Sorry Losers and Haters: How Current Public Forum Analysis of Twitter Posts Allows for Silence of Dissent and Violations of Rights

I. Introduction^[1]

Say you are on Twitter scrolling past sports highlights, animal-adoption videos, and news stories about the upcoming election. While scrolling through your timeline, you notice that someone you follow retweeted an account that prides itself as an alternative news source to the mainstream media.^[2] In the retweet, the news source reported an erroneous figure and editorialized on the misconstrued statistic.^[3] You reply with facts reported from other sources.^[4] The alternative news source blocks you.^[5]

A few days later, you talk to your friend at lunch. In between sips of a margarita, your friend says he was looking at your senator’s Twitter account and was wondering what you thought of her. Confused, you go to your senator’s Twitter feed and only see boring tweets about financial grants and Senate procedural issues. Your friend shows you his view of the senator’s account. His version shows dozens of retweets containing controversial topics and disputed statistics, all from the same alternative news source that blocked you. Your feed is completely different from your friend’s feed. You cannot see your senator’s latest thoughts, nor can you respond to your senator or others talking to her. You are cabined into a muted discourse where you cannot fully participate in the political discussion. Your ability to interact with your elected representative’s public statements is restricted because of a private Twitter user who blocked you.

As the law stands, this hypothetical is a legal reality. Current First Amendment holdings posit that government Twitter accounts—including any replies from third parties that are posted on them—are public forums, yet these holdings are fundamentally misguided and ham-fisted in application.^[6] First, as applied, they create an irresolvable tension between the public forum doctrine and the freedom of association. This conflict arises when a government account retweets a private account; the space below a private account’s tweet where users can exercise their freedom of association by excluding other users, and the replies to the government retweet—a public forum—become the same space. This creates a predicament. If the public forum designation remains under the private account’s tweet after the government retweet, that prohibits the private user from blocking others, thereby infringing on the private user’s freedom of association; however, if the private user retains the freedom to block others, that effectively allows the private user to control speech within the public forum, thus dissolving the public forum’s integrity.

Second, the holdings empower crafty politicians—who already routinely and proudly violate the First Amendment—to exploit the gap between a private account’s freedoms and a government account’s obligations, selectively excluding others from public forums.^[7] Currently, a government official may use a government account to retweet a private account and exercise the private account’s freedom of association to block dissidents from the authoring private account, effectively preventing them from interacting with the government account’s retweet. This maneuver allows government officials to censor content within a public forum under their government account’s retweet while maintaining that they have fulfilled their forum obligations because the government account did not block anyone; however, the dissidents are excluded from the forum by a government official all the same.^[8]

The solution to this conflict and gap is a proposed two-step test that maintains public forum integrity and prioritizes individual freedoms.^[9] The proposed test first designates original accounts and retweeted accounts as either government or private accounts. The test finds that government accounts may retweet other government accounts because both accounts hold identical First Amendment obligations without rights conflicts. When a government account retweets a private account, a second step—or control analysis—is triggered. When the private account is controlled by a government official in a private capacity, that private account loses its freedom of association and must unblock others. This result prioritizes the right to a public forum over officials’ private freedom of association because officials voluntarily create a public forum underneath their own private tweet.

However, if the private account is not controlled by the government official represented by the retweeting government account, the test finds that such a retweet violates the First Amendment. This result acknowledges that the ensuing public forum created by the retweet destroys the retweeted account’s freedom of association. Additionally, the government official cannot ensure public forum obligations are met because the retweeted private account may exercise viewpoint discrimination on behalf of the government account, excluding others from the public forum in the replies section under the government retweet. The test’s proposed outcome furthers public forum objectives, accounts for Twitter’s often account-wide settings, and protects fundamental freedoms. These outcomes also encourage government speech through proper channels on Twitter, creating spaces for public debate while simultaneously maintaining individual freedoms.

This Note addresses the fundamental incompatibility of online forum holdings and the First Amendment. Part II addresses Twitter’s mechanics and highlights current case law concerning First Amendment rights and the obligations of various Twitter accounts. Part II’s first section focuses on the courts’ analyses of Donald Trump’s Twitter activity and the associated forums. The second section focuses on the implication of a case concerning a county chair board’s Facebook page, where a court considered whether a government’s social media site selection could implicate First Amendment scrutiny. Part III demonstrates how the current holdings create an inherent irresolvable contradiction between individual rights and government obligations. This part also examines how the precedent allows public officials to routinely violate the First Amendment. Part IV lays out an alternative test that fits within current government speech framework and forum analysis and applies these principles more thoroughly. Part V notes that government social media behavior and norms must change and describes the ways that government accounts can respect individual rights to maintain proper public forums. Finally, Part VI introduces potential future issues and challenges to First Amendment holdings on social media.

II. Twitter and the Law

A. Twitter’s Mechanics

“Someone retweeted my tweet. Then, another account quote tweeted it, and that was retweeted by the governor’s account.” Twitter users speak in a vocabulary that would lead to very different inferences in other contexts.^[10] This section will provide Twitter vocabulary definitions and settings explanations.

This Note will refer to any Twitter account run by a government entity, office, or official as a government account.^[11] Any account that is not run by a government entity, office, or official, regardless of visibility settings, will be referred to as a private account.^[12] A protected account will refer to Twitter accounts with settings that limit the account’s visibility to approved followers.^[13] An account that is not protected will be referred to as a public account.^[14]

Users who tweet from public accounts may be retweeted, or reposted, by any other accounts.^[15] Their tweets can also be quote tweeted, placing the content of the original tweet under a comment or media the quote tweeter adds, in one tweet authored by the quote tweeter.^[16] Aside from retweets and quote tweets, any account can reply to tweets from a public account.^[17] A reply is a response to the original tweet that can also be replied to by any account—creating many different conversation threads underneath a tweet.^[18]

Protected accounts differ in several respects from public accounts. Protected accounts settings affect the visibility of every kind of tweet from the account.^[19] Only approved followers can see and reply to any kind of tweet from protected accounts, including protected accounts’ replies to public accounts.^[20] Tweets from protected accounts cannot be retweeted by other users.^[21]

Aside from account-level settings, any users on Twitter can limit who can reply to their tweets and can even change their replies’ visibility.^[22] At any point in time, authors may choose whether any account; only accounts they follow; or only mentioned accounts can reply to each tweet.^[23] However, this setting only applies to the reply function, not the retweet or quote tweet functions.^[24] Additionally, a tweet’s author may hide any reply, placing the reply in a category visible to other users but requiring an additional click to access.^[25]

Finally, any account can block any other Twitter account.^[26] When an account blocks another, the blocked account cannot view the account’s tweets, follow the account, or interact with the account at all.^[27]

B. Trending Cases: Current Law

In 1997, the Supreme Court addressed First Amendment rights on the Internet, or as the Court put it, “a vast number of documents stored in different computers all over the world.”^[28] In Reno v. ACLU, the Court distinguished government regulation of internet content from broadcast spectrum content, stating that never before “have the vast democratic forums of the Internet been subject to the type of government supervision and regulation that has attended the broadcast industry.”^[29] The Court also found the Internet was neither “invasive” nor “scarce,” two characteristics that justified the regulation of broadcast media.^[30] Therefore, the Court held that “[its] cases provide[d] no basis for qualifying the level of First Amendment scrutiny that should be applied to [the Internet].”^[31]

As the Internet has grown and become an integral part of daily life, the Court has carried First Amendment principles based on legacy media through modern times.^[32] However, cases that set out the First Amendment’s foundation for the Internet were largely decided before social media grew to shape American discourse and politics. Social media giants such as Facebook and Twitter came online in the early- to mid-2000s.^[33] Facebook grew from 1 million users in 2004, 500 million users in 2010, and 1 billion users in 2012, to 1.926 billion users in 2021.^[34] While Facebook grew, Twitter also enjoyed the social media boom after its debut in 2006.^[35] Twitter grew from 30 million users in 2010, to 255 million users in 2014, 336 million users in 2018, and 436 million users in 2021.^[36]

Due to this growth, social media sites have become part of the global, political environment. 189 countries have a presence on Twitter through an official government or governmental leader account.^[37] The number of governments using social media is matched by the number of reasons they use social media. Governments use social media to communicate health information, reflect diplomatic interactions, and signal alliances.^[38] State actors use social media during times of war to rally support or disseminate disinformation.^[39] Users based in Iran,^[40] Russia,^[41] and China^[42] have used social media to spy on countries and world leaders. Political candidates and revolutionaries have used social media to run for office^[43] and topple governments.^[44]

As social media users and uses have grown, disputes concerning users’ rights on the sites have grown as well.^[45] These disputes have involved local county officials,^[46] state senators,^[47] and United States Congress members.^[48] However, the highest-profile Twitter dispute to date involved then-President Donald Trump.^[49]

1. Presidential Tweets and Public Forums

While he was in office, former President Trump’s Twitter account, @realDonaldTrump, blocked several users.^[50] The blocked users and the Knight First Amendment Institute at Columbia University sued Trump.^[51] Though the Knight Institute’s account was not blocked by @realDonaldTrump, the Knight Institute sued because the Institute “desire[d] to read comments that otherwise would have been posted by the blocked [p]laintiffs, and by other accounts blocked by @realDonaldTrump.”^[52]

Along with standing arguments, Trump argued that because @realDonaldTrump was established before his presidency, the account was a private account and not subject to the First Amendment obligations of government accounts.^[53] Trump also claimed that blocking users was not a state action—triggering First Amendment scrutiny—because every Twitter user can use the blocking function.^[54] Alternatively, Trump asserted that his decision to interact with certain followers was part of his speech, and thus, protected by the government speech doctrine.^[55]

After finding that the plaintiffs had standing, the district court considered whether a public official could block a Twitter user in response to political views the user expressed.^[56] The court found that the individual plaintiffs’ criticism of Trump was political speech, and therefore, protected by the First Amendment.^[57] The court then concluded that Trump exercised control over the @realDonaldTrump account because he had the power to tweet and block from the @realDonaldTrump account.^[58] Thus, the court determined that the account was controlled by the government because the account was presented as being “registered to Donald J. Trump ‘45th President;’” the President’s tweets from @realDonaldTrump were official records that were required to be preserved under the Presidential Records Act; and, the @realDonaldTrump account was used for executive functions, such as appointing and removing officers and conducting foreign policy.^[59]

Therefore, the district court found that @realDonaldTrump’s tweets were government speech and exempt from forum analysis.^[60] However, the court held that the replies to the account were not government speech, and thus, subject to forum analysis.^[61] The court concluded that the interactive space under @realDonaldTrump’s account was a designated public forum because the space was generally accessible to the public, used by Trump to reach out to the public, and compatible with expressive activity.^[62] The court reasoned that because the interactive space was a public forum, viewpoint discrimination by the government was presumed impermissible when directed toward otherwise permissible speech.^[63] The parties did not dispute that the individual plaintiffs were blocked because the plaintiffs’ tweets were critical of Trump.^[64] Because the plaintiffs were blocked for that reason and excluded from a designated public forum, the court held that the plaintiffs’ continued exclusion was impermissible under the First Amendment.^[65] By blocking the individual plaintiffs and preventing them from viewing and interacting with his tweets, the court ruled that Trump excluded the plaintiffs from a designated public forum.^[66] Such exclusion was viewpoint discrimination, according to the court, and violated the First Amendment.^[67]

Trump appealed the ruling.^[68] The Second Circuit Court of Appeals held that “[b]y blocking the [i]ndividual [p]laintiffs and preventing them from viewing, retweeting, replying to, and liking his tweets, the President excluded the [i]ndividual [p]laintiffs from a public forum, something the First Amendment prohibits.”^[69] The Second Circuit also held that workarounds, such as creating a new Twitter account; logging out and taking screenshots of @realDonaldTrump’s tweets to post in original tweets; or logging out, looking up replies to @realDonaldTrump tweets, and logging back in to find the reply to @realDonaldTrump to which the user could also send a reply, were burdens on the individual plaintiffs’ speech and did not cure the constitutional violations.^[70] The Second Circuit agreed that the initial tweets from @realDonaldTrump were government speech.^[71] However, the Second Circuit wrote that the speech in question was the reply section featuring other Twitter users’ speech, not @realDonaldTrump’s tweets.^[72] Because the replies were controlled by other users, the retweets, likes, and replies of any other user were not government speech.^[73] Therefore, the Second Circuit also found that Trump violated the First Amendment by blocking other Twitter users.^[74] Ultimately, the Supreme Court granted certiorari and remanded the judgment to vacate as moot.^[75] By this time, Trump had left office and been suspended from Twitter.^[76]

2. Public Forums on Exclusionary Sites

While the Trump lawsuit is the highest-profile case, other disputes have percolated throughout the country and the courts, highlighting First Amendment nuances on social media. In Davison v. Randall, the Fourth Circuit Court of Appeals addressed whether a county board chair violated a constituent’s First Amendment rights by banning him from a Facebook page.^[77] In its analysis, the Fourth Circuit found that the interactive portion of a government official’s Facebook page constituted a public forum, and therefore, carried a prohibition against viewpoint discrimination.^[78]

The Fourth Circuit also acknowledged the constituent’s argument that even the government official’s decision to use Facebook as a forum violated his First Amendment rights.^[79] The constituent argued that Facebook hinders blocked parties from fully participating in public forums because Facebook allows private users to block other profiles, subsequently preventing the blocked profile from responding to the blocking profile’s replies to government accounts.^[80] Such control by third parties of the discussion under government posts, the argument continued, allowed third parties to exclude speech from the forum through a method that would be impermissible in a physical forum.^[81] The Fourth Circuit noted that this legal theory had not been addressed by the courts before, but that such third-party control could conceivably violate the First Amendment “if the private website included certain types of exclusionary rules.”^[82] Nonetheless, the Fourth Circuit did not decide that argument’s merits because the amended complaint in district court was dismissed for both prejudice and futility.^[83] The Fourth Circuit affirmed the denial of the motion for leave to amend the complaint on the basis of prejudice, but not futility.^[84]

III. First Amendment Tensions Created by the Courts

These above cases represent a general legal approach that focuses on accounts rather than individual posts or tweets. This approach is not compatible with the First Amendment. These holdings do not recognize that each tweet has a matrix of visibility settings from each user involved in publishing the tweet and the subsequent combination of each accounts’ settings, rights, and obligations. This layering creates inherent conflicts and allows gamesmanship by public officials to subvert the First Amendment. The results reveal the courts’ misunderstandings of online forums and warrant a rethinking of First Amendment analysis of governmental social media posts.

A. Inherent Contradictions: The Disintegration of Public Forum Integrity or Elimination of the Freedom of Association

In First Amendment analysis of Twitter posts, courts have assessed Twitter accounts as a whole, lumping every tweet and corresponding interactive section together, to discern speech and forum status.^[85] This analysis assumes that account owners can dictate who can reply to, retweet, and otherwise interact with every one of their tweets. But with this assumption, the courts miss a fundamental feature of Twitter that makes these rulings full of irresolvable conflicts: retweets.

One of Twitter’s main appeals is a user’s ability to interact with others. Indeed, the majority of tweets on Twitter are retweets.^[86] Roughly 80% of tweets from the most active tweeters are retweets and replies (49% and 33%, respectively).^[87] Government accounts on the site have the same capabilities as other Twitter accounts.^[88] Government accounts often retweet private accounts.^[89] The interactive spaces below government accounts’ tweets have been classified as public forums.^[90] Notably, courts have not differentiated between tweet types, such as when a government retweets or quote tweets a private user.^[91] Therefore, a government account retweeting a private account creates a public forum underneath that private tweet and the government retweet.

When a government account retweets a private account, the private account’s settings and block list control the tweet’s visibility.^[92] The government account’s retweet republishes the original tweet, with the original author’s settings retained on the retweet.^[93] The government account’s settings can augment private accounts’ settings, but the government account cannot strip away private accounts’ settings that limit a tweet’s visibility.^[94] In other words, the government, by retweeting a private account, adopts the private user’s settings and block list at that point—and any point in the future—with respect to the retweeted private account’s tweet.^[95]

Importantly, these private accounts’ actions are not restricted by the First Amendment.^[96] Unlike a government account, a private account’s blocking activity is not constrained by law, but rather is protected by the First Amendment’s freedom of association.^[97] Private accounts can block any other account on Twitter for any reason, at any time, for any duration.^[98] Private accounts can block a user—or many users—for speech they find disagreeable, therefore exercising their freedom to practice viewpoint discrimination.^[99] Private accounts can block another user simply because the user is an insufferable Oklahoma Sooners fan;^[100] because the user thought the Houston Astros were scapegoats; or, because the user consistently and loudly espoused specific beliefs.

Applying the courts’ holdings here illustrates the fundamental contradiction within them. Governments’ First Amendment obligations and private accounts’ freedom of association are both exercised in the same areas on Twitter: the interactive spaces under private tweets that governments adopt through retweeting.^[101] A reply to a government account’s retweet is a reply to that private account’s original tweet.^[102] When a private account blocks a user, the blocked user cannot see or reply to the tweet, including if a government account retweets that tweet.^[103] For blocked users, the block creates a situation where, from their view, an original tweet and government retweet do not exist.^[104]

Thus, either a private account retweeted by a government account may exercise its freedom of association by blocking a user for any reason, thereby excluding blocked users from a public forum; or, the government must infringe upon the private account’s freedom of association by requiring a retweeted private account to unblock other users so that the blocked users can interact in the created public forum.

For example, every Twitter user blocked by the Progressive Caucus,^[105] the Arizona Senate GOP,^[106] or RNC Research^[107] accounts cannot see or interact with retweets of those accounts from U.S. Representative Ilhan Omar or former Arizona Governor Doug Ducey. Those who have been blocked by firefighter unions^[108] and trucking interest groups^[109] cannot see or interact with retweets of those by Oklahoma City Mayor David Holt or Ducey. Oklahomans who have been blocked after online beef with the Oklahoma Pork Council cannot see or interact with Oklahoma Governor Kevin Stitt’s retweet of the trade group’s tweets.^[110]

These examples are illustrative, but they are not merely hypotheticals. Jonah Goldberg, editor in chief of The Dispatch,^[111] replied to a @RNCResearch tweet with one word, “Childish,”^[112] and the next day, Goldberg tweeted that he had been blocked by @RNCResearch with an accompanying screenshot.^[113] This interaction shows that the problem illustrated by these examples is not speculative. Many private accounts retweeted by public officials likely have participated—and, in the future will participate—in viewpoint discrimination by blocking other users based on the content of their tweets. Therefore, through retweets, these elected officials are inadvertently ceding their ability to ensure the First Amendment is upheld or, alternately, are calling the freedom to associate into question. Blocked individuals are excluded from public forums created every day without their knowledge.

The Fourth Circuit speculated that if the government chose a social media site that allowed only certain viewpoints to post and comment, a compelling argument could be made that the venue selection violated the First Amendment, despite the lack of government action when the restriction is being imposed by a private company.^[114] This, the court noted, would be no different from a municipality choosing a setting for a town hall meeting that refused entry based on a viewpoint.^[115]

As Twitter is currently designed, this hypothetical could already occur in countless online public forums. When a government account retweets a private account, the government account has no way to monitor the private account’s blocked user list.^[116] The private account’s blocked list is private and unpublished.^[117] The private account, retweeted by the government account, could block all accounts disagreeing with the tweet. The account could block all Democrats—or Republicans—from replying. The government account would be none the wiser that insidious viewpoint discrimination was taking place in the public forum created by its retweet.^[118] The government cannot, at any point in time, ensure viewpoint discrimination is not occurring for that specific tweet and forum unless the government official runs the private account being retweeted.^[119]

Twitter’s settings make government accounts retweeting private accounts a minefield of potential First Amendment violations. Additionally, good faith efforts by the government to ensure that everyone can see its retweets would not be successful or constitutional. Such efforts would likely involve asking the retweeted account whether the account has blocked any other accounts and why the user blocked them.^[120] The impracticable expectation of compliance, trust, and recall required of the private account owner is well outside the bounds of ensuring that viewpoint discrimination does not occur in a public forum.^[121] Aside from practical concerns, the request, once again, severely implicates the private account’s freedom to associate and the security that comes with that freedom.^[122]

Designation of public forums underneath a private account’s tweets would not only obliterate the freedom of association but would also vanquish the user’s online safety. Twitter users block others for many reasons, including harassment and abuse.^[123] Twitter is often the setting of insults,^[124] racist tweets,^[125] sexist remarks,^[126] and even death threats.^[127] If the First Amendment allows the government to infringe upon the freedom of association by requiring that private accounts open their replies to all who exercise speech within the forum’s bounds,^[128] the user would likely be exposed to many harmful, constitutionally protected statements.^[129] Users who would normally be able to block another account for any reason would be burdened by receiving replies that they would otherwise prefer to block. Public forums underneath every government tweet, and retweet, would be significantly detrimental to users’ online experiences^[130] and their health.^[131]

B. It’s a Feature: Exploiting the Gap Between the Law and Twitter’s Settings

Twitter’s settings allow public officials to exploit the gap between jurisprudence and reality to violate the First Amendment. Courts distinguish between a government account and a private account in First Amendment analysis.^[132] Public officials have accounts they use for private purposes.^[133] These officials may use private accounts to tweet about their families, political campaigns, or local sports teams.^[134] Public officials’ private accounts retain no obligations under the First Amendment and remain free to block anyone, for any reason.^[135] However, public officials also control their government accounts, where the First Amendment does impose obligations.^[136] Therefore, if government accounts desire to lessen the impacts of their dissenters on Twitter, they can utilize this gap to effectively do so.

Ostensibly, public officials use their personal accounts to disseminate content and aggregate followers that they can keep past their terms.^[137] For example, Texas Governor Greg Abbott’s government account, @GovAbbott, retweeted his “personal” account @GregAbbott_TX several times in January 2022. These retweets were about National Law Enforcement Day,^[138] suing the Biden administration to oppose vaccine mandates,^[139] Texas business growth,^[140] authorizing border patrol agents to shoot when shot upon,^[141] and the Occupational Safety and Health Administration (OSHA) dropping its vaccine mandate.^[142] The tweets strike a tone and encourage engagement that contrasts with his government account’s tweets and retweets of other accounts, which focus on grant funding,^[143] the Houston Ship Channel,^[144] human-trafficking awareness,^[145] and a week of prayer.^[146] Abbott’s government account also retweeted a Texas Department of Public Safety tweet about its efforts to secure the border.^[147] Tellingly, the Texas DPS tweet tagged Abbott’s personal account, despite being retweeted by Abbott’s government account.^[148]

Among these controversial tweets and retweets, Governor Greg Abbott tweeted from his “personal” account (@GregAbbott_TX) that he would be suing the Biden administration to fight President Biden’s mandate that Texas National Guard members be vaccinated against COVID-19.^[149] Abbott’s official government account, @GovAbbott, retweeted the personal account’s tweet.^[150] Like most private tweets retweeted by his government account, this tweet received a disproportionate amount of engagement compared to other @GovAbbott tweets.^[151]

As the law stands and with how Twitter works, Abbott could silence dissenters and critics of his decision to sue the President.^[152] Abbott could press two buttons on his phone to switch from his government account that retweeted his “private” tweet to his personal, “private” account. There, Abbott (staying below the occasional stray tweet standard to keep his personal account’s private status),^[153] could exercise his freedom of association as a private citizen to block anyone criticizing his decision. The private account’s block would prevent those dissenters from seeing any tweets or commenting on any future tweets from his private account. The block would also prevent the user—exercising constitutional speech within a public forum underneath the government account’s retweet—from interacting with any tweet from the personal account, @GregAbbott_TX, that has been retweeted by the government account, @GovAbbott. Abbott could then tweet occasional updates from his personal account, tap a few buttons, and retweet the tweet from his government account. Each successive private tweet and government retweet would be received with less and less criticism from those he wishes to silence. Each tweet and retweet would cull dissent. Each tweet and retweet would effectively exclude citizens from the public discourse and a public forum.^[154]

In this hypothetical scenario, the governor has successfully exploited the gap between Twitter and the law. He has stifled dissent, discussion, and protected speech because he tweeted the policy from his private account and retweeted the tweet from his government account. Abbott never broke current laws because he didn’t block anyone from the government account.^[155] The government account never acted to exclude others;^[156] the private account did, thus controlling which private individuals could participate in the public forum underneath his government account’s tweets.

C. Cheat Codes: Elected Representatives Relish Violating the First Amendment

Elected representatives act outside the law and within the gambit of this gap regularly. Many elected representatives unabashedly block their constituents. Alaska State Senator Lora Reinbold blocked a critic from her legislative, government Facebook page.^[157] U.S. Representative Thomas Massie blocked critics of a Christmas card he posted on his government Twitter account.^[158] Another U.S. Representative, Ken Calvert, blocked critics from his official Twitter account after he was criticized for a Washington Times op-ed he authored and posted on Twitter.^[159] U.S. Representative Alexandria Ocasio-Cortez blocked a Twitter user who criticized her for comparing migrant detention centers to concentration camps.^[160] Importantly, Rep. Ocasio-Cortez blocked the user from her “private” account, @AOC.^[161] After settling the suit and unblocking the critic, Rep. Ocasio-Cortez stated that she “reserve[s] the right to block users who engage in actual harassment or exploit my personal/campaign account, @AOC, for commercial or other improper purposes,” though that restriction is unconstitutional in public forums.^[162] However, because the suit was settled, a court did not determine whether the @AOC account was acting as a government account.^[163] Two suits from other individuals blocked by @AOC were pending at the time of the statement.^[164]

Aside from blocking, public officials use creative methods to exercise viewpoint discrimination against replies they dislike. These officials, sensing a lack in the courts’ precision, “hide” replies instead of blocking accounts.^[165] For example, Texas Attorney General Ken Paxton hides replies that mention his indictment or feature his booking photo,^[166] demand his resignation,^[167] and even GIFs that offer colorful reactions to his tweets.^[168]

Oklahoma City Mayor David Holt has hidden replies that question the coronavirus vaccine^[169] or Oklahoma City’s vaccination rate,^[170] or use profanity.^[171] Holt has also hidden replies that ask him to pick up his campaign yard signs.^[172] Ironically—and tellingly—these hidden replies were responding to a tweet Holt posted sharing an article Holt wrote himself extolling the virtues of pluralism.^[173]

Attorney General Paxton and Mayor Holt—and likely other elected representatives—take part in viewpoint and content-based discrimination that violate the First Amendment by hiding replies.^[174] These replies, ranging from GIFs to pleas concerning the blight of campaign materials, are the core of protected speech.^[175] Yet, politicians feel empowered to act egregiously, repressing government petitions because the courts have not specifically spoken on the issue of hidden replies. Given an inch—or left an inch of unspecified territory—politicians take a mile. They minimize valid dissent and criticism, manipulate discussion, and obliterate the purpose and integrity of public forums.^[176]

The courts’ holdings, which define public forums on Twitter, make no distinction between the types of tweets from government accounts nor do they make distinctions based on the individuals controlling the accounts.^[177] Because the law fails to specify these distinctions, violations of rights are sanctioned with every ruling. Savvy elected representatives continue to hone their abilities to take advantage of this contradiction that violates the rights of everyday citizens, and to intentionally induce greater violations. These First Amendment violations will continue unless the courts treat Twitter and the rest of social media with the modern lens they require, and public officials with the skepticism they have earned and deserve.

IV. Admin Privileges: How Courts Should Rule

Today, most Americans say most of their friends share their views.^[178] 98–99% of Americans live in “partisan bubbles,” where members of one party have “almost zero interactions” with members of another party in their neighborhood.^[179] In 2020, 58.2% of voters in the presidential election lived in a landslide county—a county where presidential candidate Donald Trump or Joe Biden won by 20% or more.^[180] Joe Biden won 85% of the nation’s counties that have a Whole Foods and only 32% of the counties that have a Cracker Barrel—the widest gap ever.^[181]

Sorting happens online as well. People are three times more likely to follow a Twitter account if the account matches their partisan persuasion.^[182] Twitter users derive greater enjoyment from content that endorses their opinion, cultivating echo chambers.^[183] This feedback effect is equivalent across the partisan divide: Republican and Democrat.^[184]

People who take part in ideologically homogenous groups tend to move to extreme points in agreement with their prior inclinations.^[185] Group polarization means, for example, that a group that is moderately against a war effort will, after conversation amongst themselves, strongly oppose the war effort.^[186] And in today’s America, these ideologically homogenous groups are common, whether the groups are represented by votes, Twitter followers, or Whole Foods locations.

Misinformation can also thrive through partisan ties. Misinformation travels “significantly farther, faster, deeper, and more broadly than the truth” on Twitter.^[187] False news stories are 70% more likely to be retweeted than true stories.^[188] None more prominent than Trump’s election lies, which were the subjects of some of his most popular tweets.^[189] In an analysis of 2020 election misinformation, researchers found that of the top 100 “repeat spreaders” of misleading content, all but two were supporters of then-President Trump.^[190]

Hopeful reactions by Twitter, such as labeling tweets as misinformation, do not seem to work. Messages labeled as misinformation by Twitter spread further and longer than those messages that did not have the misinformation label.^[191] This spread of misinformation cannot be attributed to bots, but instead, to real users retweeting inaccurate news items.^[192]

Online public forums offer an opportunity to break out past the sorting and partisanship that defines America. The marketplace of ideas is needed when lies are popular, potent, and partisan.^[193] Online public forums should be preserved as places for discussion about public ideals, calling out falsehoods, and holding elected officials accountable. However, the court holdings discussed above do not propagate these purposes. These holdings allow government officials to capitalize on the incongruity between the law and Twitter’s settings. The courts are laying the groundwork for politicians to cultivate dangerous echo chambers, tamp down dissent, and manipulate perception. To combat this problem, this Note argues that courts should conduct government speech and forum analysis, by employing a two-step approach to online speech on social media. The two-step approach supports public forum integrity, maps the proper forum analysis on tweets, negates gamesmanship by government officials, and elevates individual rights.

The proposed test’s first step is forum analysis and requires courts to look at all other accounts involved with the tweet in question from an account. This step would serve to label accounts as either private or government accounts.^[194] However, courts applying this test would determine not only the retweeting account’s status but also the original author’s account status. This step is like the courts’ current analysis but offers a more comprehensive look at accounts involved with the visibility of the government account’s posts, as opposed to one account’s general characteristics.^[195]

If a court finds that both accounts in question are government accounts, it should follow current forum analysis precedent to find that the interactive space under the government accounts’ tweets is a public forum.^[196] Both accounts are already prohibited from blocking other accounts and hiding replies.^[197] Retweets between these specific accounts should be permitted because both accounts have identical First Amendment obligations.^[198] There, the analysis should end. However, should the court find the retweeting account is a government account and the tweet’s author is a private account, further analysis should follow.

The second step of this analysis focuses on the individuals controlling the accounts. If the government official retweets an account that the official controls in a private capacity, the court should hold that the public official voluntarily created a public forum underneath the official’s own private account’s tweet.^[199] The court should then mandate that all First Amendment obligations from the government account be carried over to the official’s private account, requiring the private account to unblock any accounts previously blocked.^[200]

If a government account retweeted a private account that the public official does not control, the court should find that the public official has created an improper public forum.^[201] The court should then hold that any retweet by a government account of a private account not controlled by the government official contravenes the government’s First Amendment obligations and invades the private account’s freedom of association. Therefore, the court should hold that any government account retweet of a private account not controlled by a government official is unconstitutional.

The two-step approach solves many problems that appear in current holdings. One benefit is that the test fully supports the tradition of public forums. If a public forum exists, third-party control over a tweet’s visibility cannot be monitored by the government.^[202] Viewpoint discrimination or other forms of discrimination that are not allowed in public forums could run rampant online without any limitation.^[203] This test, clear-eyed about Twitter’s settings, eliminates selective retweets that cater public forum availability to partisans and restricts any government concession that impedes its ability to maintain a proper public forum.

While this test brings clarity and assurance to the power of public forums on Twitter, it also protects individual rights. A private user retweeted by a government official should not be burdened by a public forum created through a government account’s retweet. Further, a government account cannot ensure it is complying with First Amendment obligations without a substantial invasion of an individual’s rights.^[204] This solution values the rights of the private individual while also elevating the purpose of public forums online.

The test also negates any gamesmanship by elected representatives.^[205] The test does not hinge solely on one account’s classification.^[206] Instead, the test acknowledges that government officials can easily spread information between multiple accounts. The courts have approached their Twitter-forum analyses at the account level, but the analyses’ underlying purpose is to discover how politicians use their social media accounts.^[207] This test better serves that purpose. When the user behind a government account and personal account is the same person, retweeting a private account from a government account reveals the private tweet’s governmental purpose and the user’s choice. If the government official would like to spread a policy position for both governing and campaigning reasons, the official can tweet the information from the government account and retweet the tweet from the private account.

This two-step approach also protects government officials. Government officials can protect their rights by not engaging in government speech through their private accounts. There, while maintaining a private account that complies with current law, a government official can enjoy the benefits of social media and privacy settings on Twitter. A government official maintaining a private account can still block users, hide replies, and maintain a protected account.

If the government official wishes to enjoy the platform of the office and gather followers on a “private” account, this test does not stop them from doing so. The test merely prevents government officials from acting in a manner that contravenes First Amendment strictures. Public officials already must distinguish between government and private speech to determine whether they may block others.^[208] This test only seals the gap the courts have not covered and adds certainty for government users.

While this control on government accounts seems broad in description, in reality, the effect is likely marginal. Many “private” or “personal” accounts are likely government accounts that disseminate government speech to the detriment of officially labeled accounts.^[209] These “private” accounts should be a place for public conversation and likely are, according to standards set out by the courts.^[210] This test takes the ideals behind current jurisprudence and properly places them within Twitter’s context.

This proposed test’s main drawback is that it does not allow for government accounts to retweet private accounts not controlled by the government account holder. While many government accounts will not find this control appealing, there are many ways for government accounts to maintain the same capacity to spread information without violating the First Amendment.

V. GG: How Governments Can Respect the First Amendment and Current Law

While the courts have not yet spoken on many issues presented by the First Amendment and forum analysis in the modern era, government accounts can still take proactive steps to ensure their constituents enjoy the breadth of the First Amendment’s benefits. This can be done in several different ways, all of which disconnect the private speech’s content from its privacy settings and pass the proposed test’s scrutiny.

The first method a government account may use to replace retweeting a private account is posting a screenshot of the private account’s tweet. The screenshot separates the private account’s tweet—including the private account’s settings—from the government’s tweet. The government tweet is now a picture of another tweet that cannot be modified by the private account. The private account may or may not be tagged in the tweet or its replies.^[211] Regardless, the private account’s tweet is amplified, achieving a retweet’s goal. Additionally, the private account may block other accounts without affecting the government account’s tweet because the content of the private account’s tweet is preserved on the government’s account. The additional steps necessary to post the photo are not overly prohibitive, and only take a few seconds.^[212] Such a method does not overly burden government accounts.

The second method that government accounts could use is a “classic retweet.”^[213] Originally, the retweet was a user-created function as Twitter did not have a retweet button.^[214] Users manually copied a tweet’s text, typed “RT,” and the original author’s handle or “@,” and hit send.^[215] In response to the “RT” popularity, Twitter debuted the retweet button in 2009.^[216] While current retweets pose a difficult First Amendment analysis, classic retweets do not. While still called a retweet, this action is no different than any other tweet. The original tweet exists in its original form, and the “retweet” is copied content that is separately tweeted out by the “retweeter.” This mechanism allows a retweet unburdened by an authoring account’s settings because the “retweet” is just another government account tweet. The classic retweet creates two separate threads for discussion—one controlled by the tweet’s author and the other controlled by the government account—preserving individual rights and public forums. This method spreads information while preserving the resulting public forum and the private account’s right to block others.

The main downside to this approach is that the process is more complicated than pressing a button. This method also requires some skill from government officials, who may not have the requisite social media adeptness. Additionally, tagging a user in a tweet can bring a flood of unexpected mentions.^[217] These downsides can be mitigated though, because the private account can use Twitter’s settings to eliminate the flow of mentions and users to their private account.

Government accounts may use these techniques to preserve their constituents’ rights and the public forums they create. However, these workarounds do not adequately address other concerns that lie ahead with public forum designations on Twitter and social media.

VI. Future Glitches

While concerns about Section 230 flurry^[218] and social media companies ask for more regulation to address their core incompetencies,^[219] other proximate issues about current holdings loom.

A. Social Media Policies and Public Forum Doctrine

Current cases have held that the government may not choose a venue that practices viewpoint discrimination^[220] and may not practice viewpoint discrimination in public forums.^[221] With these restrictions on venue choice and viewpoint discrimination, current holdings are likely open to attack because Twitter and other social media sites’ moderation policies are fundamentally at odds with a public forum designation.

Twitter has its own freedom of association and the ability to moderate its content as it sees fit.^[222] This conflict between the First Amendment and viewpoint discrimination seemed to be solved when Elon Musk purchased Twitter.^[223] In a press release after the deal was announced, Musk said, “Free speech is the bedrock of a functioning democracy, and Twitter is the digital town square where matters vital to the future of humanity are debated.”^[224] Musk has characterized himself as “a free speech absolutist”^[225] and pledged to bring such ideals to Twitter’s moderation.^[226] However, his personal identification and pledge have been befuddling in application.

On December 1, 2022, after Musk claimed to be a free speech absolutist and promised to apply such principles to Twitter’s moderation, Ye (formerly known as Kanye West) tweeted an image of a swastika inside the Star of David.^[227] In a tweet explaining Ye’s resulting suspension, Musk said, “[H]e again violated our rule against incitement to violence. Account will be suspended.”^[228]

Assuming that Musk meant Ye’s tweet was “incitement to violence” within the First Amendment context, one would be woefully disappointed in Musk’s First Amendment knowledge. Ye’s tweet hardly fits within Brandenburg’s holding that speech is protected “except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action.”^[229] Further, Ye’s tweet is reminiscent of the Skokie case, where the Supreme Court reversed an injunction against “[m]arching, walking or parading in the uniform of the National Socialist Party of America; [m]arching, walking or parading or otherwise displaying the swastika on or off their person; [d]istributing pamphlets or displaying any materials which incite or promote hatred against persons of Jewish faith or ancestry or hatred against persons of any faith or ancestry, race or religion.”^[230]

If Musk has realized how expansive First Amendment protections are, he also has likely realized that moderating along the bounds of the First Amendment presents extreme and significant challenges to social media sites. Judge Keenan addressed this incongruity in her concurrence in Davison v. Randall.^[231] Social media companies have policies against unfavorable content, such as hate speech.^[232] These policies are enforced in several ways that conflict with public forum designations on these sites.^[233] Twitter, for example, treats replies that “detract from the conversation” differently from other replies.^[234] The site also prohibits users from using “hateful images or symbols” in their profile images or profile headers.^[235] Additionally, the site bars any references to protected classes that “spread fearful stereotypes about a protected category, including asserting that members of a protected category are more likely to take part in dangerous or illegal activities.”^[236] Twitter also prohibits slurs, tropes, and targeted misgendering of transgender individuals.^[237] Finally, Twitter is testing a new “Safety Mode” that “aims to reduce disruptive interactions.”^[238] The feature blocks accounts on other accounts’ behalf for “potentially harmful language” or for “sending repetitive and uninvited replies or mentions.”^[239]

Most of this speech is protected speech.^[240] This protected speech would not be excluded from a public forum, nor would it be treated differently from any other speech,^[241] yet Twitter’s rules dictate that these protected statements be barred from a public forum under a government account’s tweet. Even if the statements are not hateful, Twitter determines statements’ visibility based on its view of productive speech.^[242] Viewpoint and content-based discrimination like this is incompatible with public forums. These policies, as Judge Keenan noted, also allow public officials to exercise viewpoint and content-based discrimination covertly by reporting content protected by the First Amendment but barred by the sites.^[243] The officials can report the user, subjecting the content to the site’s determinations.^[244]

Twitter’s new “Safety Mode” also looms over political conversation on the site. A public forum does not prohibit someone from engaging in speech when they want to make an unexpected reply, nor does it prohibit a citizen from speaking in the public forum often.^[245] Such participation would normally be lauded as civic participation. Yet, Twitter’s new feature takes the opposite view and would likely exclude the most civic-minded users from the virtual public forum.^[246]

While social media sites’ policies likely pose future challenges for the public forums on the site, they also allow for third-party control of discussion to the point of First Amendment dubiousness.

B. Third-Party Control Exerted by Commenters Blocking Others

The plaintiff in Davison v. Randall argued that Facebook and other social media sites allow users participating in the designated public forums online to exclude others from responding to some speech in the forum.^[247] On Twitter, one account may reply to a government account’s tweet and also block another account that would like to participate in the forum. The blocked account’s participation could be a reply to the original reply, or another reply to the original tweet meant to dissuade others from agreeing with the original reply. Here, one account could block others and still participate in the public forum without the blocked account seeing the content. A physical analog to such a situation would be someone speaking in a designated public forum, but only allowing some people to hear the speech, and only allowing certain people to respond, which would clearly violate the First Amendment.^[248]

This situation also pits the freedom of association against the public forum obligations because one cannot unblock a person from one Twitter thread, but instead must unblock the blocked account from all aspects of their account.^[249] The tension here, and the actions taken by Twitter users that bleed over into the designated public forums in the replies to the government speech, show that a public forum determination is likely problematic.

C. The Veil Before Litigation

However, with the many different issues that arise from Twitter’s settings and government officials’ actions, litigation will likely be slow to define the rules of public forums on Twitter. Users are not notified when they are blocked by another on Twitter, and one would have to visit a profile or see a reference to a tweet to realize his account was blocked.^[250] The plethora of ways that content circulates on the website coupled with Twitter’s ever-changing settings, creates a moving, time-sensitive target that makes even the initiation of litigation challenging—not to mention attempts to bring any litigation to its logical conclusion and create rulings that may define and limit this behavior on the site.

VII. Conclusion

Current holdings applying the First Amendment to social media settings allow dissenters to be silenced and impose great limitations on public forums. As it stands, the law creates a conflict between public forum doctrine and an account’s freedom of association. This precedent also allows for gamesmanship by elected officials to avoid their First Amendment obligations.

The two-step test proposed by this Note solves many of the problems presented by current case law. First, the test ensures public forum integrity is upheld by minimizing the potential for insidious third-party control over speech in a public forum. The test also protects against an invasion of an individual’s freedom of association. Further, the test addresses public officials’ attempts to game the gap between current law and Twitter’s settings. The test encourages government actions that comport with the principles behind public forums. These resulting methods of communication achieve the goals of public forums and the purpose of government accounts while respecting all individual rights and freedoms.

While some may critique this test as missing the forest for the trees or being too dependent on one social media site’s settings, this test hopes to inform the development of First Amendment principles online. As courts have begun to see these cases more often, and with different inflections, they have not fully considered the ramifications of finding public forums online. This proposed test keeps the rights and liberties of every party in online forums front of mind. Additionally, the test can be utilized in any setting where multiple parties, including those with First Amendment obligations, can republish online content. This proposal helps to answer some of the questions about the First Amendment on Twitter and social media; however, future problems arising from Twitter and other social media networks’ changing views of moderation pose great challenges to a body of online, First Amendment law based on physical settings.

Matthew W. Welsh