Margot Kaminski

Flash mobs

The past ten years have shown that the Internet is a highly useful tool for organising crowds. In the early 2000s, the phenomenon of the ‘flash mob’ emerged, when improvisational comedy groups started using public space for whimsical performative group gatherings.¹ The creation of flash mobs as they now exist, using social media to gather strangers for such an event, has often been credited to Harper’s senior editor Bill Wasik.²

The organiser of a flash mob would email, blog or use Twitter (a ‘tweet’) to send instructions to a specific group of people or the world at large, describing when and where to meet and what activity the group should do. A ‘flash mob’ has several features: the announcer usually does not know the full membership or size of the eventual crowd; the crowd is told the time, place and sometimes the purpose of the gathering; the crowd acts on the announcement with no apparent incentive, operating within parameters set by the announcer.³

At first, the flash mob served as comedic street performance, with groups meeting to shop en masse in a department store for a ‘love rug’, perch on a stone ledge in Central Park making bird noises, dance silently in public areas or freeze in place in crowded Grand Central Station.⁴ Flash mobs have become so well recognised as performance art that mainstream advertisements now reference them.⁵

Unsurprisingly, groups started using flash mobs for political organisation.⁶ These political protests are sometimes called ‘smart mobs’, a term coined by Howard Rheingold in 2002.⁷ Howard Dean’s presidential campaign proposed a flash mob in Seattle, and protesters staged a flash mob against Russian Prime Minister Vladimir Putin in St Petersburg.⁸ In May 2006, a blogger used the software LiveJournal to propose a flash mob of ice cream eaters in Oktyabrskaya Square in Minsk, the capital of Belarus.⁹ Police showed up and arrested ice cream eaters, and images of the arrests were circulated around the world.¹⁰

Flash mobs, more recently, were used by dissidents during the Arab Spring. In 2009, thousands of Iranians organised via Twitter to protest the elections. On 25 January 2010, more than 50 000 people gathered in Tahrir Square in Egypt, many informed through social media tools, such as Twitter or Facebook.¹¹

Protest movements have taken place via flash mob in Europe and the United States, as well. In November 2011, members of the Occupy Rome movement expressed solidarity with Egyptians in Tahrir Square with a flash mob in front of the Egyptian Embassy in Rome, Italy.¹² Also, in August 2011, a flash mob gathered in Denver to dance to Lady Gaga’s song Born This Way as a protest for marriage equality in the United States.¹³

As an organisational tool, the flash mob has notable benefits for would-be revolutionaries. The organiser is able to appeal to a wide audience, increasing the size of potential gatherings and protecting individual protestors with sheer numbers. The organiser usually does not know the scope of the audience, or their identities. So, if the organiser is detained, the gathering may still occur without them, and no membership list will be turned over to the authorities. Even so, authorities can often track where, and when, flash mobs will occur and arrest people. But sometimes, that’s the point: the ice cream eaters of Belarus planned on circulating publicity about the inevitable arrests so as to show the overreach of the totalitarian state.¹⁴

The same tools that are useful for organising spontaneous performance art or starting a revolution are also useful for committing group crimes, and it can be hard for authorities to distinguish between the two. In cities across the United States, and during the UK riots in the summer of 2011,¹⁵ social media has been used to organise flash mobs for more nefarious purposes.

‘Flash robs’, as they have been labelled, use flash mob tactics to overwhelm local police forces or shop owners by announcing the time and place of a potential robbery. In early 2010, flash mobs of young people turned violent in Philadelphia.¹⁶ Other flash mobs resulting in arrests or injuries were reported the same year in Boston, South Orange, NJ, and Brooklyn.¹⁷ In June 2011, 40 young men swarmed a Sears store in Philadelphia and stole thousands of dollars’ worth of merchandise.¹⁸ In July 2011, a flash mob robbed a Victoria’s Secret store in Washington, DC.¹⁹ Similar robberies or group violence have also occurred in Maryland,²⁰ Illinois,²¹ Ohio,²²Nevada,²³ Minnesota²⁴ and Canada.²⁵ Often the events are captured on surveillance video, and it sometimes appears that participants are posing for the camera.²⁶

Frequently, coverage of these events assumes that they were started through social media. There has, however, been some pushback on this understanding. District of Columbia (DC) police have said that they have no evidence the group robberies in the DC area were organised through social media.²⁷ It also appears that some of the violence in Ohio was not started as a flash mob, though it may have been exacerbated by social media once the violence started.²⁸ Also, an investigation into one of the Maryland group robberies the media had termed a ‘flash rob’ found that the thieves had hatched their plan on a bus, not through social media.²⁹

Whether or not this group violence is the result of social media, after a year of flash mob-associated violence, the term ‘flash mob’ itself has taken on a different connotation. ‘Mob’ is no longer a tongue-in-cheek term. It now means something closer to its original definition: ‘a large or disorderly crowd; especially: one bent on riotous or destructive action’.³⁰

Several governments have spoken of, or experimented with, shutting down, monitoring or regulating social media to prevent the harm that can result from flash mobs. In the UK, British Prime Minister David Cameron proposed limiting the use of social media.³¹Two men were jailed in north-west England for inciting disorder through social networking sites.³² In San Francisco, police blocked cell phone service in the public transportation system, the BART, to prevent suspected protests.³³ In Philadelphia, the FBI has monitored social-networking sites for flash mob activity.³⁴ Philadelphia has also imposed 9 pm curfews on young people to prevent gatherings.³⁵

In Cleveland, City Council members proposed an emergency ordinance, Section 605.091, ‘prohibiting the improper use of social media to induce persons to commit a criminal offense’. The Cleveland Mayor vetoed the initial proposed ordinance and it did not become law.³⁶ In December 2011, however, Cleveland adopted a revised version of the ordinance that prohibited acts that incite riots, listing ‘electronic media devices’ as criminal tools.³⁷ The Mayor did not sign the revised ordinances, purportedly because they ‘mirror state laws already in place’.³⁸ This new Section 605.011 bans inciting to riot, and adds computers and cellular telephones to a list of items that can be considered criminal tools when used illegally.³⁹

If flash mobs and flash robs continue to occur, governments are likely to formalise their attempts to combat them. The question in the United States is: how much protection does the First Amendment afford?

The answer is complicated. From a US perspective, the organisation of flash mobs represents a classic First Amendment problem: speech entwined with action. More specifically, the problem is what to do about speech made by an individual that potentially inspires a large group to act. The Supreme Court struggled for much of the last century with how to address this dynamic, finally lighting upon the current doctrine of incitement in Brandenburg v. Ohio in 1969.⁴⁰

The rise of flash mob activity in the past decade, however, will likely revive this debate in US courts, since new media appears (at first glance) to raise new problems. Attempts to penalise the organisation of flash mobs also reveal the underexplored tension between Brandenburg’s protection of speech to large groups, and criminal penalties for speech-based activity, such as conspiracy.

The history of incitement cases in the United States

The rise of flash mobs and flash robs has restarted discussion of the dichotomous nature of group organisation. On the one hand, flash mobs have the potential to be used for valuable political speech and political organisation, which many believe are at the core of the First Amendment.⁴¹ On the other hand, flash robs can cause great damage, and governments are perhaps afraid of the power of uncontrolled group action.

The regulation of individual speech commanding group action has had a tortuous history in the United States. Over the twentieth century, the Supreme Court shifted from permitting the regulation of motivating speech as dangerous, to protecting it as fundamental to First Amendment values. This struggle reflects the paradox of verbal efficacy in this country: we value political speech highly, but the more effective that the speech is, the more dangerous it is and, thus, the more likely it is to be regulated.

The Supreme Court’s approach to speech that advocates illegal action changed vastly over the course of the twentieth century, largely in response to the political backdrop of the times. The two central struggles have been over how to distinguish between abstract doctrine and incitement, and whether the advocated action must be imminent for the speech to be permissibly regulated.

The earliest cases involved the federal Espionage Act of 1917⁴² and the Sedition Act of 1918,⁴³ and arose among widespread fear in the United States of the rise of Bolshevism in Russia. In 1919, the Supreme Court addressed advocacy of (or incitement to) illegal action for the first time in three cases: Schenck v. United States,⁴⁴ Frohwerk v. United States⁴⁵ and Debs v. United States.⁴⁶ In Schenck, a unanimous Court upheld the conviction of protestors who had distributed leaflets urging resistance to the draft, finding that the First Amendment does not protect words that are ‘used in such circumstances and are of such a nature as to create a clear and present danger that they will bring about the substantive degree of evils the Congress has a right to prevent’.⁴⁷ The Court, famously, reasoned that the First Amendment ‘would not protect a man in falsely shouting fire in a theater and causing a panic’.⁴⁸

However, in Frohwerk and Debs, the Court looked not to ‘clear and present danger’, which requires a real immediate threat of danger, but to the speech’s ‘natural tendency and reasonably probable effects’.⁴⁹ This lower-threshold ‘bad tendency’ test eliminates consideration of how imminent the illegal action is, and can easily stretch to penalise advocacy of doctrine rather than action.

The Court again applied the bad tendency test to uphold convictions in Abrams v. United States⁵⁰ and in Gitlow v. New York.⁵¹ In 1927, the Court in Whitney v. California used the bad tendency test to affirm the conviction of a member of the Communist Labor Party who had signed a resolution recognising the liberation of workers through organisation.⁵²

Although the majority of the Court promulgated the bad tendency test through the 1920s, as early as 1919 a split began to form. Justice Holmes dissented in Abrams in 1919, insisting on a return to scrutinising a threat’s immediacy under the clear and present danger test. Holmes wrote that speech should be protected unless it ‘so imminently threaten[s] immediate interference with the lawful and pressing purpose of the law that an immediate check is required to save the country’.⁵³ Justice Holmes joined Justice Brandeis in a concurrence in 1927 in Whitney, concluding that speech advocating illegal action should not be regulated if it ‘falls short of incitement and there is nothing to indicate that the advocacy would be immediately acted upon’.⁵⁴

In the 1930s and 1940s, it appeared that the Court might shift to the clear and present danger test.⁵⁵ But, in 1951, the Court back-pedalled in Dennis v. United States.⁵⁶The Court considered in Dennis ‘whether the gravity of the “evil,” discounted by its improbability, justifies such invasion of free speech as is necessary to avoid the danger’.⁵⁷

The Dennis test does not mention imminence of the action, and the more ‘evil’ the speech is, the less the court needs to consider the probability of its actual occurrence. Justice Douglas dissented, emphasising the difference between abstract doctrine and incitement, and reiterating the importance of imminence.⁵⁸

The Court took up the distinction between doctrine and incitement in Yates v. United States,⁵⁹ where it clarified that Dennis had not ‘obliterated the traditional dividing line between advocacy of abstract doctrine and advocacy of action’.⁶⁰ However, the Court did not discuss whether imminence was required in Yates because it did not need to address the question.⁶¹

Several cases in the 1960s prefigured what is now the established doctrine on incitement.⁶² Then, in 1969, the Court held in Brandenburg v. Ohio that government may not restrict advocacy of illegal action unless it ‘is directed to inciting or producing imminent lawless action and is likely to incite or produce such action’.⁶³ Brandenburg is still the governing law.

Brandenburg addressed the conviction of a Ku Klux Klan leader under the Ohio Criminal Syndicalism statute, which made illegal ‘advocat[ing] the duty, necessity, or propriety of crime, sabotage, violence, or unlawful methods of terrorism as a means of accomplishing industrial or political reform’ and ‘voluntarily assembling] with any society, group, or assemblage of persons formed to teach or advocate the doctrines of criminal syndicalism’.⁶⁴

The accused man had spoken at a rural Ohio Ku Klux Klan rally to 12 hooded figures, some of whom carried firearms. A Cincinnati television station covered the rally and broadcast portions of the speech that night. During the speech, the organiser stated that ‘if our President, our Congress, our Supreme Court, continues to suppress the white, Caucasian race, it’s possible that there might have to be some revengeance taken’.⁶⁵ He also announced a Fourth-of-July march on Congress.

The Court held that such speech was protected by the First Amendment, and explained that the statute ‘purports to punish mere advocacy and to forbid, on pain of criminal punishment, assembly with others merely to advocate the described type of action’.⁶⁶

At the core of Brandenburg is the recognition that the conversion of speech into action requires a meeting of minds between speaker and listener. To be held responsible for the actions of those listening, the speaker must, in fact, intend to cause the action: the speech must be ‘directed’ by the speaker to inciting action.⁶⁷ Brandenburg incorporates the listener’s point of view too, by looking to whether the speech itself is likely to produce unlawful action. In other words, the listener must be able to clearly understand and follow through on the speaker’s intent. Then, Brandenburg dictates that the state can regulate this meeting of minds only when the possible action is actually likely and imminent. In other words, there must be real danger, and physical closeness, between the speaker and the possibility of action. Even if a speaker intends to incite illegal action, and the listener understands the message, if the action is prevented because of circumstances, or is preventable because it is meant to take place at a more remote time, the state may not regulate the inciting speech.

At the Supreme Court level, it appears that the analysis of incitement ends with Brandenburg, and one case, Hess v. Indiana,⁶⁸ affirming it. In Hess, the appellant was convicted in Indiana under the State’s disorderly conduct statute for shouting ‘[w]e’ll take the fucking street later’ or ‘[w]e’ll take the fucking street again’ at an antiwar demonstration.⁶⁹The Court emphasised the importance of the recipient side of the Brandenburg evaluation, and held that since Hess’s statement was not directed to any person, or group of persons, it could not be incitement and, therefore, was protected under the First Amendment.⁷⁰ The Court also emphasised imminence and likelihood, and the importance of the speaker’s intent to incite.⁷¹

Thus, if flash mobs are to be regulated as incitement, the regulation must abide by the Brandenburg test. The speech being regulated must be intended to produce imminent illegal action; the speech must be likely to produce such action; the action must be imminent; and the action must be likely to occur. Any statute regulating flash-mob organising as incitement to illegal action would have to address the speaker’s intent, how likely it is that an audience would understand the command, the imminence of the action and the likelihood of the occurrence of illegal activity.

This appears to be a high threshold to pass, as long as Brandenburg remains good law. But the inquiry does not end there. Some courts of appeals have exhibited discomfort with Brandenburg’s strictness, suggesting a tendency to limit its holding. This discomfort reflects an unresolved tension between the Brandenburg test and how speech is treated in criminal law, especially in aiding and abetting, solicitation and conspiracy.

How lower courts have dealt with Brandenburg

Since Brandenburg, multiple cases have come before lower courts addressing speech that could be analysed as incitement under Brandenburg. In a number of these cases, courts have avoided the Brandenburg standard.⁷²

Some courts have sidestepped Brandenburg’s incitement test, instead placing the speech into other categories that have lower thresholds to regulation. The Ninth Circuit, for example, chose to analyse online speech as a ‘true threat’ instead of incitement. Sitting en banc in Planned Parenthood v. American Coalition,⁷³ the Ninth Circuit held that the public disclosure of the names and addresses of abortion providers on websites constituted true threats and, thus, were not protected speech. The Court distinguished Brandenburg by explaining that ‘while advocating violence is protected, threatening a person with violence is not’.⁷⁴ Due to the fact that other abortion providers listed on posters had been killed after the posters had been released, putting a new individual’s name on the ‘wanted’ posters constituted a true threat.⁷⁵

As the Ninth Circuit chose to analyse these postings as true threats, rather than incitement, it did not have to address Brandenburg’s imminence or likelihood requirements. The dissent observed that the postings were merely intimidation, not true threats, because the speaker could not control the outcome of the threat.⁷⁶

Other courts have avoided Brandenburg by finding an underlying crime, or analysing the speech as commercial speech, which has a lower threshold to regulation. For instance, the Third Circuit in United States v. Bell⁷⁷ held that it was permissible for the Court to bar a person from promoting and selling unlawful tax advice. The Third Circuit refused to analyse the unlawful tax advice as incitement to illegal activity, stating that the injunction ‘should be grounded on aiding and abetting violations of the tax laws and on false commercial speech rather than on incitement of illegal activity’.⁷⁸ Similarly, the Eighth Circuit found in United States v. Holecek that the actual preparation of tax returns by the appellant took his speech, which had advocated violation of tax laws, outside of Brandenburg’s domain and into the actual commission of a crime.⁷⁹

Courts of appeals have also recognised that, in spite of Brandenburg’s holding, speech can be prosecuted as aiding and abetting a crime, rather than inciting it. In Rice v. Paladin,⁸⁰ the Fourth Circuit held that the First Amendment did not bar a civil lawsuit against the publisher of a book of instructions on how to be a hit man. The publisher, Paladin, stipulated that it intended to attract and assist criminals, that the particular murderer followed instructions from the book, that it intended and had knowledge that the book would be used by criminals to plan and execute murder-for-hire, and that in publishing and selling the book it had assisted the particular murderer in these particular murders.⁸¹The Fourth Circuit recognised that speech that constitutes criminal aiding and abetting does not enjoy the protection of the First Amendment, and civil aiding and abetting is also not protected.⁸² The Court reasoned with respect to Brandenburg that ‘speech which, in its effect, is tantamount to legitimately proscribable nonexpressive conduct may itself be legitimately proscribed, punished, or regulated incidentally to the constitutional enforcement of generally applicable statutes’.⁸³ Thus, the ‘provision of instructions that aid and abet another in the commission of a criminal offense is unprotected by the First Amendment’.⁸⁴ As such, the Fourth Circuit did not look to imminence, or to likelihood of action.

Conspiracy provides another loophole through which courts have escaped resorting to incitement analysis. Even though it is not in front of a court of appeals, United States v. Stone⁸⁵ is of interest because of its treatment of conspiracy. In Stone, the defendant was found to have conspired to use force, as distinct from having conspired to advocate the use of force. Stone distinguished between ‘mere advocacy or conspiracy to advocate, which is subject to the Brandenburg test, and conspiracy to engage in otherwise unlawful, unprotected conduct, which is not subject to the Brandenburg test’.⁸⁶ Thus, the district court found that Brandenburg was not applicable. It concluded that because ‘[t]he gist of the crime of conspiracy is agreement to violate the law’, it is ‘both possible and permissible to charge that criminal statutes were violated entirely by means of speech’.⁸⁷ The court noted that ‘[n]umerous crimes under the federal criminal code are, or can be, committed by speech alone’.⁸⁸

Stone builds on a line of cases in the Second Circuit concerning conspiracy and Brandenburg. The Second Circuit, in Rahman, explained that ‘[w]ords of this nature – ones that instruct, solicit, or persuade others to commit crimes of violence – violate the law and may be properly prosecuted’.⁸⁹ Furthermore, in United States v. Rowlee, the Second Circuit observed that conspiracy is not subject to a First Amendment defence, because the crime being punished is the act of conspiracy, which does not implicate speech.⁹⁰

Conspiracy is a particularly fascinating paradox because it is a crime that intimately involves speech as part of the crime: you must show agreement. Speech is not merely evidence of the crime; it is part of the crime itself.

Criminal law: crimes and speech

These cases bring us to the post-Brandenburg paradox: while government regulation of incitement to lawless action must meet the Brandenburg standard, criminal law still penalises all sorts of speech that closely resemble incitement. Despite Brandenburg, then, states may be able to prosecute the organisers of flash mobs for aiding and abetting, or conspiring to violate the law, or soliciting illegal activity.

It is worth clarifying that, in the United States, state and local governments can regulate the presence of physical crowds⁹¹ and any crimes committed within them. Crimes committed during a flash mob can be prosecuted against the individuals who commit them, and those who aid and abet those crimes. But, regulating the speech that causes the crowds to occur in the first place is a different matter, and conspiracy appears to reach substantially into Brandenburg’s territory and conflict with its holding.

Some crimes intimately involve speech. As discussed above, speech that instructs on how to commit a crime may be punishable as aiding and abetting.⁹² Additionally, the crimes of solicitation and conspiracy involve speech. For instance, a person is guilty of solicitation under the Model Penal Code’s definition if that person, with the purpose of promoting or facilitating the commission of a substantive offence, ‘commands, encourages or requests another person to engage in conduct that would constitute the crime’.⁹³ Likewise, under the Model Penal Code, a person is guilty of conspiracy if they agree to commit an offence, attempt to commit an offence, solicit another to commit an offence, or aid another person in the planning or commission of the offence.⁹⁴ Thus, speech that constitutes an agreement to violate the law may be punishable as conspiracy, and speech that invites participation in an illegal activity may be punishable as solicitation.

Under Brandenburg, governments should not be able to specifically penalise the use of social media to incite illegal activity unless that activity is both imminent and likely. Despite Brandenburg, however, prosecutors may be able to prosecute the organisers of flash mobs for aiding and abetting a crime, conspiring to commit a crime or soliciting a crime. When a flash mob organiser gives clear directions of the time and place of a robbery, they might be seen to be aiding and abetting, or soliciting, that robbery. More disturbingly, when a flash mob organiser asks a group merely to gather, with no mention of any contemplated crime, prosecutors may still attempt to bring conspiracy charges against the organiser, and that is quite problematic.

This apparent paradox rests on two foundations: firstly, there is no First Amendment problem with the ‘evidentiary use of speech to establish elements of crime or prove intent’;⁹⁵ secondly, speech is an integral part of the core act, or actus reus, of crimes such as conspiracy and solicitation.

The Supreme Court has acknowledged both of these concepts several times since Brandenburg. In 1993, in Wisconsin v. Mitchell, the Court held that a racial bias may be taken into account during sentence enhancement, even though that bias is provable only through speech.⁹⁶ The Court observed that the First Amendment ‘does not prohibit the evidentiary use of speech to establish the elements of a crime or to prove motive or intent’.⁹⁷ In 1982, in Brown v. Hartledge,⁹⁸ the Supreme Court recognised that the state may ban ‘agreements to engage in illegal conduct’, and the ‘fact that such an agreement necessarily takes the form of words does not confer upon it, or upon the underlying conduct, the constitutional immunities that the First Amendment extends to speech’.⁹⁹ Similarly, ‘solicitation to enter into an agreement… remains in essence an invitation to engage in an illegal exchange… and may properly be prohibited.’¹⁰⁰ The Court cited to a number of its previous decisions about commercial speech, where it had explained that the government may ban commercial speech related to illegal activity.¹⁰¹

The Court went on, in later cases, to make it clear that speech related to illegal activity may be regulated more generally. In New York v. Ferber, concerning child pornography, the Court observed that freedom of expression does not extend ‘its immunity to speech or writing used as an integral part of conduct in violation of a valid criminal statute’.¹⁰²

The Supreme Court’s acknowledgement that some speech is coextensive with crime need not lead to a direct conflict with Brandenburg. Aiding and abetting through speech, while not yet addressed by the Supreme Court, appears to be restricted to giving highly specific instructions about how to commit a crime.¹⁰³ Solicitation of a crime also appears to require a high degree of specificity, as it is based on a deliberate request that a specific crime be committed.

Conspiracy, however, may be perilously broad.¹⁰⁴ Unlike Brandenburg’s queries about both the speaker’s intent and the audience’s understanding, and imminence and likelihood that the event will occur, the Model Penal Code version of conspiracy requires only that the person being prosecuted must have agreed to commit the crime, with no showing that anybody else has agreed to commit the crime with them, and no showing that the crime was likely to occur.¹⁰⁵ This unilateral approach is followed by many states.¹⁰⁶ The Model Penal Code also makes proof of an overt act unnecessary where the offence is sufficiently grave.¹⁰⁷ A person tweeting ‘let’s meet in Tahrir square to cause trouble for the cops’ might be prosecuted for conspiracy to riot in states that follow the Model Penal Code, even if no other person has agreed to commit a crime, and the crowd never formed – and was never likely to form. All that would need to be shown was that the flash mob organiser agreed to riot.¹⁰⁸ It is worth noting, however, that many states now require an overt act in furtherance of the conspiracy, and some require a substantial step in furtherance of the conspiracy.¹⁰⁹

Common law conspiracy, or the default conspiracy when a state does not have a conspiracy statute, requires proof of bilateral agreement: agreement between two parties.¹¹⁰ At first glance, this appears to be more favourable for the defendant, because a prosecutor must show that at least one other person agreed to commit the crime. A conspiracy at common law, however, need not be based on an express agreement, and a conspirator may be guilty even if they do not agree to every part of the substantive offence.¹¹¹ If a person tweets ‘let’s meet in Tahrir square to cause trouble for the cops’, and somebody in fact shows up, that might be evidence of a conspiracy because showing up expresses agreement. Common law conspiracy may involve conspiracy to commit an act that is not itself illegal.¹¹² Common law conspiracy also requires only an agreement, with no overt act in furtherance of it.

As with the Supreme Court’s earlier struggle with incitement, some of the expansiveness of the crime of conspiracy comes from fear of the nature of group activity. An early Kentucky case recognised that some actions that would not otherwise be criminal may be aggravated by the nature of group action. When done by individuals, these acts ‘had only the essence of crime, but by means of the conspiracy, an unfair and mischievous advantage of the aggressors is recognised’.¹¹³ The Kentucky court cited, as an example, the ‘conspiracy to use violence, as a riot, which derives its indictability from the plurality of persons concerned’.¹¹⁴ Government fear of group action, then, may have been muted by Brandenburg, but may continue to persist in the breadth of possible conspiracy charges.

Difficulties for Brandenburg posed by new media

Because flash mobs may appear highly threatening, there is a danger that courts will, increasingly, sidestep Brandenburg and analyse flash mob organisation as conspiracy. In part, this is because the new technology of social media, at first glance, appears to confront courts with new problems. With social media, people can contact larger groups than ever before, and do so instantaneously. Conversely, flash mobs may appear to be outside of Brandenburg’s scope because they can be organised far in advance. Brandenburg essentially limits regulation of incitement to control of the mob leader who is physically present and wielding the axe.

None of these features, however, should change the analysis under Brandenburg. Imminence can still be a requirement for regulation of incitement: if a flash mob is organised a week in advance, that would give police time to prepare for it, just as organising a riot through pamphlets in the past gave police adequate time to prepare. Brandenburg’s imminence requirement, thus, is still reasonable; it allows regulation of speech only when there is danger that illegal action is so imminent that law enforcement cannot otherwise counter the threat.

Likelihood of action should also remain part of the equation. Many calls for flash mobs are not likely to be followed, but if the organiser appeals to a group known to commit crimes via flash mob in the past, then the speech may exhibit a greater likelihood of incitement to illegal activity.

Another feature of the new media that counsels for preserving Brandenburg is that online communication is not ephemeral. The state can go back through records of communication and see how the organisation occurred.¹¹⁵ It is easier to trace who in fact is aiding crimes, so the state should not overreach to get those who merely organise crowds.

In many ways, courts will be encountering the same problem with flash mobs as they once did with union organisers: the tension between conspiracy and incitement. The new media may have changed many things, but they likely have not changed the analysis under Brandenburg.

Other possible First Amendment protections

Despite the fuzziness of the interaction between Brandenburg and criminal law, courts should be reluctant to enforce regulation of flash mob activity. Flash mobs implicate not just the speaker’s individual right to free speech, but also the association rights of both the speaker and the eventual gatherers. Freedom of association protects the ‘right to associate for the purpose of engaging in those activities protected by the First Amendment – speech, assembly, petition for the redress of grievances, and the exercise of religion’.¹¹⁶ The Supreme Court has recognised that the individual right to a protected activity creates a ‘correlative freedom to engage in group efforts towards those ends’.¹¹⁷

It is noteworthy that freedom of association has been used to shield members of a group from prosecution for activities done by other members. The Court recognised in NAACP v. Claiborne that states cannot impose liability on non-violent protesters merely because of their association with those engaged in violence.¹¹⁸ ‘For liability to be imposed by reason of association alone, it is necessary to establish that the group itself possessed unlawful goals, and the individual held a specific intent to further those illegal aims.’¹¹⁹

In this regard, flash mobs should be governed by NAACP v. Claiborne. If this proves to be the case, individuals who do not actually commit the crimes should not be held liable for them unless the group itself can be shown to possess unlawful goals, and the individual can be shown to wish to further those aims. This standard would apply to the organiser of the mob as much as the individual appearing in response to the organisational call. Essentially, the organiser should not be held liable for crimes committed during the flash mob unless the group itself possessed unlawful goals, and the organiser, as an individual, intended to further those aims. More work is necessary to determine how this standard interacts with current state anti-riot laws.

Conclusion

The ‘right of peaceable assembly is… equally fundamental’ as the right of freedom of speech.¹²⁰ The potential of inspirational speech to be beneficial or harmful is far from a new dilemma in the United States. As governments contemplate how to handle the phenomenon of flash mobs, courts should keep in mind the fundamental expressive value of group organisation recognised by the Supreme Court.

Mobs, after all, are not just disorderly crowds. The word ‘mob’ can also refer to ‘the lower classes of a community’.¹²¹ Controlling crowd behaviour, on the one hand, may protect the body politic from violence produced by mob mentality; but, on the other hand, it may control which socioeconomic classes are permitted to participate in speaking and organising. That result is fundamentally anti-democratic in nature.

Thus far, flash mobs have not been handled with these principles in mind. One flash mob organiser pleaded guilty to misdemeanour shoplifting and conspiracy charges.¹²² A Maryland Delegate plans to introduce a bill charging each member of a flash rob with stealing the total value of all merchandise stolen by the group.¹²³Rather than analysing flash mob organisation as conspiracy, or holding individuals accountable for group action, courts should apply both Brandenburg and Claiborne, and look to:the intent of the speaker;

Adding the intent of the group itself bolsters Brandenburg so that one rogue individual does not create a problem for the original organiser.

In all likelihood, however, governments will overregulate flash mob activity as a result of community fear and advocacy by storeowners. Flash mobs may be the vehicle for bringing the conflict between incitement and conspiracy before the courts. We will have to hope that the democratic capacity of new media is not squelched in the name of public order.