Government Speech Doctrine: The Rule in One Paragraph
The government speech doctrine holds that a message delivered by the government as its own speech is not subject to the First Amendment limits that restrict government regulation of private speech, so the government may choose the content and viewpoint of that message without granting equal access to opposing views. The doctrine turns entirely on authorship. A message counts as the government’s own speech only where the state has taken the message up as its own through the history of the medium, public attribution of the message to the state, and control the state actually exercises over content; regulation, funding or licensing of a private speaker never by itself transfers authorship of the message to the state. That sentence carries the whole field. Where the three markers converge, the message belongs to the state, and a disappointed private claimant has no free speech right to insert a competing message into the same channel. Where the markers diverge, where a medium has a tradition of private expression, where the public attributes the words to a private speaker, or where the state merely pays for, permits or regulates expression produced by others, the message remains private speech and ordinary First Amendment scrutiny applies with full force.

What is government speech in simple terms?
Government speech is a message the state itself delivers as author, such as a warning on a currency note or a monument inscription chosen by a city. Because the state is the speaker, the First Amendment rule against viewpoint discrimination among private speakers does not apply to that message.
The simplicity of that answer conceals a demanding classification exercise, because almost every contested case involves a channel that the state created, funds, owns or administers in some respect. A city owns the park in Pleasant Grove City v. Summum in 2009, a university collects the fee in Board of Regents of the University of Wisconsin System v. Southworth in 2000, and the Department of Agriculture supervises the advertising program in Johanns v. Livestock Marketing Association in 2005, yet ownership and payment pointed in different directions across those decisions. The distinguishing work therefore falls on evidence rather than on labels. Courts ask how the medium has been used across time, whether a reasonable observer familiar with the setting would understand the state to be speaking, and whether officials selected, edited, approved or rejected the specific content at issue. A state that stamps a slogan on a license plate, inscribes a name on a public building or issues a proclamation through the office of a governor satisfies all three inquiries with little difficulty. A state that opens a meeting room, a student fund or a speciality plate program to diverse private designs satisfies few of them, even though public money and public property stand behind the channel.
Two consequences follow from a finding of government speech, and each explains why litigants contest the classification so fiercely. The first consequence is freedom of selection. The state may promote vaccination and omit scepticism about vaccination, may honour one historical figure on a monument and decline another, and may design a tourism campaign around a single image of a state without presenting rival images, because a speaker choosing a message necessarily excludes other messages. The second consequence is political rather than judicial accountability. The remedy for objectionable government speech lies in elections, appropriations, open records laws and public criticism, not in a court order compelling the state to carry speech it has rejected. The Free Speech Clause of the First Amendment, ratified in 1791 as part of the Bill of Rights, was written as a restraint on government interference with the speech of the people. It was not written as a charter requiring the government to remain neutral in its own pronouncements, and the doctrine preserves that original allocation of roles.
Why Governments Must Speak at All
A state that cannot communicate cannot govern, because nearly every function assigned to public authority depends on the transmission of an authoritative message to a population that must be able to recognize the message as official. Tax administration illustrates the point with unusual clarity. The Internal Revenue Service publishes forms, instructions and notices that state filing obligations in a single official voice, and the system of voluntary compliance built after the adoption of the Sixteenth Amendment in 1913 presupposes that a taxpayer can distinguish an instruction issued by the Treasury Department from commentary produced by a private preparer. If every form had to carry competing interpretations with equal prominence, the instruction would cease to function as an instruction. The same logic governs currency. Coins and notes issued under the authority of the United States Treasury bear inscriptions, portraits and denominations selected by the state, and the credibility of the medium depends on uniformity. A dollar bill that accommodated private messages alongside the denomination would fail at the single task currency exists to perform, which is to be accepted without inquiry into the views of the person presenting it.
Public health supplies a second family of examples in which selective official speech is not a convenience but a condition of effectiveness. The Centers for Disease Control and Prevention issues guidance on vaccination schedules, on the handling of infectious disease and on the risks attached to tobacco, alcohol and contaminated food, and each guidance document takes a position. A campaign urging immunisation against measles states that immunisation is safe and desirable; a warning printed on a cigarette package under federal law states that smoking causes disease. Neither communication could survive a rule requiring balanced presentation of the contrary position, because the purpose of the communication is persuasion in a chosen direction. Surgeon General Luther Terry issued the landmark report on smoking and health in 1964, and the authority of that report derived precisely from the fact that a named public office had weighed evidence and announced a conclusion. Democratic government requires officials to reach conclusions in public, and a doctrine that penalised conclusion reaching would leave agencies able to fund research but unable to tell the public what the research had established.
Recruitment, warning and record keeping extend the argument into settings where the audience is narrower and the need for a single voice is equally pressing. The United States Army recruits soldiers through advertising that presents military service in favorable terms, and no principle of free expression requires the Army to purchase equal time for messages discouraging enlistment. The National Weather Service issues hurricane and tornado warnings for defined geographic areas, and a warning derives its protective value from brevity, urgency and unmistakable official origin. Official records perform a quieter version of the same function. A deed recorded in a county register, a judgment entered by a court clerk, a statute printed in the United States Code and a proclamation issued after a census each speak with the voice of the institution that maintains the record, and the legal effect of the document depends on that attribution. Even ceremonial speech fits the pattern. A governor delivering an inaugural address, a mayor dedicating a memorial in a municipal park and a president delivering the State of the Union address under Article II of the Constitution all speak for an office rather than as private citizens, and listeners understand the words in exactly that way. The doctrine therefore begins from a premise about institutional necessity rather than from a desire to expand state power. Government speech exists because governing requires speech, and the harder questions concern the boundary between the speech a state must be allowed to control and the speech of private persons that the state must leave alone.
The Boundary Problem: Speaker Identity Before Speech Rights
The First Amendment provides that Congress shall make no law abridging the freedom of speech, and the Supreme Court has applied that prohibition to the states through the Fourteenth Amendment since Gitlow v. New York in 1925. The prohibition is directed at a particular relationship. It restrains the government acting as regulator of the speech of others. It does not restrain the government acting as a participant in public discussion, because a participant choosing a message is not abridging the speech of anyone else merely by declining to adopt a different message. That distinction sounds elementary, and its consequences are structural. Every free speech dispute involving a public channel, a public fund or a public program must begin with a question of identity that precedes any question of rights. The court must first decide who is speaking. Only after that classification can the court know which body of doctrine governs, because the rules for government regulation of private speech and the rules for government selection of its own speech are not variations of one test. They are opposite answers to the same demand for neutrality.
Viewpoint neutrality illustrates the opposition most sharply. When the government regulates private speech in a public forum, discrimination among viewpoints is the gravest defect a rule can carry, and the Court treated it as fatal in Rosenberger v. University of Virginia in 1995 when a university fund refused payment for a student publication because the publication addressed its subject from a religious perspective. Neutrality is the price the state pays for administering a channel that belongs, in the relevant sense, to private speakers. When the government speaks for itself, neutrality would destroy the activity. A state historical commission that erects a marker describing a battle, a university president who issues a statement of institutional policy and a public health agency that recommends a vaccine all discriminate among viewpoints as an unavoidable feature of stating a position. Applying forum doctrine to those acts would convert every official statement into a constitutional violation, and applying government speech doctrine to a genuine forum would convert every act of censorship into an editorial choice. Classification therefore precedes scrutiny in a strict sense. A court that reaches the neutrality question before settling authorship has already assumed the answer to the only question that matters.
The boundary problem is difficult because modern government rarely speaks through a single unmistakable channel and rarely confines regulation to channels that are plainly private. Programs mix the elements. A state funds an activity and attaches conditions, as in Rust v. Sullivan in 1991. A state compels contributions toward a message, as in Johanns v. Livestock Marketing Association in 2005. A city accepts some donated monuments and refuses others, as in Pleasant Grove City v. Summum in 2009. A state issues license plates bearing designs proposed by organizations, a practice examined in Walker v. Texas Division, Sons of Confederate Veterans in 2015. In each setting the physical facts are similar enough to mislead. Public money moves, public property is used, public officials make decisions and private persons supply words, designs or funds. The analytical task is to resist deciding the case from any one of those facts in isolation. History of the medium disciplines the inquiry by asking whether the channel has traditionally carried official messages or private ones. Attribution disciplines it by asking how the message is reasonably understood by the audience that receives it. Control disciplines it by asking whether officials exercised editorial authority over the content in fact, through selection, revision or a retained power of final approval, rather than through the background authority that any funder or licensor possesses. Misclassification carries costs in both directions. Treating private speech as government speech licenses suppression without scrutiny, because the label ends the case. Treating government speech as private speech leaves officials unable to deliver a coherent program, because every funded message generates a claim to equal funding for its opposite and every monument generates a claim to equal space.
The Older Roots in Subsidy and Funding Decisions
The doctrine acquired its foundations in disputes about money rather than in disputes about monuments, because funding forced the Court to explain why a government that may refuse to pay for an activity altogether may also pay for one version of an activity and decline another. Rust v. Sullivan in 1991 presented that question in a sharply defined statutory setting. Title X of the Public Health Service Act provided federal funds for family planning services, and regulations issued by the Secretary of Health and Human Services prohibited a funded project from engaging in counselling concerning abortion as a method of family planning or in referral for abortion. Dr. Irving Rust and other recipients and physicians challenged the conditions as discrimination against a viewpoint. The Court, in an opinion by Chief Justice William Rehnquist, upheld the regulations by a vote of five to four. The reasoning rested on a distinction between funding an activity and suppressing speech. The government had chosen to fund a program dedicated to a defined set of services, and the conditions defined the scope of the program itself. A doctor who wished to counsel about abortion remained free to do so outside the funded project, in a setting that was not supported by Title X money and not presented as part of the Title X service.
That reasoning established two propositions that later cases refined rather than abandoned. The first proposition is that the government may select the activities it funds without funding their alternatives, because a subsidy is not a penalty and a refusal to subsidise is not a prohibition. The lineage reaches back to Maher v. Roe in 1977 and Harris v. McRae in 1980, where the Court held that a decision to fund childbirth and not abortion created no constitutional violation merely because the choice made one course easier to pursue than the other. The second proposition concerns the character of speech inside a funded program. When the government disburses funds to convey a message of its own choosing, it may insist that the funds be spent on that message. A program designed to encourage a particular outcome cannot function if the persons delivering it are constitutionally entitled to use the same money to deliver the opposite message. Chief Justice Rehnquist described the Title X project as a setting in which the government was the speaker in substance, employing private professionals to transmit a position the government had adopted for the program. On that understanding the counsellor inside the project spoke within a role the government had defined, in much the same way that an employee delivering an official campaign speaks within the duties of the office.
The opinion in Rust v. Sullivan also contained the seed of its own limit, and later decisions drew the line with greater precision. The government may define the content of a program it funds, but the power to define a program is not a general power to use funding as leverage over speech that occurs outside the program and on the speaker’s own resources and time. Legal Services Corporation v. Velazquez in 2001 tested that boundary. Congress had created the Legal Services Corporation to fund legal assistance for persons unable to afford counsel, and a restriction barred funded attorneys from challenging the validity of welfare laws, even when such a challenge was the appropriate legal argument in a client’s case. Justice Anthony Kennedy wrote for a majority of five that the restriction was unconstitutional. A legal services attorney speaks for a client within an adversarial system that depends on independent professional judgment, and the restriction distorted that system by prohibiting an argument that the program itself existed to present. The funded lawyer was not transmitting a government message, because the program had been designed to facilitate private expression through professional representation. Agency for International Development v. Alliance for Open Society International in 2013 carried the limit into a second setting. Federal legislation funding programs against HIV and AIDS required recipient organizations to adopt a policy explicitly opposing prostitution and sex trafficking. Chief Justice John Roberts wrote for the majority that the condition crossed the line. A condition that defines the scope of a funded program is permissible, because the government is entitled to decide what it is buying. A condition that reaches beyond the program and compels the recipient to affirm a belief as the price of participation restricts speech outside the federal program itself. Read together, the three decisions produce a funding rule with two halves. Inside the program, the government may specify content and insist on fidelity to the funded purpose. Outside the program, the recipient retains the full protection of the First Amendment.
Funding Without Ownership: Rosenberger and Southworth
The student fee cases of the nineteen nineties supplied the necessary counterweight to the funding line, because each involved public money collected and distributed by a state university and each held that distribution of money did not make the resulting expression the speech of the state. Rosenberger v. University of Virginia in 1995 concerned a student activities fund financed by a mandatory fee of fourteen dollars per semester and administered to support extracurricular student publications and organizations. Ronald Rosenberger and other students published Wide Awake, a journal that addressed campus subjects from a Christian perspective, and the university refused reimbursement for printing costs on the ground that the publication constituted religious activity under the fund guidelines. The Court held the refusal unconstitutional by a vote of five to four, in an opinion by Justice Anthony Kennedy. The fund, the Court reasoned, was not a vehicle for a university message. The university had created the fund to encourage a diversity of private views among students, had made it available to a wide range of publications with incompatible positions and exercised no editorial control over the content of the journals it supported. A program of that design is a forum for private speech, described in the opinion through the metaphor of a forum occurring in a metaphysical rather than a physical space, and viewpoint discrimination inside such a forum triggers strict scrutiny that the religious exclusion could not survive.
The reasoning in Rosenberger v. University of Virginia drew a line that funding alone could not erase. Money passed through the hands of a public institution, and the institution set eligibility rules, collected the fee and disbursed the payments, yet none of those facts converted student journalism into university speech. Attribution pointed toward the students, because readers understood Wide Awake as the work of its student editors. History pointed toward private expression, because student publications at the University of Virginia, founded by Thomas Jefferson in 1819, had traditionally served as independent voices rather than as organs of the administration. Control pointed in the same direction, because the university disclaimed editorial authority and intervened only through the exclusion under challenge. The case therefore stands for a proposition that the doctrine requires. Payment is evidence to be weighed, not a conclusion. A government that finances a plurality of speakers on neutral terms has built a forum, and the First Amendment governs the forum under the rules for private speech.
Board of Regents of the University of Wisconsin System v. Southworth in 2000 addressed the mirror image of the same arrangement. Students at the University of Wisconsin at Madison, led by Scott Southworth, objected that a mandatory student activity fee compelled them to subsidise private organizations whose political and ideological positions they opposed. The Court upheld the fee program, again in an opinion by Justice Anthony Kennedy. Compelled funding of private speech by a university does not violate the rights of objecting students where the program is administered under the requirement of viewpoint neutrality, because the university purpose is to facilitate the free and open exchange of ideas among students rather than to deliver a message of its own. The neutrality requirement did constitutional work in both directions. It protected funded groups against the exclusion imposed in Rosenberger v. University of Virginia, and it protected objecting students in Board of Regents of the University of Wisconsin System v. Southworth by ensuring that their compelled contributions supported a forum rather than a position. Together the two decisions clarify why the identity inquiry cannot be reduced to a search for public money. In a forum program the state acts as an administrator under a duty of even handedness, and its lack of a message is the defining feature of the arrangement. Officials may set subject categories, require recognized student organization status and allocate limited funds under neutral criteria, but selection among viewpoints is forbidden precisely because the speech belongs to the students. The contrast with a commissioned program is instructive. When the Department of Health and Human Services defined Title X services in Rust v. Sullivan, the government specified the content of the service it purchased. When the University of Virginia administered its student activities fund, the university specified no content at all and was forbidden to begin specifying content by excluding a religious perspective. The difference lies in the design and representation of the program, in whether the institution holds the channel out as its own voice or as a space for the voices of others.
Johanns and the Compelled Subsidy for an Official Message
Johanns v. Livestock Marketing Association in 2005 forced the doctrine to answer the objection that arises when private persons are not merely taxed in general but are compelled to pay for an identified promotional message with which some of them disagree. The Beef Promotion and Research Act of 1985 authorised the Secretary of Agriculture to establish a program of generic beef promotion financed by an assessment of one dollar per head of cattle sold, commonly called the beef checkoff. Ranchers and importers were required to pay the assessment, and the funds supported advertising campaigns built around the slogan Beef. It is What is for Dinner. The Livestock Marketing Association and several producers objected that the compulsion required them to finance speech promoting an undifferentiated commodity, including beef produced by competitors, contrary to their preference for promoting their own identifiable products. The Court held by a vote of six to three that the promotional program was government speech and that the compelled subsidy therefore raised no First Amendment claim. Justice Antonin Scalia wrote the majority opinion, and Mike Johanns, Secretary of Agriculture, stood as the named petitioner in his official capacity.
Control supplied the decisive evidence in Johanns v. Livestock Marketing Association, and the opinion examined control as a matter of institutional fact rather than as a formal assertion. The Secretary of Agriculture appointed the members of the Cattlemen’s Beef Promotion and Research Board, held authority over the content of promotional campaigns, approved the messages before dissemination and supervised the operating committee that administered the program. Federal law and regulation specified the program objectives, and the Department of Agriculture retained final authority over the speech produced under it. The advertising was therefore not the expression of an independent trade association that the government happened to finance. The association operated inside a structure in which a cabinet officer bore legal responsibility for the message and could direct, revise or block it. Attribution reinforced the conclusion in a quieter way. The campaigns were presented as generic industry promotion rather than as the views of any individual rancher, so no payer was understood by the audience to be endorsing a position personally. The compelled contribution funded a message attributed to the program and to the government structure behind the program, not to the members of the Livestock Marketing Association as speakers.
The holding depends on a distinction between compelled speech and compelled subsidy that the Court had developed in cases involving union fees and bar dues, including Abood v. Detroit Board of Education in 1977 and Keller v. State Bar of California in 1990. A person ordered to utter a message, to display a slogan on private property or to affirm a belief suffers an injury to expression that the First Amendment addresses directly, as West Virginia State Board of Education v. Barnette established in 1943 when the Court protected students who refused to salute the flag. A person required to contribute money to a government program stands in a different position, because taxation routinely finances government speech that individual taxpayers oppose. Citizens pay for military recruitment advertising, for public health campaigns and for presidential communications without acquiring a right to withhold the portion of a tax attributable to each message. Johanns v. Livestock Marketing Association placed the checkoff on the taxation side of that line once the message was classified as the government’s own. The limits of the decision are as significant as the holding. The result would differ if the government compelled a private person to carry a message in a manner that attributed the message to that person, or if the funded program were in substance the speech of a private entity operating outside effective government control. Justice David Souter, writing in dissent, pressed the attribution concern from the opposite direction, arguing that the advertisements gave the audience no indication that the government was the speaker. The disagreement exposes the analytical center of the case. Both the majority and the dissent treated authorship as the question on which everything else depended; the dispute concerned whether appointment power, approval authority and statutory supervision amounted to the kind of control that makes a promotional campaign the speech of the state.
Pleasant Grove City v. Summum and the Doctrine Named in Full
Pleasant Grove City v. Summum in 2009 gave the doctrine its fullest statement and its most durable factual setting, because a public park concentrates every element of the inquiry in a small and visible space. Pleasant Grove City in Utah maintained Pioneer Park, a public park of 2.5 acres containing 15 permanent displays, most of them donated, including a Ten Commandments monument presented by the Fraternal Order of Eagles in 1971. Summum, a religious organization founded in 1975 and based in Salt Lake City, sought permission to erect a permanent monument setting out the Seven Aphorisms of Summum near the Ten Commandments monument, and offered to pay all costs of installation. The city refused. Summum sued, arguing that Pioneer Park was a public forum in which the city had accepted one religious monument and could not discriminate against another religious viewpoint. The Supreme Court ruled unanimously for the city. Justice Samuel Alito wrote the opinion, holding that the selection and display of permanent monuments in a public park constitutes government speech and that the Free Speech Clause therefore does not require a city to accept a monument it has chosen not to display.
Justice Alito grounded the holding in the history of the medium, and the historical account did analytical work rather than decorative work. Governments across centuries have used monuments to communicate official messages about identity, memory and value. A city that erects a statue of a founder, a state that builds a memorial to soldiers who died in a named war and a nation that places an obelisk on the National Mall in Washington all employ durable physical objects to state what the community honours and how the community understands its past. The Washington Monument, completed in 1884, the Lincoln Memorial, dedicated in 1922, and countless courthouse statues of Justice belong to a tradition in which the monument is understood as the voice of the polity that authorised, sited and maintains it. That tradition distinguishes a permanent monument from the leaflets, speeches and demonstrations that also occur in parks and that forum doctrine protects as private speech. A park is a traditional public forum for transient expression by members of the public. A permanent installation occupying public land for decades is a different medium with a different history, and the Court refused to let the forum status of the surrounding ground determine the classification of the objects placed permanently within it.
Public attribution supplied the second ground. Justice Alito reasoned that observers reasonably associate a permanent monument displayed on public land with the government that owns the land and maintains the display, because governments do not ordinarily permit installations of that permanence without adopting the message the installation conveys. The city had taken ownership of the donated monuments, including the Ten Commandments monument from the Fraternal Order of Eagles, and bore responsibility for their upkeep, repair and continued presence. Donation therefore did not preserve private authorship. The act of acceptance, followed by ownership and maintenance across decades, transferred the message into the voice of the city in the eyes of the public that encountered it. Control supplied the third ground, expressed in Pleasant Grove City v. Summum as selective acceptance. The city had exercised discretion over every proposed permanent display, had enacted criteria for monuments in Pioneer Park and had rejected at least one earlier proposal. Space itself imposed a practical limit that reinforced the constitutional point. A park of 2.5 acres cannot accommodate a permanent monument from every group that seeks to place one, and a rule requiring acceptance of each qualifying donation would produce either an unintelligible clutter of competing permanent statements or a decision by the city to accept no monuments at all. Justice Alito treated that dilemma as evidence that selection is inherent in the medium. A government maintaining a park must choose, and the choice among monuments is an exercise in stating an official account of the community, not an act of censorship against the groups whose proposals are declined. Summum remained free to distribute literature in the park, to hold gatherings there and to erect its monument on private land, and those preserved channels confirmed that the city had regulated no private speech at all. The opinion closed by locating accountability outside the Free Speech Clause, in the political process and in other constitutional provisions that continue to apply to government speech. A city that adopts a monument remains bound by the Establishment Clause of the First Amendment, by the Equal Protection Clause of the Fourteenth Amendment and by the judgment of voters. Government speech is therefore not speech outside the Constitution. Government speech is speech outside the specific prohibition on abridging the freedom of speech of private persons, and that prohibition was the only basis of the Summum claim.
Walker and the Specialty License Plate
The specialty plate dispute reached the Supreme Court after Texas refused a design proposed by the Texas Division of the Sons of Confederate Veterans. The design featured a Confederate battle flag, and the Texas Department of Motor Vehicles Board rejected the design after public comment showed strong objection to the symbol. The organization argued that the rejection sorted private designs by viewpoint and therefore violated the First Amendment. In Walker v. Texas Division, Sons of Confederate Veterans in 2015, Justice Breyer wrote for a majority of five and held that specialty license plates are government speech, so Texas could refuse the design without satisfying the rules that govern regulation of private expression in a forum.
The reasoning began with the function of a license plate itself. A plate is issued by the state, carries the name of the state, serves as government identification for a vehicle, and remains tied to official registration records. Drivers display plates because the state requires display, and observers associate plates with the state that issued them. Specialty designs modify the plate within a system the state created and administers, and final approval rests with a state board. A specialty design therefore does not merely appear on state property through a licensing arrangement.
History reinforced the reading of plates as official media. States have long used license plates to convey state messages, from names and slogans to commemorative themes chosen by legislatures. The specialty program expanded the range of designs, yet the expansion occurred inside a medium that had carried state identification and state promotion for generations. Private groups proposed designs, motorists selected among approved designs, and the state retained authority to approve or reject each proposal before manufacture. The majority treated proposal and selection as participation inside a state program rather than as creation of an open channel for private speakers.
Public perception formed the second strand of the analysis. A reasonable observer encountering a specialty plate sees the state name across the top, a registration number assigned by the state, and a design approved under state authority. The observer has little reason to treat each design element as the personal statement of the driver alone, even though the driver chose the design from an approved list. Choice among government approved options differs from authorship of the options themselves.
Control formed the third strand, and the control in Texas was direct rather than nominal. State law required board approval for every specialty design, the board reviewed proposals on the merits, and the board exercised that authority in the disputed decision itself. No design reached the road without affirmative state approval. The majority contrasted that arrangement with a forum in which the state records or distributes private messages without selecting content. Regulation alone would not have carried the same weight.
The holding carried limits that later cases developed. Government speech remains subject to other constitutional constraints, including the Establishment Clause and the guarantee of equal protection, and a classification in one program settles nothing about a different program with a different structure. The three considerations in Walker v. Texas Division, Sons of Confederate Veterans later hardened into factors because courts needed a method for separating adoption from permission. History of the medium, public attribution, and control actually exercised supply that method.
Matal v. Tam and the Viewpoint Limit
Two years after the plate decision, the Court refused to extend the government speech label to federal trademark registration. Simon Tam, lead singer of the dance rock band The Slants, sought registration of the band name as a trademark. The Patent and Trademark Office refused registration under the disparagement clause of the Lanham Act, which barred registration of marks that disparage persons, institutions, beliefs, or national symbols. Tam chose the name to reclaim a slur directed at Asian Americans, and the refusal rested on a finding that the term disparaged a substantial composite of Asian Americans. In Matal v. Tam in 2017, Justice Alito wrote the principal opinion, and all participating justices agreed that the disparagement clause violated the First Amendment.
Can the government refuse a trademark because it offends a group?
A refusal based on offense to a group is viewpoint discrimination when the mark itself remains private speech. Registration records a private mark, and the government may enforce neutral rules on source confusion or deception, but offense alone does not supply a lawful ground for denial.
The threshold question concerned authorship, and the answer was that a trademark belongs to its owner. A mark identifies the source of goods or services in commerce, and consumers associate a mark with the producer rather than with the federal government. Registration places the mark on the Principal Register, grants procedural and evidentiary advantages, and supports enforcement against infringers, yet registration does not change the origin of the expression. The examining attorney reviews an application for statutory compliance, and approval records that the mark satisfies the registration criteria.
The distinction between subsidy and speech carried additional weight. A government program that spends funds to convey a chosen message may select content because the government itself is the speaker, as in a public health campaign drafted by an agency. Trademark registration operates differently. The applicant pays a fee, creates the mark in private commerce, and seeks legal protection for private use. The benefit conferred by registration does not transform the beneficiary into a government mouthpiece, just as copyright registration does not make a novel an official publication. A comparable insistence that classification precedes regulation runs through the Supreme Court campaign finance cases, where treating spending as speech, rather than as a government administered benefit, decided the reach of the limits under review.
With authorship settled as private, the disparagement clause failed because the discrimination turned on viewpoint. A mark that praises a group could register, while a mark that disparages the same group could not, even where both marks addressed the same subject. The clause therefore favored one side of a debate about the dignity and standing of identifiable groups. Government regulation of private speech on the ground that the ideas expressed offend listeners is the core evil that viewpoint doctrine targets, because offense depends on the position taken rather than on a neutral feature such as fraud, confusion, or false advertising.
Matal v. Tam also clarified the boundary between a register and a forum of official messaging. The decisive point remained attribution joined to control over meaning. The office controlled eligibility for a legal status, and applicants controlled the expressive content of their marks. Where control extends only to the conditions of a register and the message originates in private commerce for private purposes, the register stays private in character. A stamp of approval, standing alone, is evidence of compliance with a regulatory scheme.
Professional Speech and Compelled Disclosure After NIFLA
The California Reproductive Freedom, Accountability, Comprehensive Care, and Transparency Act required licensed clinics that serve pregnant women to post a notice stating that California provides free or low cost access to comprehensive family planning services and to prenatal care and abortion. Unlicensed facilities faced a separate notice disclosing the absence of a license and of licensed medical personnel. Crisis pregnancy centers challenged both notices as compelled speech. In National Institute of Family and Life Advocates v. Becerra in 2018, Justice Thomas wrote for a majority of five and held that the licensed notice likely violated the First Amendment and that the unlicensed notice failed the applicable disclosure standard.
The majority rejected the idea that speech by licensed professionals forms a separate category receiving diminished protection. Doctors, lawyers, counselors, and other professionals speak in settings shaped by licensing, ethics rules, and duties of competence, and those settings support certain regulations of professional conduct that incidentally affect speech. A state may require informed consent for a medical procedure, discipline a professional for misleading a client, and enforce standards against fraud. Each power regulates conduct within a professional relationship or targets deception inside that relationship. A broader claim that professional speech as such deserves reduced scrutiny found no support in the case law, apart from two established pockets involving commercial disclosure and speech tied directly to the performance of a procedure.
The licensed notice fell outside both pockets. The disclosure rule associated with Zauderer v. Office of Disciplinary Counsel of the Supreme Court of Ohio in 1985 permits a state to require purely factual and uncontroversial information in commercial advertising, provided the requirement is not unjustified or unduly burdensome and is reasonably related to preventing deception of consumers. The California notice did not advertise a service offered by the clinic, did not correct a deceptive claim made by the clinic, and compelled the clinic to publicize services provided by the state, including abortion, a subject on which the centers held opposing moral convictions. Content based compulsion of that kind demanded heightened scrutiny, and the state failed to show that the notice was the least restrictive means of informing women about state programs.
Authorship explains why the notice counted as compelled private speech rather than as the message of the state delivered through a convenient channel. The notice appeared on clinic walls and in clinic waiting rooms, where patients reasonably understood communications as coming from the provider caring for them. California drafted the words, yet placement inside the professional setting transferred the practical burden of speaking to the clinic and borrowed the credibility of the clinic for a message the clinic opposed. A state remains free to publish the same information on government websites, in government offices, and through government advertising purchased with public funds.
The unlicensed notice illustrated a different failure. Disclosure of unlicensed status is factual in isolation, yet California required the notice in large type, in as many as 13 languages depending on the county, and in advertisements of every kind, including materials with limited space. The burden threatened to drown out the facility message itself, and the state offered an inadequate justification for the breadth of the requirement. Licensing creates a regulatory relationship that supports rules tied to the licensed activity, to competence, and to deception.
Brunetti and the Register That Stayed Private
Erik Brunetti founded the clothing brand FUCT in 1990 and sought federal registration of the brand name as a trademark. The Patent and Trademark Office refused registration under the provision of the Lanham Act barring marks that consist of or comprise immoral or scandalous matter. The refusal treated the brand name as a vulgar term, even though the letters functioned as the source identifier Brunetti had used in commerce for decades. In Iancu v. Brunetti in 2019, Justice Kagan wrote for a majority of six and held that the immoral or scandalous bar violated the First Amendment because the bar discriminated on the basis of viewpoint.
The statutory terms drove the outcome. Marks expressing ideas that align with conventional moral approval qualified for registration, while marks expressing ideas that defy conventional approval did not, even where both sets of marks addressed the same subjects of religion, morality, and social conduct. Registration decisions under the bar had allowed marks praising religious belief and refused marks criticizing religious belief, and had sorted political messages by the favor or disfavor the messages expressed. A standard built on approval and disapproval of ideas is not a neutral filter for vulgarity alone.
Authorship received less extended treatment than in Matal v. Tam, because the earlier decision had settled the character of the register. A trademark remains the expression of the owner who created the mark and used the mark in commerce, and entry on the federal register records legal status rather than adopting the mark as an official message. Iancu v. Brunetti confirms that settlement by deciding the case on viewpoint grounds, a framework that applies only where the regulated expression is private. Had registration converted a mark into government speech, viewpoint selection among marks would have raised no First Amendment problem of that kind.
The confirmation matters for every approval system that touches expression. Governments maintain registers of trademarks, copyrights, professional licenses, corporate names, and parade permits, and each register involves review followed by an entry, a certificate, or a stamp. A certificate proves that an application satisfied the criteria for entry. A certificate does not prove that the government composed the entry or will be understood by the public as the speaker behind the entry.
Limits remain inside the register itself. Trademark law may refuse marks that mislead consumers about source or content, may refuse generic terms that competitors need to describe their goods, and may enforce rules against marks that falsely suggest a connection with a person or institution. Each limit ties to the commercial function of a mark rather than to approval of the ideas a mark expresses. A bar framed around immorality and scandal failed because the criterion measured social approval of a message. After Iancu v. Brunetti, the register stayed private, and the reasoning supplies a durable question for other registers, asking whether the state selected the message as a state message or merely recorded a private message under a regulatory standard.
Shurtleff v. City of Boston and the Three Factors Stated
Boston City Hall Plaza contains three flagpoles. The city flies the flag of the United States on one pole and the flag of Massachusetts on another, and the third pole ordinarily carries the flag of Boston. For years the city allowed private groups to request use of the third pole for flag raisings connected to ceremonies on the plaza. Harold Shurtleff and Camp Constitution, a group dedicated to teaching about the Constitution and the role of religion in American history, requested permission to raise a Christian flag, and a city official refused, citing concern that flying a religious flag on city property would violate the Establishment Clause. In Shurtleff v. City of Boston in 2022, Justice Breyer wrote for a unanimous Court on the central holding and concluded that the flag program was not government speech.
What are the three factors in the Shurtleff test?
The test asks about the history of the medium, whether the public attributes the message to the government, and how much control the government exercises over content. No single factor controls, and a program with private tradition, private attribution, and no review points away from government speech.
The history of the Boston program weighed heavily against the city. Flag raisings on the third pole had become a civic ritual performed by private groups, with group members raising their own flag during their own ceremony while city staff played no role in selecting the design or the message. The city described the plaza program in public materials as a forum for the community, invited groups to participate, and approved 284 flags over the history of the program without rejecting a single request before the Camp Constitution application.
Public perception pointed in the same direction. A passerby watching a group gather on the plaza, conduct a ceremony, and raise a flag associated with the group had reason to attribute the flag to the group holding the event rather than to the municipal government. Flags on the first two poles carried official meaning because the city selected and flew those flags as symbols of the nation and the Commonwealth. The third pole operated under a different practice, with rotating flags tied to scheduled private events, and context rather than location supplied the signal an observer would read.
Control decided the balance. Boston exercised no review of flag designs, kept no records of the messages flown in most instances, imposed no content criteria in a written policy, and delegated the ceremony entirely to the requesting group. The absence of control was not a technical gap in paperwork. Control is the mechanism by which a government takes a message up as a government message, editing, approving, or rejecting content before display. A government that performs none of those acts has permitted expression rather than authored expression, however prominent the display site may be.
The holding followed from the classification. Because the program was not government speech, the plaza program functioned as a forum for private expression, and exclusion of the Christian flag because of the religious viewpoint of the flag was discrimination that the First Amendment forbids. The Establishment Clause concern did not justify the exclusion, since accommodation of private religious expression in a forum open to comparable secular expression does not establish a religion. Justice Kavanaugh, Justice Alito, and Justice Gorsuch each wrote separately, debating the broader doctrine, while the majority confined the decision to the record of the Boston program.
The History Factor: Media That Have Carried Official Messages
History in the government speech inquiry means the history of the medium as a channel of communication, read through the practice of the government that claims authorship. Courts ask whether the medium has traditionally carried messages of the state, whether private speakers have traditionally used the same medium for private messages, and whether the specific program under challenge developed along one line or the other. A monument in a public park offers the clearest official example. In Pleasant Grove City v. Summum in 2009, the tradition of government monuments supported the conclusion that donated monuments accepted for permanent display in Pioneer Park were government speech, because cities have used monuments to express public identity since the nineteenth century and because permanent placement forces selection among competing proposals.
License plates present a mixed history that courts resolve by attention to the program rather than to the object in isolation. Plates began as registration markers, acquired state names and slogans, and developed into carriers of state promotion long before specialty designs invited proposals from private groups. The private proposal step arrived late in the life of the medium and operated inside state approval from the start, so the older official tradition retained force. A different sequence produces a different result. Where a display practice begins as an open community ritual and continues for years without official selection, the tradition that accumulates is a tradition of private use on public property. The Boston flagpole had that sequence, and the sequence mattered more than the formal fact that a municipality owned the pole.
Proclamations, seals, and official mottos sit at the strongest end of the historical scale. A gubernatorial proclamation issues in the name of the state, follows a settled ceremonial form, and has no tradition of private authorship behind the form. A state seal authenticates official acts and appears on documents precisely because observers treat the seal as the mark of the state itself. Specialty media with commercial or expressive origins sit nearer the opposite end. Trademarks developed in private commerce to identify source, and public bulletin boards developed as posting space for community notices. A long record of private expression weighs against an official speech claim because the claim asks the court to recharacterize a channel that speakers and observers have treated as private for generations.
Evidence for the history factor takes concrete forms. Statutes creating a program, regulations governing submissions, records of approvals and rejections, and publicity materials describing the program each show how a medium functioned over time. A program with written selection criteria applied across hundreds of decisions tells a different historical story from a program with no criteria and a perfect record of approval. A government may redesign a program for the future through genuine selection and announced standards, yet recharacterization operates prospectively through changed practice. A label applied during litigation receives skepticism because the history factor tests whether claimed authorship matches the life of the medium before the dispute arose.
The Public Perception Factor: Who a Reasonable Observer Takes to Be Speaking
Attribution turns on context rather than on location alone. The question asks whether a reasonable observer, familiar with the setting and the practice surrounding the display, would take the government to be the speaker. Placement supplies one signal among several. A message carved into the base of a courthouse monument, printed on a state issued identification card, or read aloud by an official at a state ceremony carries the marks of official authorship in the setting itself. A message posted on a community board among hundreds of private notices, raised by a private group during a private ceremony, or printed on a product label in a store carries the opposite marks, even where a government owns the board or issued a permit for the ceremony.
Names, seals, and disclaimers refine the signal. A display that bears the name of a private organization, identifies a private sponsor, and omits any state seal points toward private authorship, while a display issued under a state name with a state seal and no private attribution points toward the state. Disclaimers carry weight in proportion to clarity and placement. A prominent statement that a program presents private views, printed in program materials and repeated at the display site, can separate the government from messages the government hosts. A disclaimer buried in application paperwork that no observer at the display will encounter carries little weight, because perception forms at the point of display.
Rotation and variety supply perception evidence of a practical kind. A pole, board, or exhibition space that carries a different private message every week, on subjects ranging from ethnic festivals to commercial promotions to religious observances, teaches observers that the channel hosts private speakers in sequence. A channel that carries a single consistent message, or a curated set of messages aligned with a government program, teaches the opposite lesson. A register containing hundreds of thousands of private marks cannot plausibly read as a collection of government statements, and a park containing a small number of permanent monuments selected over decades reads as a curated official display. Scale and selectivity shape what an observer reasonably infers.
Perception evidence has weaknesses that careful courts acknowledge. Observers differ in knowledge, and an observer familiar with a program may attribute a message differently from a passerby encountering the display for the first time. Litigation surveys can manufacture the perception a party needs, and official assertions about public understanding can serve as a litigation position rather than as a description of practice. The factor also risks circularity, since a government that announces authorship loudly enough may create the very perception the doctrine then credits. For those reasons perception operates as one factor in company with history and control rather than as a standalone test. A weak showing can expose a claim that rests on property ownership or regulatory authority alone, as the Boston flag record showed.
The Control Factor: Selection, Review and Final Approval
Control is the factor that most often decides a government speech claim, and the control that counts is control exercised in practice rather than authority reserved in a policy document. A city, an agency, or a university may write rules claiming a right to approve every message, yet a record showing that officials approved messages without reading them, rejected none, and treated a space as open to all comers points toward private speech. Courts examine who selected the speaker, who reviewed the wording before it appeared, who could require changes, and who held final approval when a dispute arose.
The Boston flagpole program in Shurtleff v. City of Boston illustrates failure on each step. Boston owned three flagpoles outside City Hall and allowed community groups to raise a flag of their choosing on the third pole during events. Between 2005 and 2017 the city approved 284 flags without reviewing their content and without rejecting a single request. Officials did not select the flags, did not propose wording or imagery, and did not treat the raised flag as a statement by Boston itself. When the city refused a flag bearing a Christian cross requested by Harold Shurtleff and Camp Constitution, the Supreme Court held in 2022 that the program, as administered, was not government speech. The absence of review was not a technical defect. It showed that the city had created a channel for private expression and had surrendered the editorial function that authorship requires.
The contrast with Pleasant Grove City v. Summum is sharp because the municipal conduct differed at every point. Pleasant Grove City owned the monuments in Pioneer Park, solicited donations under criteria tied to the history of the community, and accepted only a small number of displays from a larger pool of offers. The city took title to each accepted monument, maintained the surrounding grounds, and retained power to remove a display. A Ten Commandments monument donated by the Fraternal Order of Eagles stood under those conditions, while a proposed Seven Aphorisms monument from Summum was declined. The Supreme Court held in 2009 that permanent monuments in a public park were government speech, a conclusion grounded in selection, ownership, and continuing custody rather than in a bare assertion that parkland belonged to the city.
A middle pattern appears when a state designs a program, sets detailed content rules, and enforces them before money or space is released. In Rust v. Sullivan, federal regulations for a family planning program barred counselling and referral concerning abortion as a method of family planning, and project staff operated inside limits defined in advance by the Department of Health and Human Services. In Johanns v. Livestock Marketing Association, a federal statute and a Secretary of Agriculture approval process governed promotional messages financed through a beef checkoff, with the Secretary appointing board members and approving campaign material. Control in those settings shaped the message before the message reached an audience. A grant agreement that reserves a right of approval but leaves topic choice, drafting, and revision to a recipient shows weak control. The parallel in spending law is the treatment of conditions attached to federal grants, where a condition that defines the scope of a funded project is read differently from a condition that reaches private expression beyond the project. A commissioning process in which officials draft instructions, review scripts, demand edits, and sign off on the final version shows strong control, even if a contractor speaks the words aloud. Review after publication, without power to prevent release, resembles regulation more than authorship.
Platforms, Moderation and the Limits Shown by Moody v. NetChoice
Large social media platforms present the control question in a different setting because the speaker arranging a feed, the users supplying posts, and the state seeking to regulate both occupy distinct positions. In Moody v. NetChoice, decided in 2024, the Supreme Court considered laws enacted in Florida and Texas that restricted the ability of major platforms to remove, demote, or label user content and that imposed disclosure duties concerning moderation decisions. Trade associations including NetChoice challenged the laws on First Amendment grounds, arguing that decisions about which posts to display, and in what order, formed a curated product protected as private expression. The states defended the laws in part by treating platforms as conduits for the speech of others, closer to common carriers than to editors.
The Court vacated judgments from the lower courts and returned both cases for further proceedings, with an opinion by Justice Kagan stressing method over final resolution. A facial challenge, the opinion explained, requires an assessment of which applications a law covers and whether unconstitutional applications substantially outweigh lawful ones. Neither lower court had performed that accounting across the full range of platform functions, which extend from public feeds to private messaging, email hosting, and payment-linked services. The failure was analytical rather than incidental. A rule governing a curated news feed raises different questions from a rule governing a direct message between two users, and a court cannot settle the feed question by pointing to functions that involve little editorial choice.
Within that procedural holding, the reasoning treated a platform feed as an expressive compilation. A service such as Facebook or YouTube selects and arranges third party material into a sequence presented under its own name, applies community standards, and removes material that violates those standards. That activity parallels the editorial discretion protected in cases concerning newspapers and parades, even though the volume is vast and much of the sorting is automated. The feed is not a state message because no state selects its contents, and it is not a mere transmission pipe because the platform exercises judgment about inclusion and prominence. A state law compelling a platform to carry material it would exclude therefore interferes with private editorial choice, subject to the application by application analysis the Court required. The longer line of Supreme Court decisions that shaped internet law supplies the background for that treatment, because those decisions consistently analysed online services as speakers and editors rather than as state administered channels.
Government pressure on platforms raises the adjacent problem examined in Murthy v. Missouri, also decided in 2024. Missouri, Louisiana, and several individual plaintiffs alleged that federal officials had coerced or significantly encouraged platforms to suppress posts concerning COVID-19 and election administration, turning private moderation into state action. The Supreme Court held that the plaintiffs lacked standing, finding the record insufficient to trace particular moderation decisions concerning their posts to particular federal communications, especially where platforms enforced their own policies independently. Justice Barrett wrote for the majority, and the merits of the coercion question remained unresolved. The standing result left a boundary in place without defining it fully. Persuasion, public criticism, and information sharing by officials differ from threats tied to adverse government action, and the line drawn in National Rifle Association of America v. Vullo, decided the same term, treats coercive threats aimed at punishing disfavored speech as unconstitutional. Between open advocacy and coercion lies a contested zone in which emails, meetings, and public statements must be read together, with attention to whether a platform changed a decision it would otherwise have kept.
Compelled Speech: The Doctrine on the Other Side of the Line
Government speech doctrine answers who is speaking. Compelled speech doctrine answers a different question, whether the state may force a private person to become the speaker. When the state speaks for itself, it may choose a message and refuse another. When the state conscripts a private voice, the First Amendment ordinarily forbids the conscription, even if the message is one the state could lawfully deliver through its own employees, signs, or publications.
West Virginia State Board of Education v. Barnette, decided in 1943, supplies the foundation. West Virginia required public school pupils to salute the flag and recite the Pledge of Allegiance, with refusal treated as insubordination and parents exposed to prosecution. Jehovah’s Witnesses objected on religious grounds, viewing the salute as an act of idolatry forbidden by their faith. The Supreme Court held the requirement unconstitutional in an opinion by Justice Jackson, reversing a decision from three years earlier that had upheld a similar rule. The state could teach patriotism, present the flag, and lead voluntary observances. It could not prescribe an orthodoxy and compel each child to declare acceptance through word and gesture. The defect lay not in the content of the pledge alone but in the compulsion applied to a dissenting speaker in a public classroom.
What is the difference between government speech and compelled speech?
Government speech occurs when the state speaks for itself and may choose its message without viewpoint neutrality. Compelled speech occurs when the state forces a private person to carry or utter a message the person rejects. The first doctrine grants the state freedom as speaker; the second protects the private speaker against conscription.
Wooley v. Maynard, decided in 1977, extended the principle from a classroom ceremony to an object carried in public every day. New Hampshire required the motto Live Free or Die to appear on vehicle license plates and prosecuted George Maynard, a Jehovah’s Witness, for covering the motto with tape. The Supreme Court held that the state could not force an individual to serve as a mobile billboard for an ideological message he found morally objectionable. New Hampshire retained ample means of communicating its own views through official vehicles, public buildings, and state publications. What it lacked was authority to place its motto on the conscience of a private motorist and punish concealment. The case also clarifies a boundary with Walker v. Texas Division, Sons of Confederate Veterans, where specialty plate designs were treated as state speech because Texas designed, controlled, and issued the plates as its own program. In Wooley v. Maynard, the compelled element was the personal display obligation imposed on the plate holder.
The modern commercial setting appeared in 303 Creative LLC v. Elenis, decided in 2023. Colorado prohibited businesses offering services to the public from refusing service based on sexual orientation, and the dispute concerned Lorie Smith, a website designer who sought to create wedding websites expressing a view of marriage she endorsed while declining sites celebrating marriages she opposed. The Supreme Court held that Colorado could not compel Smith to create expressive designs speaking a message inconsistent with her beliefs. The majority treated custom wedding websites as expressive activity rather than as a neutral good sold off a shelf, and it distinguished regulation of conduct, such as rules governing access to ordinary goods and services, from regulation that dictates the content of expression itself. Critics warned that a broad reading could weaken public accommodations law. Supporters answered that the holding turns on expression tailored to a particular occasion and message.
The Khatibi Problem: Licensed Teaching and the Regulation Argument
Licensed professions test the boundary because the state sets entry standards, defines scopes of practice, disciplines licensees, and often prescribes training that private instructors then deliver. The question is whether prescription alone makes the resulting instruction an official message, or whether the instructor remains a private speaker delivering material under compulsion. Khatibi v. Yip placed that question before the Supreme Court through a petition that the Court declined to hear, leaving a lower court answer in force without endorsing its reasoning.
California required continuing medical education courses for license renewal to include instruction on implicit bias under a statute enacted in 2019 and effective January 1, 2022. Accredited providers offered the courses, physicians selected among approved offerings, and instructors presented the material in settings ranging from hospital lecture halls to recorded online modules. The statutory requirement specified that qualifying instruction address the effects of implicit bias on patient care, including disparities in treatment and strategies for recognising and mitigating bias. The state did not draft a single script, appoint the lecturers, or deliver the lectures through employees of the Medical Board of California.
Dr. Azadeh Khatibi, an ophthalmologist, together with the organization Do No Harm, challenged the requirement as compelled speech. The claim rested on a straightforward premise. Physicians had to complete qualifying education to retain a license, qualifying education had to contain instruction the plaintiffs disputed, and private instructors had to teach that instruction to audiences required to receive it. On that account, the mandate used the machinery of licensure to place a state favored message inside professional training, regardless of whether the state had written every slide. The challengers treated the case as an application of Barnette and Wooley v. Maynard to professional education, with the license serving as the lever of compulsion.
The United States Court of Appeals for the Ninth Circuit held in 2025 that the courses were government speech because of the breadth of state regulation over medical practice and medical education in California. The reasoning emphasised the pervasiveness of licensing, accreditation, and statutory direction, and it treated that regulatory density as sufficient to convert the mandated instruction into a message of the state. Under government speech classification, the compelled speech claim failed at the threshold, since a requirement that a speaker deliver the government’s own message in a government program does not trigger the ordinary prohibition on viewpoint based compulsion. The breadth rationale is the element that drew national attention, because it locates authorship in the scale of regulation surrounding an activity rather than in selection of a speaker, attribution of the message to the state, or review and approval of course content by officials.
On November 16, 2026, the Supreme Court denied certiorari in Khatibi v. Yip, No. 25-1240. Justice Gorsuch noted that he would have granted review. Justice Thomas, joined by Justice Alito, dissented from the denial and argued that heavy regulation does not convert private instruction into an official message. The dissent warned that a contrary rule would allow a state to evade compelled speech limits across licensed occupations by the simple device of regulating an occupation heavily and then prescribing the views its members must learn or repeat. A denial of certiorari decides nothing on the merits; the Court states that position as a standing rule of its practice, and the denial in No. 25-1240 carries no approval of the Ninth Circuit reasoning and no rejection of the challengers’ argument. The Ninth Circuit judgment remains binding within that circuit and persuasive nowhere by command, while courts elsewhere remain free to reach a different conclusion on comparable facts. The practical consequence is a divided landscape in which the same structural arrangement, a license conditioned on state prescribed instruction delivered by private teachers, may be classified differently depending on the circuit in which a challenge is filed. Whether a later decision displaces the reasoning altogether is a separate question, governed by the rules the Supreme Court applies when it overrules its own precedents.
The Doctrine Across Its Leading Cases
The sequence of leading decisions settles, case by case, which settings the Supreme Court has treated as government speech and which it has refused, with the limit each decision stated in its own reasoning. Monuments and specialty plates crossed the line on records of ownership, history, and control, while clinic notices and trademark registration stayed on the private side because the state was regulating or recording the speech of others.
| Case | Year | Setting | Speaker identity holding | Limit stated |
|---|---|---|---|---|
| Rust v. Sullivan | 1991 | Federally funded family planning program | Government speech | The government may define program content but may not suppress speech outside the program |
| Johanns v. Livestock Marketing Association | 2005 | Beef promotion checkoff | Government speech | Compelled funding is permissible where the message is controlled by the government |
| Pleasant Grove City v. Summum | 2009 | Park monuments | Government speech | Selection and ownership of permanent displays place authorship with the city |
| Walker v. Texas Division, Sons of Confederate Veterans | 2015 | Specialty license plates | Government speech | Plate designs closely identified with the state remain subject to state choice |
| Matal v. Tam | 2017 | Trademark register | Private speech | Registration records a private mark and does not convert it into a state message |
| National Institute of Family and Life Advocates v. Becerra | 2018 | Clinic notices | Private speech | The state may not compel licensed clinics to deliver a notice favoring state services without justification |
| Shurtleff v. City of Boston | 2022 | Municipal flagpole program | Private speech | An open program administered without content review is a channel for private speakers |
| Khatibi v. Yip | 2026 | Continuing medical education, certiorari denied | Government speech in the Ninth Circuit | Certiorari denial decides nothing on the merits and binds no other court |
Read across the rows, acceptance clusters where the state designed a program and controlled its output, and refusal clusters where the state administered a system that private speakers used for their own messages. The table also shows the doctrine expanding through new settings without abandoning its central demand, a record tying the message to the state as author rather than as regulator, funder, or registrar.
The Complication: Regulation Dressed as Authorship
The strongest case against the thesis begins with substance rather than form. A state that prescribes curriculum content for a license, specifies the conclusions training must convey, and conditions the right to practise on completing that training has done more than regulate the manner of private teaching. It has chosen a message, required its delivery, and selected the audience by law. On that view, insisting that the instructor remains the author elevates paperwork over reality. The physician required to sit through implicit bias instruction, and the instructor required to present it, experience the message as an official position backed by the licensing power of California, whether or not a state employee wrote the slides. If authorship follows effective control over what must be said, a detailed curriculum mandate is control in its most consequential form, control over content itself rather than over scheduling, fees, or record keeping.
That argument draws strength from the reasoning adopted by the Ninth Circuit in Khatibi v. Yip, where the breadth of regulation over medicine supported classification of mandated course content as government speech. It also draws strength from a broader worry expressed by Justice Alito in a concurrence in Shurtleff v. City of Boston, joined by Justice Thomas and Justice Gorsuch, warning that the government speech inquiry can be manipulated and that courts must look past labels a government attaches after a dispute arises. A government sophisticated enough to draft a program description, insert approval language into a handbook, and describe every funded activity as its own message could, under a loose test, convert large fields of private expression into official speech and remove viewpoint neutrality and compelled speech protections in a single step. The worry cuts in both directions. A test too attentive to formal indicia invites manipulation by the state, while a test based on regulatory density invites the same result by a different route, since modern licensing regimes are dense by design across medicine, law, accounting, engineering, and teaching.
The answer preserves the thesis by returning to what authorship means in every other setting where the doctrine applies. A speaker is the party held out to an audience as the source of a message and the party exercising editorial judgment over its content. Regulation, however extensive, ordinarily leaves both features with the regulated party. A building code prescribes much about a house without making the state the builder, and a medical licensing regime prescribes much about practice without making the state the author of every lecture delivered under its rules. Where providers choose instructors, instructors choose examples and emphasis within a topic mandate, and officials neither review nor approve the material presented, the state has commanded that a subject be taught. Commanding a subject is an exercise of regulatory power subject to constitutional limits, and audiences in professional courses ordinarily attribute views to the instructor and the accrediting provider rather than to the Medical Board of California.
One narrowing is required, and the thesis accommodates it without collapse. Speech the state commissions and approves word by word is its own even when a contractor delivers it. A training film scripted by agency staff, reviewed through successive drafts, approved by a named official, and presented as an official production satisfies history, attribution, and control together, and the use of an outside narrator or production company changes nothing. The same conclusion follows for a curriculum in which officials write the text, require verbatim delivery, and forbid deviation. In those settings the private party functions as an instrument of transmission, comparable to a printer or a broadcaster under contract. The line therefore runs between prescribing that instruction occur and prescribing the instruction itself in finished form, with evidence of drafting, review, and approval deciding close cases.
The Forum Doctrines That Sit Beside Government Speech
Forum doctrine and government speech doctrine govern the same physical and institutional spaces through opposite starting questions. Government speech analysis asks whether the state is the speaker. Forum analysis assumes private speakers are present and asks what kind of place or program the state has opened to them. Classification decides the standard that follows. A finding of government speech ends the Free Speech Clause inquiry into viewpoint choice, because a speaker may choose a position. A finding that the state has opened a forum triggers viewpoint neutrality within that forum, with additional rules varying by forum type. The same flagpole, fund, or meeting room can fall on either side of the line depending on how the state designed the program and how officials ran it in practice.
The categories received their classic statement in Perry Education Association v. Perry Local Educators’ Association, decided in 1983, a dispute concerning access to a school mail system. A traditional public forum consists of places held in trust for assembly and debate across long usage, with streets, sidewalks, and public parks as the standard examples named in cases reaching back to Hague v. Committee for Industrial Organization in 1939. In such places, content based restrictions face strict scrutiny, viewpoint discrimination is forbidden, and reasonable rules governing time, place, and manner remain permissible where they serve a significant interest and leave open alternative channels. The state holds these spaces as a trustee for public discussion rather than as a proprietor free to select favored messages.
A designated public forum arises when the state intentionally opens a nontraditional place or program for public discourse, as a university did with meeting rooms in Widmar v. Vincent and as a city may do with a theatre or a channel set aside for public access. While the forum remains open, the standards of a traditional forum apply, including the bar on viewpoint discrimination. The state may close a designated forum by changing its policy and practice, but it may not keep the forum open in name while admitting only speakers who support official policy. Evidence of intent comes from policy statements, past practice, and the nature of the property, the same evidentiary sources that decide control in government speech cases, read toward a different conclusion.
A limited public forum is opened for particular groups or particular subjects, such as a school facility reserved for youth organizations or a fund established for student groups at a public university, the setting in Rosenberger v. Rector and Visitors of the University of Virginia. Within the limits the state has set, viewpoint discrimination remains forbidden, and exclusions must be reasonable in light of the purpose of the forum. A nonpublic forum covers property not opened for public discourse at all, such as a military base in Greer v. Spock, internal mail systems, and airport terminals treated as nonpublic in International Society for Krishna Consciousness v. Lee. In a nonpublic forum, restrictions need only be reasonable and viewpoint neutral. The Boston flagpole in Shurtleff v. City of Boston failed as government speech and was analysed as a forum opened broadly to community groups, which is why refusal of the Christian flag amounted to viewpoint discrimination. Had Boston selected flags as its own statements, forum doctrine would never have applied. Sequence is therefore decisive in litigation. A court first decides who is speaking, and only if private speakers occupy the space does it classify the forum and apply the corresponding standard.
Federal, State and Municipal Speakers: Whose Message Counts
The doctrine operates at every level of American government, and the identity of the speaking body shapes how attribution and control are assessed. At the federal level, messages issued through institutions such as the United States Mint, the Bureau of Engraving and Printing, and federal public health agencies carry a long history of direct state authorship. Currency designs, postage stamps, and official census materials have been understood for generations as statements of the national government itself, selected through administrative processes answerable to Congress and the President. When a federal agency commissions an advertising campaign or prescribes the wording on a federal form, the control element is usually straightforward because a single chain of authority approves the final content before release. That chain of authority is the same structural mechanism examined in the law of presidential removal power, where control over officers is what keeps executive action answerable to an elected head.
State governments present a more varied picture because authority is divided among legislatures, governors, agencies, and independent boards. A state department of transportation that designs highway markers, a state university that sets degree requirements, and a state licensing board that prescribes course content each exercise a different kind of control, and courts examine the actual approval practice rather than the formal label attached to the program. The specialty plate program in Walker v. Texas Division, Sons of Confederate Veterans illustrated the state level clearly: the Texas Department of Motor Vehicles Board retained final authority to approve or reject every design, and that retained authority supported treatment of the plates as speech of Texas itself.
Municipal speakers raise the attribution question in its sharpest form because city property often sits in spaces residents use for private expression. A city council that selects monuments for a public park, maintains a flagpole at city hall, or publishes a municipal newsletter speaks through channels that residents may also seek to use. The flagpole program in Shurtleff v. City of Boston showed the danger of assuming municipal ownership equals municipal authorship: Boston owned the pole, yet the city had raised 284 flags through a program with no content review, and the absence of exercised control defeated the claim that each flag was a message of the city.
School boards occupy a distinctive position within the municipal and state layers. A local board of education selects textbooks, approves curricula, and prescribes graduation requirements, and those choices have long been treated as educational judgments of the district rather than as a forum for the private views of every teacher or student. The reasoning rests on function: a curriculum must choose among competing histories, scientific accounts, and literary works, and a requirement of viewpoint neutrality across every selection would make coherent instruction impossible. Courts therefore give boards wide latitude over what a school teaches, while treating student-initiated expression and access to library materials under different rules that protect private speech from suppression.
Running through each level is the same justification for tolerating viewpoint choice in official speech. An elected body, or officials answerable to one, can be removed at the next election, budgets can be revised by a legislature, and a mayor or governor can be held responsible for a slogan, a monument, or a campaign in a way no private speaker can be. That electoral accountability does not answer every objection, but it explains why the doctrine permits a city council to endorse one commemorative message and decline another while forbidding the same council from excluding one private speaker from an open forum because of disagreement with a viewpoint. The speaking institution must be identifiable, the message must be traceable to a decision that institution actually made, and a political remedy must exist for citizens who reject the message. Where any of those conditions fails, the claim that a message counts as the voice of a federal, state, or municipal speaker weakens considerably.
The Critics and the Swallowing Problem
A substantial body of legal scholarship warns that the doctrine expands by relabeling, and the warning has a precise structure. Once a court accepts that speech inside a government program is speech of the government itself, the ordinary rule against viewpoint discrimination falls away, funding conditions that would be unconstitutional as direct regulation become permissible, and private speakers who believed a channel was open to them discover that the channel was an official channel all along. The critics named here return to that sequence from different directions, yet each identifies the same structural risk: classification decides the outcome before any balancing begins.
Caroline Mala Corbin has examined the boundary between compelled speech and government speech and argued that the two doctrines blur in ways that burden dissenters. When the state compels a private party to carry or fund a message and then defends the compulsion by calling the message its own, the individual objection recognized in cases such as West Virginia State Board of Education v. Barnette and Wooley v. Maynard loses force. Corbin treats that movement as a transfer of protection from the dissenting individual to the speaking state, and the concern sharpens where the compelled carrier is closely identified with the message in public perception.
Abner S. Greene approaches the problem from the direction of constitutional rights held against the government. In his account, official speech is necessary and legitimate, yet a government that defines the scope of its own speech also defines the space left for private rights, and courts that defer to the label allow the state to contract the First Amendment by definition. Greene presses for attention to the effect on the private speaker who is displaced, funded, or conscripted, rather than attention only to the institutional needs of the speaking agency.
Mary-Rose Papandrea has focused on media institutions and the dependence of the press on official channels. Her analysis highlights a setting where formal classification misses practical reality: government officials communicate through press briefings, official accounts, and access decisions that shape coverage, and a doctrine that treats every official communication as fully discretionary speech of the state can insulate exclusion of disfavored outlets from the scrutiny that would apply to exclusion from a forum. The worry is less about monuments than about gatekeeping over the flow of public information.
Robert C. Post supplies the theoretical frame that connects the other accounts. Post distinguishes the domain of democratic legitimation, where government must be able to form and express a position, from the domain of public discourse, where citizens form the opinions to which government must remain responsive. Government speech doctrine becomes dangerous, in that frame, when the state imports the managerial freedom appropriate to the first domain into the second and manages public discourse itself. A park, a funding program, or a communications platform reclassified as an instrument of official messaging escapes the viewpoint neutrality that keeps public discourse open.
The swallowing problem follows directly. Every forum the state creates can be redescribed after the fact as a vehicle for official messages, every subsidy can be redescribed as a purchase of the government voice, and every license condition can be redescribed as a script. If redescription alone sufficed, viewpoint neutrality would apply only where the state had neglected to assert authorship, and the exception would consume the rule. The critics therefore converge on a demand for proof external to the assertion itself: a history of the medium being used for official messages, genuine public attribution, and control exercised in practice. Assertion without those proofs, in the critical account, is not authorship but avoidance.
The Defense: Democratic Accountability and the Need to Govern
The strongest defense of the doctrine begins with a practical observation: governing requires choosing messages, and a state forbidden to choose cannot act. Mark G. Yudof supplied the early systematic account in When Government Speaks, published in 1983, at a time when the case law consisted largely of forum and funding decisions without a unifying theory. Yudof argued that official communication is not an incidental activity of modern government but a central instrument of policy, used to educate, to warn, to recruit, to reassure, and to explain the law itself. A constitutional rule that subjected every official message to viewpoint neutrality would not produce a neutral state; it would produce a state unable to state a position on the policies it exists to carry out.
Justice Alito gave the accountability version of the defense its canonical judicial form in the majority opinion in Pleasant Grove City v. Summum, decided in 2009. The opinion reasoned that government speech is accountable through the political process in a way private speech is not: officials who approve a monument, a motto, or a public campaign answer to voters, and citizens who object to the message can organise, campaign, and replace the officials responsible. The check on a foolish or offensive official message, in that account, is electoral rather than judicial, and courts intervening to impose neutrality would displace a remedy the Constitution already supplies through democratic institutions.
The functional argument extends the accountability argument. A public health campaign against smoking must be permitted to say that smoking causes disease without carrying an equivalent message from a tobacco manufacturer. A school curriculum must be permitted to teach that the Earth orbits the Sun without presenting a contrary account as equally valid. A national currency must carry one set of symbols and inscriptions rather than a rotating selection demanded by every citizen who objects to the design. In each setting the state is not suppressing a forum; the state is performing a task that consists of expression, and the task fails if the expression must be neutral among viewpoints.
Defenders also point to the alternative doctrines waiting on the other side of classification. If official messages were treated as forums, every monument selection would oblige a city to accept a rival monument, every license plate design rejection would become a viewpoint discrimination claim, and every change to a curriculum would invite litigation from any group whose preferred material was omitted. The resulting accumulation would be unmanageable in a park, incoherent in a classroom, and paralysing in a public information campaign. Forum doctrine was built for property the state holds open for private expression, and applying that doctrine to the state speaking in its own capacity misdescribes both settings.
The defense, stated at its strongest, still contains an internal limit that responsible versions of the argument accept. Accountability works only where the speaking official is genuinely answerable and the message is genuinely attributable, so the public can connect the words to the officeholder who approved them. A message issued through layers of contractors, licensees, and private intermediaries diffuses responsibility until no voter can identify whom to punish or reward. The Yudof account and the reasoning in Pleasant Grove City v. Summum therefore support the doctrine most strongly in its core applications, monuments chosen by a council, campaigns approved by an agency head, curricula adopted by an elected board, and support it weakly or not at all where attribution and control dissolve. Democratic accountability is the justification, and the justification fails on its own terms when accountability cannot be traced.
Hard Cases at the Edges
The edges of the doctrine are where classification does the most work and where the factors point in different directions. Public library book selection is a leading example. Librarians at institutions such as a municipal public library choose which books to acquire with limited budgets and limited shelf space, and every acquisition decision excludes other titles. Selection has generally been treated as a discretionary institutional judgment, because a library cannot purchase everything and a neutrality rule across acquisitions would be unworkable. Removal of books already on the shelves presents a different question, because the motive for removal can be suppression of ideas rather than management of a collection.
That distinction stood at the center of Board of Education, Island Trees Union Free School District No. 26 v. Pico, decided in 1982. A plurality of the Supreme Court concluded that a school board may not remove books from a school library simply because board members dislike the ideas contained in those books and wish to prescribe orthodoxy in politics, nationalism, religion, or other matters of opinion. The judgment produced no single majority rationale, and its limits have been debated for decades, yet the plurality reasoning marks a boundary the government speech label does not erase: discretion over acquisition is broad, while removal motivated by suppression of a viewpoint invites constitutional scrutiny that ordinary collection management does not.
Public officials blocking critics on social media accounts created a newer edge problem with a different doctrinal door. In Lindke v. Freed, decided in 2024, the Supreme Court set a test for when an official who blocks a constituent or deletes comments engages in state action. Under that test, an official acts in an official capacity only where the official possessed actual authority to speak on behalf of the state and purported to exercise that authority in the relevant posts. The inquiry is fact specific: disclaimers, account history, use of government staff or resources, and the nature of the posts all bear on whether a page that looks personal functions as an instrument of the office. Classification again precedes the merits, because private accounts remain private even for officeholders, while accounts held out as channels of the office carry constitutional obligations toward the constituents who are excluded.
Government advertising campaigns sit closer to the core of the doctrine, yet even there the edges show. A campaign approved and funded by an agency, such as a federal anti-drug campaign or a state tourism campaign, is official speech in the ordinary case, and the agency may choose a slogan and reject alternatives. Difficulty arises when a campaign recruits private messengers, conditions funding on delivery of an approved script, or presents a paid message as the independent view of a community group. Attribution then diverges from authorship, and the public receives a message that appears private while its content was prescribed by the state.
Vanity and specialty plates after Walker v. Texas Division, Sons of Confederate Veterans illustrate how a single factor can dominate an edge case and leave residue for later disputes. Texas prevailed because the board approved every design and plates had historically carried state messages, yet drivers display plates on private vehicles, pay extra fees for specialty designs, and propose many of the designs themselves. Justice Breyer and the dissent treated those features as evidence of private expression on government property. The majority classification settled plate programs with board approval, while leaving open how much driver initiation and payment would matter in a program with weaker state control. Edge cases of that kind recur because modern government communicates through hybrid channels by design, and hybrid channels resist a single label applied to the whole program rather than to the specific message in dispute.
What Is Settled and What Remains Open
Several propositions command broad agreement across the case law. Permanent monuments selected by a government for a public park are speech of that government, as Pleasant Grove City v. Summum established, and a city may accept one donated monument and decline another without creating a forum. Specialty license plates issued under a program with state design approval are likewise government speech after Walker v. Texas Division, Sons of Confederate Veterans. Compelled financial support for an official campaign stands on different footing from compelled support for private speech, a distinction traced through Johanns v. Livestock Marketing Association, where assessments funded a federal beef promotion program classified as a message of the government. The three considerations identified in Shurtleff v. City of Boston, the history of the medium, public attribution, and the extent of control exercised over content, supply the governing framework for classification disputes. The viewpoint limit recognized in Matal v. Tam is equally settled in the opposite direction: registration of a trademark does not convert a private mark into government speech, and a statutory bar on disparaging marks imposed viewpoint discrimination on private expression.
The open questions cluster where private speakers deliver messages under state direction. Licensed professional instruction is the sharpest example. In Khatibi litigation, docket No. 25-1240, the question concerned whether mandatory instruction delivered by licensed private instructors under a state-prescribed program becomes speech of the state or remains speech of the instructors subject to First Amendment protection. The Supreme Court denied certiorari on November 16, 2026. A denial of certiorari carries no ruling on the merits and settles no circuit conflict, so the classification of state-mandated teaching by private licensees remains governed by divergent lower court reasoning.
Commissioned speech by contractors presents a parallel uncertainty. A government that hires a firm to design a campaign clearly speaks through the firm where the agency drafts or approves every word, yet long-term contracts for communications services, educational content, and public engagement often leave drafting discretion with the contractor. Courts have not produced a general rule allocating authorship along that spectrum of discretion, and outcomes turn on contract language and approval practice in individual disputes.
Algorithmic curation by government platforms and official accounts online raise questions the older factors fit awkwardly. A state agency that operates a website, ranks resources through a recommendation system, or moderates comments through automated filters makes expressive choices at a scale no human approval process reviews item by item. History offers little guidance for media with no settled tradition, attribution is ambiguous where users supply the content being ranked, and control is exercised through design choices made long before any individual message appears. Whether the Shurtleff factors can govern curation decisions, or whether forum analysis returns for interactive state platforms, remains unresolved. Official accounts online compound the difficulty, because a single account can carry official announcements, personal commentary, and constituent discussion in an interleaved stream. Settled doctrine answers the monument and the plate; the feed, the mandated lesson, and the contractor draft await a classification rule equal to their structure.
A Worked Example: One Flagpole, One Plate, One Course
Consider first a city flagpole opened to all applicants without review. A municipal program invites community groups to request a flag raising at city hall, applications are granted in the order received, no official reads the flag design or the accompanying message before approval, and 284 flags are raised over the life of the program for groups of every description. History offers no support for official authorship, because flags flown by private groups at a shared pole have not traditionally carried messages of the city in such a program. Attribution points the same way: an observer familiar with the open application process understands the flag to express the views of the requesting group, in the same manner as a speaker at an open meeting. Control is absent in practice, not merely limited on paper. Under the reasoning of Shurtleff v. City of Boston, the flag is private speech in a forum the city created, and exclusion of a group because officials dislike its message is viewpoint discrimination. The city could redesign the program into a genuine official channel by reserving the pole for flags the council itself selects, but a program that never exercises selection cannot claim the benefit of selection after a dispute arises.
Consider next a specialty plate program with a design approval board. A state legislature authorises specialty plates, a board reviews every proposed design, the board rejects designs that conflict with state messaging standards, plates carry the state name, and the state has issued plates bearing official slogans and commemorations for decades. History supports official authorship because license plates have long served as a medium for state identification and state messages. Attribution supports the same result because motorists and observers associate plates, issued by the state and required by law, with the state itself even where a driver selected a specialty design. Control is exercised in fact through the approval board. Under Walker v. Texas Division, Sons of Confederate Veterans, the plate is government speech, and rejection of a proposed design does not violate the First Amendment rights of the proposing organization. The driver payment and private proposal weaken the claim at the margins, yet exercised approval authority carries the classification.
Consider finally a licensing course with a state-prescribed topic but private instructors. A state requires licensees in a profession to complete a course on a topic the legislature has mandated, accredited private instructors write the lesson plans and deliver the teaching, the state prescribes the subject and the required hours but does not review scripts, slides, or classroom statements, and certificates issue in the name of the instructional provider. History is mixed, because professional instruction has traditionally been speech of the teacher even when the state required attendance. Attribution favors the instructors, because students understand a lesson to express the judgment of the person teaching, not a verbatim message of the legislature. Control reaches the topic and the schedule while stopping short of content. That combination lands closer to private speech delivered under regulation than to government speech, and the dispute posture of Khatibi, docket No. 25-1240, shows why the distinction matters: if the teaching were classified as a message of the state, the state could dictate the viewpoint of every lesson; classified as private, a mandate compelling instructors to teach a state-favored position faces compelled speech scrutiny. The three scenarios diverge not because the state is involved in each, since involvement is constant, but because history, attribution, and exercised control align in the plate program, fail together at the flagpole, and split in the classroom.
Speaker Identity Decides Cases Before the Merits
Classification of the speaker is the decision that quietly determines most government speech disputes, and later argument about discrimination or compulsion usually ratifies a conclusion reached at that first step. A monument, a plate, a flag, and a lesson can each be described plausibly as state action in a loose sense, because public money, public property, or public authority touches each one. The doctrine insists on a narrower question, whether the state adopted the specific message as its own voice through a medium historically used for official statements, under circumstances where the public attributes the words to the state, and through control the state actually exercised over what was said. Where those elements converge, the message belongs to the government and viewpoint choice follows. Where they diverge, the message remains private and the ordinary protections of the First Amendment apply with full force.
The single point worth carrying away from the whole line of cases is that funding, licensing, and regulation describe relationships between the state and a speaker, while authorship describes a relationship between the state and a message, and only the second relationship triggers the doctrine. A researcher tracking a sequence of decisions across decades may keep the distinctions straight in a personal case notebook, yet the distinction matters far beyond any record system, because courts apply it to live disputes over curricula, campaigns, platforms, and mandated instruction. Confusing support for a speaker with adoption of speech collapses the boundary that keeps public forums open and keeps compelled speakers free.
That discipline also explains why the sequence of analysis matters in litigation and in program design alike. A court that begins with the merits, asking whether exclusion seems unfair or whether a message seems valuable, will reach results that shift with the sympathies of the panel. A court that begins with authorship, examining the approval record of the Texas Department of Motor Vehicles Board, the selection minutes of the Pleasant Grove City Council, or the unreviewed application file behind 284 flags in Boston, produces a classification that later courts can test against evidence. Program designers face the mirror image of the same lesson: a government that intends to speak must build the record of speaking, through selection, approval, and attribution, before a dispute forces the question.
The structural implication reaches beyond any single program. A constitutional order that permits government to govern through speech must also prevent government from expanding its voice by declaration, absorbing every channel it funds or regulates into an immunity from viewpoint rules. The three considerations from Shurtleff v. City of Boston serve that preventive function when courts apply them as evidence of authorship rather than as labels to be recited. History disciplines novelty, attribution disciplines appearance, and exercised control disciplines assertion. A state that wishes to speak may do so openly, through channels the public recognizes as its own and through decisions for which officials answer. A state that wishes to regulate private speakers must accept the constitutional limits on regulation. Speaker identity, established by proof rather than proclamation, keeps those two powers distinct.
Frequently Asked Questions
Q: What is the government speech doctrine?
The government speech doctrine holds that when government speaks on its own behalf, the Free Speech Clause does not require viewpoint neutrality in the message selected. A city choosing a monument, a state choosing a license plate design, and an agency choosing a campaign slogan each make expressive decisions that would be unconstitutional if imposed on private speakers in a forum. The doctrine rests on the premise that government must be able to state a position in order to govern, and that the remedy for disagreement with an official message lies in elections and legislative oversight rather than in a judicial order requiring equal carriage of opposing views. Classification as government speech therefore removes the dispute from ordinary forum analysis and places it under political accountability.
Q: Why does the First Amendment not limit government speech?
The First Amendment restricts government regulation of private expression, and its text and structure assume a distinction between the state as regulator and the state as speaker. When the state speaks for itself, requiring neutrality would prevent adoption of any coherent policy position, because every official statement favors one view over alternatives. A public health agency endorsing vaccination, a legislature commemorating a historical event, and a school board adopting a curriculum each express a viewpoint by necessity. Courts reason that democratic accountability supplies the relevant constraint in that setting: voters can replace officials whose messages they reject, legislatures can revise budgets and mandates, and other constitutional provisions still bind official action even where the Free Speech Clause does not impose neutrality.
Q: How do courts decide whether speech is government speech or private speech?
Courts examine three considerations drawn from Shurtleff v. City of Boston: the history of the medium, public attribution, and the extent of control government exercises over content. History asks whether the channel has traditionally been used to convey official messages, as monuments and license plates have been. Attribution asks whether a reasonable observer familiar with the program would understand the message as a statement of the government or of a private participant. Control asks whether officials actually review, approve, edit, or reject content in practice, rather than merely owning the property or funding the program. No single consideration decides every dispute. Classification follows the combined weight of the evidence about authorship of the specific message in question.
Q: What did Pleasant Grove City v. Summum decide about monuments in public parks?
In Pleasant Grove City v. Summum, decided in 2009, the Supreme Court held that permanent monuments in a public park are government speech. Pleasant Grove City had accepted donated monuments, including a Ten Commandments monument, while declining a proposed monument from the Summum religious organization. The Court reasoned that governments have historically used monuments to express official messages, that observers attribute park monuments to the property owner, and that the city exercised selection authority by accepting some donations and rejecting others. Because the monuments spoke for the city, the Free Speech Clause did not require acceptance of the Summum monument under forum doctrine. The decision emphasised electoral accountability as the check on monument selection by an elected council.
Q: Why were specialty license plates treated as government speech in Walker?
In Walker v. Texas Division, Sons of Confederate Veterans, decided in 2015, the Supreme Court classified Texas specialty license plates as government speech for three converging reasons. License plates had historically served as a medium for state identification and official messages, carrying the state name and state selected designs. Observers closely associated plates with Texas because plates are government issued, state owned in legal form, and required for lawful driving. Texas also exercised control through the Department of Motor Vehicles Board, which reviewed and held final approval authority over every specialty design. On that combined record, rejection of a proposed Confederate flag design was a choice among official messages rather than viewpoint discrimination against a private speaker in a forum.
Q: What did Matal v. Tam hold about trademarks and government speech?
In Matal v. Tam, decided in 2017, the Supreme Court rejected the argument that federal trademark registration converts a private mark into government speech. The case involved the disparagement clause of the Lanham Act, under which the Patent and Trademark Office had denied registration to the mark The Slants, adopted by musician Simon Tam. The Court reasoned that trademarks are private expression: the government does not create the mark, does not edit its content, and registration alone does not make the mark a statement of the United States. Because the speech remained private, the disparagement clause constituted viewpoint discrimination and violated the First Amendment. The holding marks a firm outer limit on expansion of the doctrine through mere government involvement in a registration system.
Q: How is government speech different from a public forum?
A public forum is property or a channel government has opened for private expression, and inside a traditional or designated public forum, exclusion of a speaker because of viewpoint is presumptively unconstitutional. Government speech involves the opposite relationship: the message is attributed to the state itself, selected by officials, and subject to political rather than judicial control over its viewpoint. The same physical space can contain both categories at different times, as a park can host private rallies under forum rules while its permanent monuments express official choices. Classification therefore turns on the specific expressive activity under review, not on government ownership of the location alone. Ownership establishes authority to regulate a forum; authorship of a message establishes government speech.
Q: Can the government be forced to carry a message it disagrees with?
Where a message qualifies as government speech, government generally cannot be forced through the Free Speech Clause to carry a contrary message it has declined to adopt. That principle explains the outcomes in the monument dispute in Pleasant Grove City v. Summum and the plate dispute in Walker v. Texas Division, Sons of Confederate Veterans: a requirement to accept every proposed addition would transfer control over official expression to private applicants. The principle has limits outside its classification premise. If the channel is in fact a forum for private speech, compelled acceptance is not the issue and exclusion faces strict scrutiny. Other constitutional provisions, statutes, and contractual commitments can also require carriage or prohibit particular official messages in specific settings.
Q: Does government funding of a program make everything said in it government speech?
Funding alone does not transfer authorship of every statement made inside a funded program to the government. Courts distinguish programs where government hires speakers to deliver an official message from programs where government funds private expression to promote diversity of views, a distinction central to Legal Services Corp. v. Velazquez and to university funding disputes such as Board of Regents of the University of Wisconsin System v. Southworth. Relevant evidence includes the program design, the degree of content control retained, public attribution, and whether participants are understood to speak for themselves. A grant condition that defines the scope of an official project receives more deference than a condition that leverages funding to control private expression outside the project. The inquiry remains message specific.
Q: What is the difference between government speech and compelled speech?
Government speech doctrine concerns messages attributed to the state, while compelled speech doctrine concerns private persons forced to utter, carry, or subsidise messages attributed to another. The classic compelled speech decisions, West Virginia State Board of Education v. Barnette and Wooley v. Maynard, protected individuals against being made instruments for an ideological message. Difficulty arises where the doctrines meet, as in disputes over mandatory fees, license conditions, and assessments funding a promotional campaign. Classification of the funded message decides which doctrine governs: a message genuinely belonging to government receives deference under Johanns v. Livestock Marketing Association, while a message belonging to private speakers triggers protection for the objecting payer or carrier. Authorship is again the threshold question.
Q: How did Shurtleff v. City of Boston change the government speech test?
Shurtleff v. City of Boston, decided in 2022, clarified government speech analysis by organising classification around history, attribution, and control, and by warning against mechanical application of any single factor. Boston had allowed private groups to raise flags at city hall through an application program under which 284 flags were raised without content review, then declined a request to raise a Christian flag. The Supreme Court held the program, as operated, had not made each flag a message of the city. Boston owned the pole but had created a channel dominated by private participation and had exercised no selection over content. The decision narrowed expansive readings of Pleasant Grove City v. Summum and Walker v. Texas Division, Sons of Confederate Veterans by requiring proof of exercised control.
Q: What role does government control over content play in the doctrine?
Control over content is the most direct evidence that government has adopted a message as its own, and courts examine control as actually exercised rather than as reserved in a policy document. Approval of final designs in Walker v. Texas Division, Sons of Confederate Veterans, selection of accepted monuments in Pleasant Grove City v. Summum, and absence of any review across 284 flags in Shurtleff v. City of Boston explain the divergent outcomes in the leading trilogy. Control can be shown through drafting, editing, veto authority used in practice, or prescribed scripts. A contractual right to intervene that is never used carries less weight, and ownership of property or provision of funding without content direction carries less weight still. Exercised control connects attribution to a responsible decision maker.
Q: Why does the history of a medium matter in government speech cases?
History matters because a medium used for official messages over a long period gives the public a basis for attributing a new message in that medium to the government. Monuments in public parks, inscriptions on currency, and designs on license plates acquired settled public meanings through repeated official use across generations. A court can therefore infer attribution with some confidence. A newer or mixed medium, such as a flagpole opened to community groups or an interactive online platform, lacks that settled meaning, and observers may reasonably attribute messages to the private participants who supplied them. History also disciplines government claims: an agency asserting authorship over a channel that has functioned as an open forum must explain the departure from the established character of the medium.
Q: Can a public school choose its curriculum without violating the First Amendment?
A public school board generally may choose a curriculum, select textbooks, and prescribe required subjects without violating the First Amendment, because curricular decisions express the educational judgment of the district and coherent instruction requires selection among competing materials and viewpoints. Courts give boards broad discretion over what is taught in classrooms, a discretion connected to democratic control of public education through elected or appointed boards. That discretion is not unlimited in every direction. Student private expression receives separate protection under Tinker v. Des Moines Independent Community School District, library removal motivated by suppression of ideas raised distinct concerns in Board of Education, Island Trees Union Free School District No. 26 v. Pico, and other constitutional provisions constrain curricular choices that cross independent legal boundaries.
Q: Are government social media accounts government speech?
Official social media accounts present a mixed classification problem because a single account can carry agency announcements, statements by an officeholder, reposted material, and exchanges with constituents. Posts drafted and approved by an agency for official release resemble government speech, while comment threads opened to the public can function as forums where viewpoint based exclusion faces constitutional limits. Lindke v. Freed, decided in 2024, addressed the related state action question for an official who blocked a critic, requiring actual authority to speak for the state and a purported exercise of that authority in the posts at issue. Account labels, disclaimers, use of public resources, and posting history all inform the analysis. No single rule classifies every feature of an account at once.
Q: What happens when the government speaks through a private contractor?
Use of a private contractor does not by itself decide authorship in either direction. Where an agency drafts the message, prescribes the wording, or approves the final product before release, the contractor functions as an instrument of the agency and the output is ordinarily treated as government speech. Where a contractor exercises independent professional judgment, creates content without agency review, and is publicly identified as the source, the output retains a stronger claim to private character even though public funds paid for the work. Contract terms, approval practice, attribution to the public, and the degree of discretion left to the contractor all bear on classification. Courts resolve such disputes on the operational record rather than on the existence of a government contract alone.
Q: Does a denial of certiorari in Khatibi v. Yip settle the law on licensed teaching?
A denial of certiorari in Khatibi v. Yip, docket No. 25-1240, on November 16, 2026, settles no legal question about licensed teaching. The Supreme Court denies review in the large majority of petitions for reasons unrelated to agreement with the judgment below, including vehicle problems, an underdeveloped record, and absence of a mature conflict among lower courts. A denial leaves the lower court judgment in force between the parties while creating no precedent binding other courts. The underlying classification problem therefore remains open: mandatory instruction delivered by licensed private instructors under a state-prescribed topic can still be characterized as private professional speech, as compelled speech, or as government speech by different courts reasoning from history, attribution, and exercised control.
Q: How does the government speech doctrine affect public libraries?
Public libraries exercise broad discretion in selecting materials for a collection, because budgets and space require choices and every acquisition excludes alternatives. That selection function resembles other institutional judgments courts have treated deferentially. Removal and access restrictions attract closer attention where evidence suggests a motive to suppress disfavored ideas, a concern articulated by the plurality in Board of Education, Island Trees Union Free School District No. 26 v. Pico for school libraries in 1982. Public library meeting rooms, display spaces, and internet terminals can raise separate forum questions distinct from collection decisions. The doctrine therefore affects libraries unevenly: acquisition receives wide latitude, while viewpoint motivated removal, exclusion from an opened channel, and restrictions on patron private expression face more demanding constitutional analysis.
Q: What limits remain on government speech outside the First Amendment?
Classification as government speech removes the Free Speech Clause requirement of viewpoint neutrality, while leaving other constitutional and legal limits fully operative. The Establishment Clause restricts official religious endorsement, the Equal Protection Clause restricts messages and programs that deny equal treatment on protected grounds, and the Free Exercise Clause restricts targeting of religious practice. Statutes, appropriations conditions, administrative procedure requirements, and state constitutions impose additional constraints on particular agencies and programs. Political limits also operate in practice: elections, legislative oversight, budget control, public records laws, and press scrutiny expose official messages to challenge through democratic channels. Government speech is therefore a doctrine about one clause and one requirement, not a general immunity for official expression from law.
Q: Why do critics say the government speech doctrine is dangerous?
Critics argue that the doctrine is dangerous because classification can be manipulated to escape constitutional limits that would otherwise apply. A government that loses under forum doctrine has an incentive to redescribe an open channel as an official messaging vehicle, and a government that funds private expression has an incentive to claim authorship over the resulting speech. Scholars including Caroline Mala Corbin, Abner S. Greene, Mary-Rose Papandrea, and Robert C. Post have warned, from different premises, that expansive classification displaces dissenters, conscripts carriers, insulates gatekeeping, and allows management of public discourse itself. The shared concern is structural: if assertion of authorship suffices, viewpoint neutrality survives only where government forgets to assert control, and the exception displaces the rule.