Falsehoods, Foreign Interference, and Compelled Speech in Singapore

Abstract

Online misinformation endangers the infrastructure of fact essential to public discourse and presents an even greater threat where it is being utilised as a weapon by hostile state actors. In recognition of these dangers, Singapore has implemented legal measures to combat online misinformation, enacting in quick succession the Protection from Online Falsehoods and Manipulation Act (POFMA) and the Foreign Interference (Countermeasures) Act (FICA). These statutes open up novel frontiers of development for Singapore's free speech jurisprudence. Indeed, these statutes confer upon government authorities the power to compel the authors of certain material to display notices stating that the material contains falsehoods or originated from a hostile information campaign. Yet, should one accept that the constitutional right to freedom of speech extends to the freedom not to speak, the compulsion of such expressions may well be unconstitutional under Singapore's free speech guarantee. This article will study the theoretical justifications for a prohibition against compelled speech to evaluate whether Singapore free speech jurisprudence ought to recognise such a prohibition, propose a doctrinal framework to analyse compelled expressions by reference to US, UK, and Canadian jurisprudence, and critically assess how the POFMA and FICA would fare under such a doctrine.

Online Misinformation and Compelled Speech

Countries around the world have increasingly come to realise the magnitude of the problem of fake news proliferating online. Online fake news can be generated and spread with great speed, resulting in an unprecedented ability for fake news purveyors to effortlessly reach large audiences. Unchecked, the proliferation of online fake news presents an existential conundrum for constitutional democracy. It endangers the shared epistemic foundation at the core of constitutional democracy. By preying on groupthink and drawing upon emotive rhetoric, online fake news poses a threat to the liberal democratic ideal that reasoned public deliberation is the preferred way to navigate, and ideally resolve, disagreements within a political community.¹ The problem of online misinformation is exacerbated when one notes that online misinformation can be used as a weapon by hostile state actors. Indeed, credible reports have emerged of state actors waging hostile information campaigns to erode public confidence in a target country's government, or to sway public elections in a direction favourable to the attacking state.²

Fully cognisant of these dangers, the Singapore government has swiftly enacted legislation in response – the Protection from Online Falsehoods and Manipulation Act ³ (POFMA) and the Foreign Interference (Countermeasures) Act ⁴ (FICA). These statutes confer upon government authorities a diverse suite of legal powers to combat fake news and foreign interference in domestic political affairs. For example, ‘stop communication directions’ can be issued under the POFMA to require a person who has communicated a false statement of fact to stop communicating the offending statement.⁵ The FICA empowers a Minister to authorise ‘must-carry directions’ against social media or relevant electronic service providers, which would require them to put up a mandatory message with respect to specified material – for example, stating that the content in question is part of a hostile information campaign.⁶

The advent of the POFMA and FICA makes Singapore a global leader in the implementation of legal measures to combat the scourge of online misinformation, and the legal framework of these statutes is likely to be of considerable interest to a worldwide audience. In addition, the introduction of these statutes represents one of the most momentous recent developments in Singapore law. Given the nature of the powers granted by these statutes, it should be abundantly clear that both the POFMA and FICA present a variety of possible implications for the right to freedom of speech enshrined in Article 14 of the Singapore Constitution. Some of these implications are rather direct. For example, stop communication directions issued under the POFMA clearly amount to restrictions on one's freedom to speak.

Beyond these obvious implications, however, the POFMA and FICA open up novel frontiers of development for Singapore's free speech jurisprudence. Specifically, the power granted by the POFMA and FICA to compel recipients of certain directions to communicate a specific message implicates the doctrine of compelled speech. Indeed, does being compelled to make a statement offend one's right not to speak, as a correlative of one's freedom of speech? This is a question which has thus far received little attention in Singapore's free speech jurisprudence; a stark contrast to Canadian, UK, and US free speech jurisprudence, which have all accepted the doctrine. This question takes on special importance when one observes that the power most frequently used under these statutes thus far is the issuance of POFMA correction directions – directions compelling their recipient to append a message to certain specified published material stating that the material included a false statement of fact.⁷

This article therefore aims to capitalise upon this opportunity brought by the advent of the POFMA and FICA to explore whether and how the compelled speech doctrine should be accepted in Singapore constitutional law, and whether the POFMA and FICA would be implicated by such a doctrine. The article draws upon theoretical, comparative, and doctrinal analysis. It first briefly describes the rationale for the POFMA and FICA, and explains how these statutes raise the issue of compelled speech. The article also discusses the recent Singapore Court of Appeal decision in The Online Citizen Pte Ltd v Attorney-General ⁸ (‘Online Citizen’), as the latest development in Singapore law material to the focus of this article. With the foundation thus laid, the article turns to analyse the possible theoretical justifications for the doctrine of compelled speech, and examines the extent to which these justifications cohere with those underlying Singapore's free speech jurisprudence. By reference to free speech theory, as well as Canadian, UK, and US constitutional law, the article proposes a doctrinal form for the compelled speech doctrine in Singapore. Finally, the article returns full circle to the POFMA and FICA, evaluating how these statutes would fare under the proposed compelled speech doctrine.

For the avoidance of doubt, the constitutional implications of the POFMA and FICA certainly extend beyond the issue of compelled speech. For instance, the extent of the permitted legal checks on powers exercisable under these statutes merits analysis for coherence with the judicial power codified in Article 93 of the Singapore Constitution. Such issues are important to investigate. But they deserve a separate dedicated analysis. Accordingly, for clarity of scope, this article focuses on the implications of the POFMA and FICA for compelled speech in Singapore constitutional law.

The POFMA, FICA, and compelled speech

The POFMA can be best understood as forming an important part of a remedy to an emerging ‘fundamental problem’: the evidence around the world of ‘a serious loss of trust in governments, in institutions both public and private, including the political system, the media, professions, businesses, financial institutions and so on’.⁹ As Singapore's Minister for Law, Mr K Shanmugam SC, explained in Parliament, the foundations of an active democracy include trust and an infrastructure of fact facilitative of proper public discourse.¹⁰ The proliferation of online falsehoods, however, threatens to destroy these foundations. Singapore's Second Minister for Law, Mr Edwin Tong SC, added that ‘the core trick of falsehoods lie in their use to arouse anger, fear and negative emotions’, thereby exploiting ‘cognitive biases’.¹¹ He argued that ‘it is very difficult for facts to overcome falsehoods organically’, in view of the fact that ‘falsehoods move and take effect quickly’.¹² Accordingly, in view of ‘the importance of putting out corrections swiftly and circulating them vigorously’, the introduction of the POFMA was intended to provide the Singapore government with the proper tools to be able to do precisely that, where the public interest so requires.¹³

Turning to the FICA, it can be broadly divided into two main parts. The first part is directed at thwarting online hostile information campaigns conducted by foreign parties, while the second part is directed at countering foreign interference through local proxies. It is the first part that is most relevant to this article. This part of the FICA stems from a recognition that the online domain has become a new front in modern warfare – as Mr Shanmugam SC explained, ‘countries are actively developing attack and defence capabilities as an arm of warfare equal to and more potent than the land, air and naval forces’. The Singapore government was concerned about the exploitation and inducement of racial and religious tensions by hostile foreign actors, as well as the fact that political debate and public opinion are susceptible to being shaped by foreign states to be more aligned with their interests.¹⁴ While existing legislation did indeed confer powers upon the government to address such threats to a certain degree, Mr Shanmugam SC argued that there was a need to update these powers ‘to ensure that they are fit for the Internet age’.¹⁵

The prerequisites for the powers within these statutes to be exercised share some similarities. The key powers under the POFMA can only be exercised if material containing a false statement of fact has been communicated in Singapore, and if the relevant government minister considers that it is in the public interest to issue a POFMA direction.¹⁶ As for the FICA, the relevant powers under the FICA to combat online misinformation can be exercised if online communications activity has been undertaken or is suspected of having been undertaken on behalf of a foreign principal, such online communications activity has resulted in information being published in Singapore, and if it is in the public interest to issue the relevant FICA direction.¹⁷

In relation to whether the prerequisites for these powers have indeed been met, the POFMA provides that appeals can be made to the High Court against the issuance of POFMA directions.¹⁸ However, whether it is in the public interest to issue a direction is not a permissible basis of appeal under the POFMA. As for the FICA, appeals against FICA directions can be made to the relevant government minister,¹⁹ failing which a further appeal may be made to a reviewing tribunal.²⁰ Notably, the FICA ousts judicial review of decisions made under the Act, except where questions of procedural compliance with the Act are concerned.²¹ Mr Shanmugam SC justified such ouster on the ground that the ‘judicial process will not be appropriate in matters where we rely heavily on sensitive intelligence and collaboration with foreign counterparts’.²²

Turning then to the actual powers contained in these statutes, it will be observed that many of them impose straightforward restrictions on free speech. For example, ‘take-down’ directions under the POFMA, as the nomenclature suggests, require the recipient – which can include individual publishers and internet intermediaries – to take down an identified falsehood and cease communicating it.²³ Similar directions are available under the FICA as well.²⁴

Our main interest here, however, is with the provisions of the POFMA and FICA which compel recipients of certain directions to display specified messages. Indeed, a number of POFMA directions fall under what has been called the ‘corrections regime’.²⁵ Such directions would allow the falsehood to stay published, but would require a notice of correction to be displayed alongside the originally-published falsehood. These directions can be targeted specifically at viewers of a falsehood, and can be issued against both individual publishers and internet intermediaries,²⁶ or generally promulgated on online platforms whether or not those platforms are carrying the falsehood.²⁷ Similarly, the FICA provides for the issuance of must-carry directions, which can be issued to a variety of recipients to require them to display a message notifying viewers that the material being viewed is part of a hostile information campaign.²⁸

These provisions spotlight for discussion an interesting point of free speech jurisprudence. A conventional understanding of the right to freedom of speech is that it renders unconstitutional restrictions upon one's speech that are unjustifiable. This is a straightforward understanding of the right to free speech—one's right to free speech means that one ought not to be restricted in speaking. However, does this right to free speech encompass the freedom not to speak? One might argue that a natural corollary of the freedom to speak is the freedom not to be compelled to speak when one does not wish to. Yet, the effect of the POFMA and FICA directions described in the preceding paragraph is to compel the recipient to communicate a message that one would surely ordinarily not wish to communicate. If one's right to free speech indeed encompasses the freedom not to speak, then such directions may be constitutionally problematic. The power to grant such directions contained within the POFMA and FICA therefore raises for discussion a novel and thus far underdeveloped frontier of free speech jurisprudence in Singapore – the doctrine of compelled speech, which will be the central focus of discussion in this article.

Compelled Speech in Singapore Constitutional Law thus Far

The issue therefore is whether the right to free speech in Article 14 of Singapore's Constitution prohibits compelled speech; and if so, whether the POFMA and FICA would be problematic by reference to such a prohibition. The subsequent sections will aim to explore these questions by way of theoretical and comparative analysis.

The Singapore Court of Appeal decision in Online Citizen is a necessary starting point for our discussion of the compelled speech doctrine, being the only Singapore court decision thus far which has given consideration to the doctrine. The appellants in Online Citizen had published certain articles and Facebook posts. These articles and posts stated, in essence, that there was a rising proportion of local PMETs (professionals, managers, executives and technicians) being retrenched, and also that Singapore prisons officers had been instructed to use brutal methods to execute prisoners on death row. The parties responsible for each set of posts were subsequently issued correction directions under the POFMA, requiring them to attach correction notices to the originally-published material. Both parties appealed under section 19 of the POFMA to the Ministers who had issued the directions. These appeals were unsuccessful. They then each duly commenced separate proceedings in the High Court pursuant to section 17 of the POFMA to set aside the directions issued against them. The High Court dismissed both appeals, following which both appellants appealed to the Court of Appeal. Both appeals were consolidated and heard together in Online Citizen.

Of particular relevance to the central concern of this article, the appellants raised before the Court of Appeal several arguments challenging the constitutionality of the POFMA. In addressing these arguments, the court drew upon the three-step analytical framework it had articulated previously in Wham Kwok Han Jolovan v Public Prosecutor ²⁹ (‘Jolovan Wham’) governing judicial review challenges based upon Article 14, the first step of which required the court to consider whether the legislation in question did derogate from or restrict Article 14 rights in the first place.³⁰

It is the court's reasoning on this first step of the Jolovan Wham test that is of greatest relevance for present purposes. At this stage of analysis, the Court of Appeal held that correction directions do not restrict the recipient's positive right to free speech. Indeed, a correction direction leaves its recipient entirely free to continue publishing the original material. All that it requires the recipient to do is to put up the additional notice required by the correction direction alongside the original material.³¹ Accordingly, if a correction direction were to infringe upon the recipient's right to free speech, such an infringement would have to be based on a negative conception of the right to free speech – specifically, that the freedom to speak also necessarily entails a freedom not to speak.³²

The Court of Appeal therefore took the opportunity to discuss the doctrine of compelled speech for the first time in Singapore.³³ The court sought to discern the nature of the compelled speech doctrine by reference to foreign jurisprudence. It considered the US Supreme Court decision of Wooley v Maynard ³⁴ (‘Wooley’), where a Jehovah's Witness had sought to challenge New Hampshire's legal requirement that the state motto, ‘Live Free or Die’, had to be displayed on automobile licence plates. The majority in Wooley ultimately held that such a legal requirement did indeed violate the complainant's First Amendment rights, given that the First Amendment protected ‘both the right to speak freely and the right to refrain from speaking at all’.³⁵

However, the Court of Appeal in Online Citizen was more persuaded by the minority's reasoning in Wooley. In Wooley, the dissenting minority argued that there was a distinction between merely communicating a point of view and being forced to assert it as true, and thought that since New Hampshire's laws did not prevent the complainants from ‘displaying their disagreement with the state motto as long as the methods used do not obscure the license plates’, the law in question did not compel an endorsement of the message in the state motto and would therefore not fall foul of the compelled speech doctrine.³⁶ Agreeing with this distinction, the Court of Appeal in Online Citizen held that ‘one's freedom not to speak so as not to be forced to advance a position with which one disagrees cannot be said to have been inhibited, prevented or materially limited if one is compelled to display that position, but retains the freedom to qualify it appropriately in an equally visible manner’.³⁷

The Court of Appeal in Online Citizen also analysed the UK Supreme Court decision of Lee v Ashers Baking Co Ltd and others ³⁸ (‘Lee v Ashers’) through the lens of this distinction. In Lee v Ashers, the Christian owners of a bakery declined to bake a cake decorated with the slogan ‘Support Gay Marriage’, giving rise to the issue of whether the owners’ right to freedom of speech would be violated if they were forced by law to make the cake in question. The UK Supreme Court ultimately found that compelling the bakery owners to bake the cake would indeed fall foul of the compelled speech doctrine, since the owners could not be compelled to convey a message with which they profoundly disagreed.³⁹ The court in Online Citizen thought that this conclusion was a justifiable one – indeed, the court thought that the distinction between Lee v Ashers and Wooley was that the complainants in Wooley were not precluded from expressing their disagreement with the state motto, while the bakery owners in Lee v Ashers would not have had ‘the option of expressing its disagreement with [the pro-gay marriage caption] in the same medium (namely, on the cake)’.⁴⁰

It is therefore quite apparent that in the Court of Appeal's view, whether one has the ability to disavow the message being compelled is quite crucial in determining whether there has been impermissible compulsion of speech. Applying this to the POFMA correction directions at hand, the court noted that POFMA directions do not prevent their recipients from displaying their disagreement with the correction notice by posting, for example, an accompanying message that the statement in question was actually a true statement of fact.⁴¹ Accordingly, the court concluded that the issuance of the correction directions was not capable of constituting impermissible compelled speech at all, and that it was unnecessary for the court to make a conclusive pronouncement on the status of the compelled speech doctrine in Singapore constitutional law at this juncture.⁴² In the context of the Jolovan Wham test being applied here, this meant that the appellants’ challenge to the constitutionality of the POFMA failed to surmount the first step in the Jolovan Wham test.⁴³

The Court of Appeal's willingness to engage with the doctrine of compelled speech – uncharted waters in Singapore's free speech jurisprudence – is commendable. Yet, many questions about the compelled speech doctrine in Singapore remain to be decided in the wake of Online Citizen. First, should Singapore constitutional law accept the doctrine of compelled speech at all? In answering this question, it will be useful to reflect upon the theoretical justifications for the compelled speech doctrine, and on whether these justifications cohere with Singapore's Article 14 jurisprudence. Second, what doctrinal form should the compelled speech doctrine take in Singapore? Third, would the POFMA and FICA fall foul of the compelled speech doctrine thus conceived?

Each of these questions will be discussed in turn. The following discussion will be both theoretical and comparative in nature – indeed, Singapore's constitutional jurisprudence can draw fruitfully from the experience of the US, UK, and Canadian courts in reflecting upon this doctrine. The Singapore Court of Appeal's comments on the doctrine of compelled speech will also be critically evaluated in the course of the following discussion.

Theoretical Justifications for Compelled Speech

One might instinctively think that there is something objectionable about being compelled by law to express a certain position, whether or not one is free to disavow such a statement. Being compelled by law to express a view which one does not hold seems quite self-evidently disagreeable. At the same time, the law does regularly compel speech. For example, disclosure requirements, as well as the requirement to make certain oaths upon the assumption of public offices or the performance of public duties, all can be said to be rather commonplace and widely-accepted compulsions of speech.⁴⁴

The question then is what makes some instances of compelled speech objectionable and others unobjectionable. To answer this in a principled manner, it will be useful to first articulate what the theoretical justifications underlying the compelled speech doctrine are. Indeed, various justifications for the compelled speech doctrine have been offered in the case law and academic literature of jurisdictions which have accepted the doctrine. For example, in describing the compelled speech doctrine in US constitutional law, the US Supreme Court in the relatively recent decision of Janus v American Federation of State, County, and Municipal Employees, Council 31 (‘Janus’) held that restrictions on free speech generally damage ‘our democratic form of government’ and ‘the search for truth’.⁴⁵ However, compulsion of speech does “additional damage”—indeed, Justice Alito, writing for the majority, held that compelled speech coerces individuals into ‘betraying their convictions’, which is problematic since ‘[f]orcing free and independent individuals to endorse ideas they find objectionable is always demeaning’.⁴⁶

This elaboration in Janus of what makes compelled speech problematic falls into one of the major categories of justifications for the doctrine, which can be described as dignity or autonomy-based. On this view, independent of its effects on democracy or the marketplace of ideas, compelled speech is problematic for the bare fact that a compulsion to express statements represents an intrusion into personal autonomy.⁴⁷ Indeed, Vikram David Amar and Alan Brownstein argue that ‘individuals living with dignity must be able to control what they say’.⁴⁸ The harm to one's autonomy engendered by compelled speech is directed at ‘the capacity for the unwilling speaker to remain silent, without any correlated effect on whether she also wants to express speech of her own’.⁴⁹ Should a person be compelled to express statements that one does not wish to express, such action may also put the person's ‘speech in tension with her moral integrity and independence’.⁵⁰

The dignitary harm inflicted by compelled speech obtains even if the speaker is able to disavow the statement, and even if the relevant message is not attributed to the individual at all – the ‘pervasive commandeering of an individual's voice’ is a ‘serious dignitary concern’ in itself, and it is deeply problematic to be ‘transformed into an expressive tool to serve the state's purpose’.⁵¹ As Seana Valentine Shriffin argued, transforming a person into ‘a mere means for transmitting the state's message, rather than as an independently-minded thinker and speaker’ conflicts with the respect due to ‘a person with dignity’.⁵² Steven H Shriffin has also argued that forcing individuals to be ‘couriers of messages to which they are ideologically opposed’ amounts to ‘a particular kind of assault on human dignity’.⁵³

Bearing in mind the focus of this article on exploring the possibility of the compelled speech doctrine in Singapore constitutional law, a key problem that arises with this dignity or autonomy-based justification for the doctrine is that the Singapore courts have not grounded Singapore free speech jurisprudence on such justifications. Indeed, it is instructive in this regard to observe how the Singapore courts have dealt with the question of whether false speech enjoys constitutional protection. In a series of cases, the Singapore courts have repeatedly held that there is no reason to accord false speech constitutional protection, on account of the fact that such speech cannot make any meaningful contribution to the ‘competition of ideas in the marketplace’.⁵⁴ In Review Publishing Co Ltd and another v Lee Hsien Loong and another appeal, the Court of Appeal noted that ‘there is no interest in being misinformed’,⁵⁵ and in Attorney-General v Ting Choon Meng and another appeal, Chief Justice Sundaresh Menon in his dissenting judgment argued that “false speech, which has been proven as a matter of fact to be false in a court of law, can contribute little to the marketplace of ideas or to advances in knowledge for the benefit of society as a whole”, thereby placing false speech outside the realm of ‘the theoretical justifications underpinning the liberty of persons relating to free speech’.⁵⁶ These comments were all cited with approval in Online Citizen.⁵⁷

All of this provides a useful indication as to how the Singapore courts are likely to view the dignity or autonomy-based theoretical justification for compelled speech. It is quite telling that the Singapore courts were not keen on according constitutional protection to false speech. If the Singapore courts had been inclined to accept the dignity or autonomy-based rationale for free speech, this would in all likelihood have been expressed in more solicitude for the possibility that even false speech could be constitutionally protected – restrictions on such speech, after all, do restrict one's autonomy and can be said to be harmful to one's ability to fully express oneself.⁵⁸ The Singapore courts’ preference for the truth-based theoretical justification for free speech instead is therefore quite clearly illustrated by this line of cases.

This does not necessarily mean, however, that the compelled speech doctrine has no place in Singapore's free speech jurisprudence. Indeed, the compelled speech doctrine can also be justified on the basis of truth-based justifications for free speech. From the perspective of this justification for free speech, compelled speech would be constitutionally problematic to the extent that it distorts the marketplace of ideas, hampering the free exchange of ideas essential for a better apprehension of the truth.

Compelled speech can have an undesirable effect upon the marketplace of ideas in the following ways. First, if the compulsion to speak forces the speaker to be identified with the compelled message, this can have a distortive effect on public debate by suggesting greater support than actually exists for the message being expressed.⁵⁹ As Amar and Brownstein argued, ‘[t]he audience of the compelled speech may believe that private speakers with their own independent credibility and followers support the government's position when in fact they do not do so’.⁶⁰ Martin H Redish pointed out that this forced identification of a speaker with the message potentially confuses the people as to ‘the actual strength and popularity of substantive positions advocated by the government’, which can then distort public debate and electoral processes.⁶¹

Second, compelled speech can distort the marketplace of ideas by allowing the government to avoid the financial and logistical costs of conveying a message, thereby giving the government an increased ability to overwhelm the public square with its message. Such action would unfairly tilt the marketplace of ideas in favour of the message the government wishes to convey. In addition, allowing the government to avoid its usual limitations in conveying a message can give the government an advantage in the marketplace of ideas by allowing it to convey messages through media and venues which ‘it might have difficulty reaching through its own expressive resources’.⁶²

Third, compelled speech can distort the marketplace of ideas by affecting the speaker's ability to communicate his or her own message. This indirect limitation upon speech can occur by way of a ‘tax’ on speech – for example, in the US Supreme Court decision of Miami Herald Publishing Co v Tornillo, the court found unconstitutional a statute that required newspapers which criticised electoral candidates to print responses by the candidate in a comparable size and location in the newspaper, on the basis that such a requirement would amount to a “tax” discouraging critical commentary.⁶³ Compelled speech can also indirectly restrict speech if the amount of speech that a speech medium can support is limited – any compelled speech would accordingly reduce the space for the speaker to convey his or her own message.⁶⁴ Eugene Volokh described this effect of compelled speech as an interference with a ‘coherent speech product’ which threatens to drown out the message which the speaker actually wants to communicate.⁶⁵ It is worth noting, however, that this effect of compelled speech would be less relevant for speech carried out in the internet domain, given that the internet is effectively limitless as a medium of speech.⁶⁶

Fourth, compelled speech can exercise a distorting influence on the marketplace of ideas by diluting the persuasiveness of speakers. Indeed, if a person is compelled to state a position which one does not believe in, then even if the person is able to simultaneously disavow the compelled statement and dissociate from it, the speaker's persuasiveness in the eyes of his or her listeners will undoubtedly be diluted. As Redish argued, a speaker wishing to advocate for position X ‘can hardly be deemed an effective advocate for her position if she is simultaneously forced by government to advocate for position Y, or, even more concerning, position not-X’.⁶⁷ The impact of the speaker's message will be ‘severely diluted by the proximately forced utterance of the diametrically opposed position’.⁶⁸ This forced dilution of one's message can have a demoralising effect upon speakers, exerting a chilling effect upon their desire to continue engagement in the public square.⁶⁹

Fifth, and finally, compelled speech can exercise an insidious effect on the marketplace of ideas by psychologically influencing the speakers themselves. Indeed, as Ashutosh Bhagwat has argued, ‘one might be concerned that compelling people to speak the government's message might inculcate that message, and its attendant values, in the speakers themselves’.⁷⁰ By being compelled to express certain statements, a speaker may ‘subconsciously rationalise his forced statements by coming to believe them’⁷¹ – an effect of compelled speech which would have a manipulative effect on the marketplace of ideas.

Each of these truth-based justifications for a compelled speech doctrine will be evaluated in detail subsequently against the specific context of the compelled speech required by the POFMA and FICA. For present purposes, the key takeaway is that there is sufficient theoretical justification for a doctrine of compelled speech to be accepted in Singapore's free speech jurisprudence. Indeed, given that compelled speech can have a distortive effect on the pursuit of truth – a theoretical foundation for free speech accepted in Singapore – a doctrine of compelled speech would be consistent with Singapore's Article 14 jurisprudence.

Proposing a Doctrine of Compelled Speech in Singapore

The question that arises then is what form a doctrine of compelled speech should take in Singapore. This section will propose a suggested doctrinal form for compelled speech in Singapore, taking reference from foreign jurisprudence where relevant and paying attention also to the theoretical justifications upon which the compelled speech doctrine would be based in Singapore.

In summary, in determining whether there is a compulsion of speech forming a prima facie infringement upon Article 14(1) in the first place, the following analysis can be utilised: (1) Has there been compelled expression at all? If so, then (2) is there a forced identification of the speaker with the compelled message, or (3) is the speaker unable to disavow the compelled message?⁷² As long as either of the second or third requirements is met, there would be a prima facie problematic compulsion of speech under Article 14(1). If there is indeed compelled speech leading to a prima facie violation of Article 14(1), the relevant authority would have to justify the compulsion of speech by reference to the permissible restrictions upon Article 14(1) rights set out in Article 14(2). Each component of the proposed test will be discussed in turn.

Compelled Expression

It stands to reason that the first step in determining whether there has been prima facie impermissible compelled speech should be to consider whether there was indeed compelled expression at all. The rationale for such a step is straightforward – logically, for something to amount to compelled speech, the compelled activity must have some form of expressive content in the first place.⁷³ For example, taking reference from the facts in Wooley, a legal requirement to display car licence numbers would not be expressive in nature, while a legal compulsion to display the message ‘Live Free or Die’ would indeed be a situation of compelled expression.

While this is a straightforward component of a test for compelled speech, the experience of jurisdictions which have accepted the compelled speech doctrine shows us that it can raise difficult questions for courts to grapple with. A useful example is the decision of the Supreme Court of Canada in Lavigne v Ontario Public Service Employees Union ⁷⁴ (‘Lavigne’). In this case, the appellant challenged a law compelling the payment of dues to the Ontario Public Service Employees Union, on the basis that the union supported a certain set of causes which the appellant did not agree with, and that the compulsion to pay dues to an organisation in support of such causes would amount to an infringement upon the appellant's freedom of expression.⁷⁵

Justice Wilson in Lavigne held that the starting point was indeed whether the relevant activity ‘conveys a meaning’.⁷⁶ The question therefore was whether compelled payment of dues in such a situation contained expressive content. Justice Wilson came to the conclusion that it did. Holding that ‘in this day and age where money is an extremely powerful way of expressing support, the channelling of contributions is expressive indeed’, Justice Wilson concluded that ‘the fact that [the appellant] is denied the right to boycott the Union's causes prevents him from conveying a meaning which he wants to convey’.⁷⁷ Accordingly, the mandatory payment of dues did amount to compelled expression on the part of the appellant.

However, Justice La Forest in the same case disagreed on this point. He argued that he could not see how the appellant, as ‘one of many who either willingly or unwillingly contribute to the Union, can be said to be responsible for the use to which the money contributed is put’.⁷⁸ Indeed, in his view, he thought that the manner in which the union expends the funds would ‘not be regarded as an expression of the thoughts and opinions of the many individuals who contribute to the Union's coffers’.⁷⁹ Accordingly, contrary to the view of Justice Wilson, the mandatory payment of dues ought not to be considered as amounting to an expression on the part of the appellant at all.

The various opinions in Lavigne provide an example of the kinds of issues that judges can encounter at this step of analysis – whether a particular activity constitutes expressive activity at all. Looking more closely at Justice La Forest's reasoning, one might observe that there are two ways of framing the point he sought to make. The first is that Justice La Forest did not think that the mandatory payment of dues could amount to expressive content at all. Alternatively, one might argue that Justice La Forest did think that the expenditure of funds could in principle have expressive content, and that the precise ground of his disagreement with Justice Wilson was whether this expressive content should be attributed to the appellant or the union. This is a point which will be discussed in greater detail in the next section. Nevertheless, the key point to emphasise at this juncture is that while this step of analysis can raise certain difficulties, it is a question which courts can and must inevitably grapple with. Indeed, it ultimately depends on how one interprets the words ‘speech’ and ‘expression’ in Article 14(1) of the Singapore Constitution, a question which is equally relevant to free speech jurisprudence more generally and is by no means unique to the context of compelled speech.

Returning to the difficulties involved in determining what constitutes expressive activity for the purposes of compelled speech more specifically, the Supreme Court of Canada's decision in Canada (Attorney-General) v JTI-MacDonald Corp provides another useful example. One of the issues the court had to grapple with in this case was whether mandatory health warnings on tobacco packages contravened the manufacturer's freedom of expression. The court's conclusion was that the manufacturer's freedom of expression was indeed prima facie infringed, on the basis that silence can constitute expressive activity.⁸⁰ While the court ultimately held that the mandatory health warnings were constitutional on the ground that they were rationally justifiable, the court's reasoning on the nature of expressive activity was quite noteworthy. One might be sympathetic to the view that there are circumstances under which one would prefer to remain silent. Yet, if silence constitutes expressive content, this means that any compulsion to speak would be capable of triggering a freedom of expression violation. Careful legal boundaries may therefore need to be drawn to limit the reach of this principle.

One possibility might be to limit the applicability of this principle where the compulsion of factual or non-controversial expressions is concerned. A more permissive attitude towards the compulsion of factual and uncontroversial statements can be found in the US Supreme Court's compelled speech jurisprudence.⁸¹ For instance, in Zauderer v Office of Disciplinary Counsel of Supreme Court of Ohio (‘Zauderer’), the court upheld a legal requirement that lawyers placing advertisements had to also disclose that clients had to pay certain fees and costs, primarily on the basis that such information was ‘purely factual and uncontroversial information’.⁸² Applying this principle, the US Supreme Court in National Institute of Family and Life Advocates v Becerra held that a requirement for pregnancy counselling clinics to disclose information about state-sponsored abortion services concerned ‘anything but an “uncontroversial” topic’, thereby falling out of the ambit of the Zauderer principle.⁸³

It is suggested that making an exception for factual and uncontroversial expressions for the purposes of the compelled speech doctrine is reasonable. Indeed, this caveat at this stage of analysis is quite necessary to take into account the numerous and commonplace instances when one is compelled by law to make certain statements – financial disclosures and vehicle licence numbers, for example.⁸⁴ As Volokh argued, ‘the constitutionality of such pure factual compulsions is particularly important because we are all routinely required to state various facts to the government’, such as our annual income for tax purposes or police reports where we are witnesses to a crime.⁸⁵ The compulsion of such purely factual statements would also not implicate any of the truth-based rationales for the impermissibility of compelled speech described above. Accordingly, the compulsion of factual and non-controversial expressions ought not to be considered compelled expression at all under this first stage of analysis.

Forced Identification

Should the court find that there has indeed been a compulsion of speech satisfying the first step of the test, the next step ought to be to determine whether the compulsion of speech forces the speaker to identify with the message being compelled. If so – for example, if it is not clear that the compelled message is a result of government compulsion, and the compelled message is attributed to the speaker – then this would amount to prima facie objectionable compelled speech.

Including such a doctrinal requirement in the compelled speech doctrine would give effect to the first two truth-based theoretical justifications for the doctrine described in the preceding section – that is, the concern that compelled speech can distort public debate by falsely bolstering the amount of support that actually exists for the message being expressed, and also the concern that the government will be able to overwhelm the public square with its message by avoiding the financial and logistical costs of communicating messages and entering into public fora which it might ordinarily have difficulty accessing. The inclusion of this doctrinal requirement addresses both of these concerns. Indeed, if the compelled speech in question does not identify the speaker with the message at all and is unambiguously a result of government compulsion, then there is no misrepresentation of the amount of support for the compelled message. The government would also not be given any unfair advantage in the marketplace of ideas if the compelled message will be unmistakeably attributable to the government, since it will continue to labour under the same disadvantages in the public fora it seeks to access. Accordingly, such compelled speech ought not to fall foul of freedom of speech under Article 14.

This doctrinal requirement also finds support in US and Canadian constitutional law. In determining whether an instance of compelled speech is constitutionally impermissible, the Supreme Court of Canada has consistently taken into account whether the compelled speech in question forces the speaker to be identified with the message as part of its assessment.⁸⁶ Indeed, Justice Wilson in Lavigne, drawing upon a survey of Canadian and US case law, considered ‘public identification’ to be one of the relevant factors in the determination of whether compelled speech infringed upon the freedom of speech.⁸⁷ On this basis, she reasoned that the mandatory payment of dues to a union would ultimately not be constitutionally impermissible, since the payment of such dues would not result in the union's messages being attributed to the views of the payers. Moving across the border, the US Supreme Court in PruneYard Shopping Center v Robins held that allowing a group of protestors to engage in expressive activity within a shopping centre would not violate the shopping centre owner's freedom of speech, since a shopping centre was ‘a business establishment that is open to the public to come and go as they please’, and that any views expressed by members of the public in this space ‘will not likely be identified with those of its owner’.⁸⁸

The UK Supreme Court's decision in Lee v Ashers, discussed earlier, provides an example of how the inclusion of such a doctrinal step can be useful. The court's conclusion – that compelling the bakery to write a slogan on a cake would amount to forcing them to express ‘a message with which they profoundly disagreed’⁸⁹ – was clearly premised on the assumption that the bakery would be identified with a slogan on a cake they baked. Indeed, otherwise, the bakery would not be expressing any message at all.⁹⁰ However, as Kenneth M Norrie has argued, this assumption was a contentious one. Pointing out that ‘[b]roadcasters are not themselves expressing a view when they give airtime to those who offer opinions’, he suggested that it was ‘not obvious that the shop owners were “expressing” any view at all by the act of selling a cake with words on it’.⁹¹ Accordingly, given the contentious nature of this assumption, this point of analysis could have benefitted from more judicial attention – attention which will be facilitated by the inclusion of this step in compelled speech doctrine.

Inability To Disavow

The third inquiry that the courts ought to undertake in assessing whether there has been impermissible compulsion of speech is to consider the speaker's ability to disavow the message being compelled. If there is a compulsion of speech, and if the speaker's ability to disavow the compelled message is practically restricted, then this would amount to a constitutionally impermissible compulsion of speech.

Such a doctrinal requirement would account for the final three truth-based theoretical justifications for the compelled speech doctrine discussed earlier – the concern that compelled speech would indirectly restrict the speaker's ability to communicate his own message, the concern that it would dilute the persuasiveness of the speaker in relation to his own message, and the concern that it would psychologically manipulate the speaker into accepting the message being compelled. If the speaker remains fully able to disavow the compelled message, then there is no restriction on the speaker's ability to communicate his own message. There would also be no effect on the speaker's persuasiveness – the speaker's persuasiveness would depend on the force of the speaker's rebuttal to the compelled message, and is not necessarily dampened by the fact of compulsion alone. Finally, if a speaker retains the ability to vigorously disavow the compelled message, the risk of psychological manipulation into accepting the compelled message will be minimal.

The inclusion of this step would also take onboard the Singapore Court of Appeal's own reflections on the compelled speech doctrine in Online Citizen. Indeed, as described earlier, the court took the view that ‘one's freedom not to speak … cannot be said to have been inhibited, prevented or materially limited if one is compelled to display that position, but retains the freedom to qualify it appropriately in an equally visible manner’.⁹² The court also thought that the UK Supreme Court's decision in Lee v Ashers was justified on the basis that the bakery in this case did not have ‘the option of expressing its disagreement with [the pro-gay marriage caption] in the same medium (namely, on the cake)’.⁹³

Yet, while this is a sensible consideration to include in a doctrine of compelled speech, it ought to be noted that the manner in which the Court of Appeal framed this consideration can raise certain thorny issues. Indeed, it is apparent in the court's reasoning that whether the speaker is able to express one's disagreement with the compelled message in the ‘same medium’ or ‘equally visible manner’ is quite crucial to the court's analysis. Such a requirement can be challenging to apply. For example, what amounts to the ‘same medium’ or ‘equally visible manner’?⁹⁴ If a driver cannot express his disagreement with a mandatory slogan on a vehicle licence plate by amending the licence plate, but can do so on a bumper sticker or windshield sticker, does that count as the ‘same medium’? Or if a baker can express his disagreement with a customised slogan on a cake by writing a note on the cake box, would that amount to an ability to disavow the slogan in an ‘equally visible manner’?

These difficulties suggest that instead of relying on doctrinal distinctions that may be prone to formalistic reasoning, it will be prudent to frame this requirement as a pragmatic assessment of whether the speaker had the practical opportunity to disavow the compelled message. Justice Wilson in Lavigne was clearly alive to the importance of framing this inquiry in a practical manner – pointing out that ‘care should be exercised in considering whether or not one truly has the opportunity to disavow’, she added that ‘[o]pportunity must be meaningful and we should not be too quick to ascribe to persons opportunities and abilities which they do not really possess’.⁹⁵

The recent Canadian Federal Court decision in Right to Life Association of Toronto v Canada (Employment, Workforce, and Labour) ⁹⁶ (‘Right to Life’) provides a useful example of how this inquiry can be conducted in a pragmatic manner. In Right to Life, the applicants sought to challenge a mandatory attestation that served as a condition for eligibility for a government funding scheme. The attestation required applicant organisations to affirm their respect for reproductive rights – an attestation which would have betrayed the convictions of the Right to Life Association. The association therefore argued, among other things, that the attestation requirement contravened the constitutional right to freedom of speech as an instance of impermissible compelled speech.

The court applied the factors set out in Lavigne to determine if this was indeed an instance of compelled speech implicating the association's right to free speech – that is, whether the speaker was publicly identified with the compelled message, and whether the speaker had the opportunity to disavow the message.⁹⁷ On the point of whether the speaker had the opportunity to disavow the message, the government argued that the attestation did not prevent the association from expressing their disagreement with it in other fora, and therefore ought not to be considered as objectionable compelled speech.

The court rejected this argument. Adopting ‘a practical perspective’, the court acknowledged that while the association could indeed express its disagreement with the attestation in other fora, doing so would ‘call into question the honesty and credibility of the organisation’.⁹⁸ The court thought that the association ‘could not practically and credibly make the Attestation’, since doing so would have identified the association with a view on reproductive rights which it did not accept. Accordingly, on the basis of this practical assessment, the court held that the association did not have ‘a real opportunity to disavow the Attestation’, which meant that its right to freedom of speech was prima facie engaged.⁹⁹

The court in Right to Life ultimately held that this infringement was a proportionate one when weighed against ‘the important objective of creating an inclusive labour workforce’, and dismissed the challenge to the attestation.¹⁰⁰ Nevertheless, the key point to note from this case is that a pragmatic and common-sense approach has to be taken towards the assessment of whether the speaker indeed had the practical ability to disavow the message being compelled in a manner that would address the distortive effect of compelled speech on the marketplace of ideas.

Applying such an approach to the hypothetical scenarios raised earlier, a driver's ability to express his or her disagreement with the compelled message on the licence plate by placing a bumper sticker or windshield sticker would probably indeed amount to a practical opportunity to disavow the message – such a sticker would make sufficiently clear that the driver disagrees with the compelled message, thereby getting directly at the root of the compelled speech problem here. As for the baker being compelled to deliver a cake bearing a slogan with which he or she disagrees, a practical approach towards thinking about this requirement may yield the conclusion that even though the baker indeed cannot express his or her disavowal of the compelled message on the cake itself, the crucial pragmatic point is that the baker wants to make clear to any prospective audience that the baker does not agree with the compelled message – this can be accomplished readily by placing a message on the cake box near the bakery's logo, even if the disagreement cannot be expressed on the cake itself.

Should the court find that there was indeed a situation of compelled expression under the first step of analysis, and should the court find that the speaker was forced to identify with the compelled message, or that there was no practical opportunity for the speaker to disavow the compelled message, then there will be a prima facie impermissible instance of compelled speech under Article 14(1). Applying the Jolovan Wham framework that the Singapore courts have adopted to analyse infringements of the right to free speech in Article 14, the next relevant steps of the inquiry would be to consider whether the infringement was necessary or expedient in the interests of one of the enumerated purposes under Article 14(2), and whether the infringement indeed fell within the relevant and permitted purpose.¹⁰¹ The prima facie impermissible compelled speech will be found constitutional if both of these steps are answered in the affirmative.

Evaluating the POFMA and FICA through the compelled speech doctrine

By way of recap, the preceding sections have argued that a doctrine of compelled speech can cohere with Singapore's free speech jurisprudence to the extent that it can be justified on the basis of truth-based theoretical justifications centring around the distortive effect of compelled speech on the marketplace of ideas. The preceding sections have also argued that a doctrine of compelled speech in Singapore's Article 14 jurisprudence ought to incorporate a consideration of whether there was a compulsion of expression at all in the first place, whether there was a forced identification of the speaker with the compelled message, and whether the speaker had the practical ability to disavow the compelled message.

Returning full circle then to the POFMA and FICA, the question that arises is how these statutes would fare under the proposed doctrine of compelled speech. Applying the compelled speech doctrine proposed above, the first step is to consider whether the POFMA and FICA require compelled expression at all. In this regard, while POFMA correction directions and FICA must-carry directions would compel their recipients to display a specified message, these directions would not amount to prima facie impermissible compelled speech for the reason that the nature of the messages compelled under these statutes are purely factual in nature. A message that the material targeted by these directions is a false statement of fact or originates from a hostile information campaign would be purely factual in nature. For the same reason, technical assistance directions issued under the FICA to compel the recipient to provide information about the foreign connections behind certain internet accounts would not amount to compelled expression at all.¹⁰²

One might argue that the messages compelled by POFMA correction directions or FICA must-carry directions may not actually be purely factual, in the sense that they may be contentious – for example, one may contest whether the targeted material indeed conveyed a false statement of fact. The solution to this argument, however, is that the statutory framework of both the POFMA and FICA is premised on the compelled message indeed being a factual statement. If one disagrees with the messages being compelled, the appeal processes set out in both statutes can be utilised to set aside the received direction. If the appeal process succeeds, the challenged direction will be set aside, and there will be no issue with the compulsion of a contentious message. If the appeal process does not succeed, then the veracity of the message would indeed have been verified, and there will similarly be no issue with the compulsion of a contentious message. It is therefore important to observe that the determination of whether the compelled messages are indeed factual statements does not rest solely in the discretion of the relevant government authority – which would be more problematic – but is susceptible to final determination either by the courts under the POFMA or a reviewing tribunal under the FICA.¹⁰³

As such, a challenge to the POFMA and FICA, or to specific directions issued under these statutes, would likely fail at this first step in the compelled speech doctrine. In any case, even if one finds that the POFMA and FICA do indeed require prima facie impermissible compulsion of speech, a challenge will fail at the second and third steps of the proposed test. The POFMA and FICA do not force the speaker to identify with the compelled message – indeed, it is abundantly clear to all viewers that the compelled message is attributable to the government. Further, both statutes do not restrict recipients of directions from expressing their disagreement with the compelled message, as the Court of Appeal in Online Citizen noted in relation to the POFMA.¹⁰⁴ This would mean therefore that the POFMA and FICA would not trigger the proposed compelled speech doctrine at all – a conclusion that aligns with Online Citizen's preliminary reflections on the applicability of the compelled speech doctrine to the POFMA.¹⁰⁵

A reflection on the theoretical justifications for the compelled speech doctrine discussed earlier lends further support to this conclusion. Indeed, the majority of the concerns with compelled speech raised earlier are simply not implicated at all by the POFMA or FICA. The first concern was with the possibility that compelled speech would misrepresent the actual support for a certain position by forcing speakers to be publicly identified with the message being conveyed, thereby distorting the landscape of public debate. As the preceding paragraph has explained, it is plain that the POFMA and FICA would not trigger this concern.

The second concern was with the ability of compelled speech to overwhelm the marketplace of ideas in the direction of the government's preferred message by piggy-backing on private speakers and allowing the government to avoid the financial and logistical costs of conveying a message. It is suggested, however, that this concern carries much less weight in the online domain – the domain which the POFMA and FICA are principally concerned with. Indeed, there are negligible costs to promulgating messages through online fora. Accordingly, any advantage that can be gained by the government for its message through this piggy-backing effect is likely to be quite negligible. In any case, if it is plain to all viewers that the compelled message originates from the government, and if the requirement of POFMA and FICA directions is simply the appending of an additional mandatory message to the original content, it is difficult to see how there is any risk of the government's message insidiously overwhelming the marketplace of ideas. Such mandatory messages would simply be another contribution to the marketplace of ideas, reaching the same audience as the original content, giving viewers more information but ultimately leaving it up to them to make an informed decision.

The third concern was that compelled speech can have the effect of indirectly restricting speech by imposing a ‘tax’ on speech, or by crowding out the message which the speaker actually wants to convey where the amount of speech that a speech medium can support is limited. It is clear, however, that the POFMA and FICA would not trigger this concern. Indeed, POFMA correction directions or FICA must-carry directions do not limit their recipients’ speech at all, and given the unlimited capacity of the online domain, do not reduce the space within which the recipients can convey the message which they do want to express.

The fifth concern was that the compulsion of speech can exercise an insidious psychological effect on the speaker by indoctrinating the speaker to actually believe the compelled message, even if the speaker would not ordinarily have affirmed the message, thereby unfairly biasing the marketplace of ideas. It is suggested, however, that this concern would be most relevant where the compelled message is the expression of a substantive viewpoint – for example, the slogan ‘Live Free or Die’ or ‘Support Gay Marriage’. It is much less applicable to the compulsion of factual expressions like those which POFMA correction directions and FICA must-carry directions require. The salience of this concern is further reduced when one notes that the compelled speaker remains entirely free to vigorously disagree with the compelled message under the POFMA and FICA.

However, the POFMA and FICA do at first glance implicate the fourth concern raised above. The fourth concern was that compelled speech can burden a speaker's ability to communicate their own messages by diluting their persuasiveness, thereby distorting the marketplace of ideas. Indeed, it is quite clear that a compelled message stating that the recipient of a POFMA or FICA direction has stated a falsehood or is part of a hostile information campaign would diminish the persuasiveness of the speaker.

It is suggested, however, that this concern would be significantly less material in the context of the kind of speech targeted by both the POFMA and FICA. A condition for a POFMA correction direction to be issued is that a false statement of fact had been made, and one might justifiably think that diminishing the persuasiveness of the speaker in relation to the expression of such falsehoods would in fact be beneficial for the marketplace of ideas and the pursuit of truth.¹⁰⁶ A similar argument could be made in relation to FICA must-carry directions. Messages stemming from a foreign hostile information campaign – the target of FICA directions – would themselves have a distortive effect on the marketplace of ideas by seeking to tilt public discourse in a direction favourable to a foreign state actor, and an effort to diminish the persuasiveness of speakers in relation to such messages would in fact have a positive impact on the pursuit of truth. Accordingly, in view of the nature of the speech targeted by the POFMA and FICA, the messages compelled by these statutes implicate this fourth concern to a much reduced extent.

As such, at a theoretical level, the expressions which the POFMA and FICA can compel recipients of correction or must-carry directions to make would not implicate any of the truth-based theoretical justifications for the compelled speech doctrine. Indeed, such directions may even facilitate the pursuit of truth and redress distortions in the marketplace of ideas caused by falsehoods and foreign interference. The doctrinal conclusion arrived at earlier is therefore supported also from a theoretical standpoint.

Conclusion

In closing, this article has sought to make a contribution to an area of Singapore's free speech jurisprudence that has thus far been virtually unexamined – does the right to free speech encompass the right not to speak where one does not wish to do so? Capitalising on the observation that the POFMA and FICA raise the issue of whether compelled speech is constitutionally permissible, the article argued that there is a sound theoretical justification for accepting the doctrine of compelled speech in Singapore on the basis of the truth-based theoretical justifications for free speech that have been accepted in Singapore. Drawing inspiration from major common law jurisdictions which have accepted the doctrine of compelled speech, the article went further to propose the doctrinal framework by which compelled speech may be analysed in Singapore. Turning full circle back to the POFMA and FICA, the article has argued that neither of these statutes would fall foul of the compelled speech doctrine, even if the doctrine were to be accepted in Singapore. It is hoped that this article will be a useful contribution to a fascinating and yet under-explored area of constitutional law in Singapore.