NASCAR plate would likely not sufficiently interfere with a purported government message promoting safe driving, a design advocating drag racing just might. Yet of these factors, the Court’s opinions grappled only with attribution, quarrelling over whether what matters is realistic, reasonable attribution by onlookers or a more formalist, medium-based attribution inquiry.81×81. See supra note 55. The test for attribution has been a consistent point of debate in government speech cases. Compare Johanns v. Livestock Mktg. Ass’n, 544 U.S. 550, 560–62 (2005) (holding government control of the message as dispositive in identifying government speech), with id. at 578–79 (Souter, J., dissenting) (insisting upon a reasonable-observer test under which government must identify itself as the speaker). But that debate obscures the point, as neither medium nor realism is tied closely to the justification proffered for exempting these messages and their regulation from the protections of the Free Speech Clause.

Going forward, the Court should eschew purely attribution-based inquiries when deciding whether the government speech exemption applies, and instead adopt an approach more closely linked to the doctrine’s theoretical rationale. Such an approach would usher in a doctrinal coherence absent in Walker. Because the government speech exemption exists to protect government’s ability to communicate, the root doctrinal question should be whether the relevant speech would impair that ability if it were not censored. Whether or not this question would have led to a different result in Walker, it would have moored the Court to the essence of the doctrine it applied.