A Marketplace for Honest Ideas

1 Introduction

The marketplace of ideas metaphor (MoIM) invites us to imagine a vigorous trade in thoughts with a bountiful selection of ideational wares being advertised and sampled by discerning customers. Perhaps Habermas’s beloved 18th century coffee houses, or the open street cries of a village market square, are brought to mind. The thrust of the metaphor is that it is best to let producers and recipients of ideas manage their own exchanges, just as buyers and sellers do in free markets. The MoIM and its inspiring imagery have long been used to argue for a free and unregulated sphere of public discourse to match the ideal of the free economic market. Acknowledging the ongoing widespread influence of the MoIM, I propose a reinterpretation of the orthodox understanding of the metaphor by focusing on the role of honest speech regulation in economic markets. This revised interpretation is illustrated through three legal cases which highlight the importance of market participants being able to rely on the honesty of one another’s representations. Drawing on the economics literature on information asymmetry, I identify the crucial role that honesty regulations play in the proper functioning of the economic market. Therefore, if the regulation of economic markets and public discourse are analogous, then this provides a prima facie case for regulating the honesty of speech in the latter. Support for this view from a different normative approach is identified in the work of Seana Shiffrin, who argues for deontological requirements – and in principle legal regulation – for honest speech. The upshot is a wholesale re-imagining of the MoIM to help break the sometimes seductive rhetorical appeal of an efficient deregulated market as somehow an analogy for a similarly productive deregulated public sphere of speech.

2 The Traditional Metaphor

The deployment of metaphor as a feature of argument inevitably involves a not insignificant degree of vagueness and an imaginative leap. This is exacerbated when the two things the metaphor equates are themselves highly complex, as is the case with the MoIM comparing the economic marketplace and the public sphere of speech and ideas. Both of these concepts contain a large range of potential internal distinctions: what counts as a market, what goods or services are under consideration, in what places, and at what times in history. The public sphere admits of even broader diversity, as discourse ranges from the scientific to the theological, and its format from court testimony to political pamphlets.^[1] One might therefore balk at the notion that the MoIM can mean anything at all, and it certainly does not lack for critics who argue that the analogy at its heart is misleading (Bambauer 2006; Blocher 2019; Brietzke 1997; Goldman and Cox 1996; Joo 2014; Sparrow and Goodin 2001). Nevertheless despite, or perhaps even because of, their lack of specificity metaphors can in turns both inspire and capture the imagination. The MoIM’s enduring influence is such that it continues to appear as a rhetorical point of reference across a variety of scholarship, ranging from the spread of academic knowledge (Delfanti 2021; Moorman et al. 2019) to public health experts (Smith and Wanless 2020) and especially in US First Amendment studies (Dryden 2018; Weinstein 2020). One recent analysis found a rapid increase in occurrences of the phrase the ‘market for ideas’ and its derivations from the latter half of the 20th century continuing well into the 21st century across journals in economics, sociology, and political science (Hodgson 2020, 1155–6). Given this enduring impact, what I set out to do here is to reframe the MoIM, so that it evokes a new image and understanding of speech regulation in both the economic and public spheres. With this shift in perspective I aim to encourage further work on how such regulation can be conceptualized and implemented. In light of the ambiguity inherent to metaphorical and analogical inquiry and the complex subject matters at hand, this analysis will necessarily come with a degree of imprecision, but I nevertheless hope to persuade the reader that my argument makes for a compelling retelling of the MoIM.

I begin with a simple reconstruction of the orthodox version of the MoIM, by drawing on the three sources which are regularly credited with inspiring its origin (Bezemek 2015; Blocher 2008; Brietzke 1997; Ingber 1984) – Milton’s Areopagitica, John Stuart Mill’s chapter “Of the Liberty of Thought and Discussion” from On Liberty, and Justice Oliver Wendell Holmes Jr.’s dissent in the case Abrams v United States. Returning to the ‘origins’ of the metaphor in this manner does not necessarily tell us how modern interlocutors might use the term. However, despite their differences across time, location, purpose and intended audience, I believe that we can learn something about its essential elements.

My main, hopefully uncontroversial, contention is that the core of their arguments stands in opposition to the overreach of government power and the deleterious consequences of outlawing undesired speech.^[2] All three arguments advanced at least two claims: first, that we should be aware of our own fallibility in potentially misidentifying truth and falsehood, and second that truth will ultimately prevail over falsehood without explicit prohibition or censure of false views.^[3] All three authors anticipate, and are primarily concerned with, measures which prohibit views on the grounds of their perceived noxiousness or incorrectness. Despite their differences,^[4] one can identify the genesis of the MoIM. Regulation of speech is to be opposed because, among other things, it undermines the presumed goal of discourse – truth. On this model, opinions are presumed to be opposed by – and to compete with – contrary views, with truth being the outcome of this oppositional struggle. The marketplace element of the metaphor is hinted at when Milton compares thought and ideas to commercial trade in order to argue against applying commerce-like regulations to them:

Truth and understanding are not such wares as to be monopoliz’d and traded in by tickets and statutes, and standards. We must not think to make a staple commodity of all the knowledge in the Land, to mark and licence it like our broad cloath, and our wooll packs. (Milton [1664] 1875, 33)

This comparison becomes an explicitly market-based competition metaphor in the opinion of Oliver Wendell Holmes, who, unlike Milton, compares ideas to unregulated markets:

The ultimate good desired is better reached by free trade in ideas – that the best test of truth is the power of the thought to get itself accepted in the competition of the market. (Abrams v United States)

The focus of the argument in all three authors’ work is on speech regulation and in doing so they all emphasise the importance of ‘competition’ between ideas. Holmes is most explicit in making competitive – by being unregulated – markets the ideal comparator for speech. Drawing on this final iteration of the MoIM, I propose that the MoIM relies on the following basic argument:

1. The market is an unregulated space.

2. Absence of regulation is key to the success of the market.

3. The dynamics of the public forum and its exchange of ideas and speech are, in some relevant sense, analogous to the market.

4. The public forum should therefore also be free from regulation.

P1 and P2 are implied by the metaphor’s reference to the marketplace and are made most explicit in Holmes’s dissent. They tie together the three sources’ common opposition to regulatory speech prohibitions to the metaphor’s appeal to the apparent success of the free marketplace as a comparator (Volokh 2011). P3 is the crux of the analogic relationship implied by the MoIM. It is alluded to in the sources’ references to competition between ideas with truth winning out, but it has also perhaps been the element of the MoIM most subjected to sustained academic criticism (Bambauer 2006; Blocher 2019; Brietzke 1997; Goldman and Cox 1996; Joo 2014; Sparrow and Goodin 2001). Issues raised range from the lack of equivalent price discovery mechanisms in public discourse to how ideas lack the relevant properties of goods or services to different cognitive biases in how we consume ideas as opposed to goods or services. I certainly do not mean to dismiss these criticisms; however, I contend that with reframing we can see a more specific manner in which this P3 analogic relationship might hold. In the remainder of this paper I wish to turn our focus to a new way of understanding P1 and P2.

At one level it is obvious that modern markets rely for their existence on a series of regulatory institutions – property rights, independent legal dispute resolution, state-backed currencies, corporate personhood, stock exchanges – and less formalised ones such as social norms (Vogel 2018). The analogous institutions for speech are those which facilitate communication, ranging from language itself to modern communications technology.^[5] What the MoIM is generally taken to refer to, however, is not these basic structural features, but the absence of substantive regulation of the transactions themselves.^[6] Instead of thinking about market regulation simpliciter I suggest we focus on how speech in the marketplace is regulated, and in particular dishonest or misleading speech. This brings us closer to comparing like with like in the course of P3 and therefore strengthens the potential for analogic comparison.

3 Honest Speech and Three Illustrative Cases

Honest speech regulations place limits on the extent to which actors can mislead one another in the marketplace. In particular, untrue or misleading representations intended to influence the economic behaviour of others are prohibited (Fraud Act 2006, section 2). There are, of course, granular questions regarding just how candid actors are required to be in order to fulfil their obligations not to mislead. For example, publicly listed companies will have highly specific disclosure requirements, whereas under other commercial circumstances a degree of ‘bluffing’ to befuddle other participants might be a legitimate option (Carr 1968; Khorasanee 2024b). Notwithstanding these granular differences, the overarching principle is clear that speech in the economic context that is designed to mislead or that is intentionally dishonest is prohibited. In this section I set out three examples of the different ways in which these regulations work in practice. They are not intended to provide a systematic breakdown or classification of honesty regulations, but instead to provide readers with a sense of the different ways and contexts in which they can occur. I will then go on to explain the economic justification of such regulations in order to draw out how such explanations might be developed analogically for the public sphere.

The first is Gordon v Selico Ltd. This case concerned the sale of a long lease on a flat. After their purchase the claimants discovered that the flat had a serious case of dry rot and sued the seller and their agent. The common law rule in England and Wales is caveat emptor – when the buyer has no explicit or implied warranties as to quality from the seller they should beware.^[7] In this case the buyers had not requested a warranty against dry rot from the seller. They had instructed a surveyor to inspect the property before they purchased it, but the surveyor’s cursory inspection for dry rot – lifting a single floorboard – did not catch it. The lack of a warranty would have defeated the buyer’s claim but for the fact that prior to the inspection the defendants had had the flat painted – deliberately covering up the dry rot. The Court found this to have constituted a fraudulent deception of the buyers, despite the fact that the buyers could have requested warranties regarding dry rot or discovered the dry rot upon closer inspection before completing the purchase. In this instance the seller communicated deception through their silence with respect to a salient matter – a deception facilitated by their actions in painting over the rot.

A conclusion some might draw from Gordon is that regulation of misleading selling is a way to protect individual consumers. However, the actual law and how it is applied demonstrates that this is not the case. There is a whole legal regime designed specifically to provide protections for individual consumers,^[8] but the general protections against dishonest misrepresentations apply to all market participants. The most obvious evidence for this is the fact that powerful companies also avail themselves of these protections. One example is the controversy following the purchase of the company Autonomy by Hewlett Packard (HP). HP purchased Autonomy in 2011 for USD11 billion, but subsequently wrote off USD8.8 billion of the deal’s value, alleging that certain of Autonomy’s directors – who were also significant shareholders – had fraudulently misrepresented its value to them (Nuttal et al. 2012). HP complained specifically of various accounting measures used by Autonomy which HP argued inflated Autonomy’s apparent value. A slew of litigation followed, including a criminal conviction of Autonomy’s chief financial officer, Sushovan Hussain (Croft 2018), and a civil finding of fraud in the UK against both Autonomy’s chief executive officer, Dr Mike Lynch, and Hussain (ACL v Lynch). The principal lesson here is that all market participants, big and small, are entitled to not be lied to by sellers. As in the Gordon case above, the fact that HP as the buyer had the benefit of its own well-paid advisers and carried out its own due diligence was no barrier to claiming against deceit. While parties may have stronger or weaker bargaining positions and can therefore extract various concessions from one another in how they structure their deals – once the matter crosses into one of deception then parties have recourse to state machinery via the courts to redress the wrong which has occurred.

Finally, regulation to preserve honest speech in marketplaces extends beyond direct purchase and sale transactions. In a now infamous incident, on 7 August 2018 Elon Musk – CEO of publicly traded company Tesla, Inc – tweeted “Am considering taking Tesla private at $420 [per share]. Funding secured.” This suggested a price had been agreed with private investors to buy the company from its publicly held shareholders at a 22 percent premium over the ex ante stock price of the company (Massoudi, Waters, and Fontanella-Khan 2018). Tesla’s public share price increased markedly as a result. However, there were in fact no private buyers who had committed to purchasing the shares and Musk’s tweet therefore at best constituted speculative hypothesising.^[9] As a result, the Securities and Exchange Commission (SEC) – the US regulator responsible for maintaining fair and orderly public markets – sued Musk for his ‘false and misleading’ statement. The suit was settled for fines of USD20 million each applied to Musk and Tesla respectively, Musk stepping down as chair of the Tesla board of directors, and Musk agreeing to a screening process for his future communications.^[10] The principle of this case is that dishonesty can negatively impact markets without the speaker being party to any identifiable seller-buyer relationship. Neither Musk nor Tesla were accused of selling shares as a result of – or directly profiting from – this tweet. Nevertheless, the intense share price volatility which followed his statement indicated that many others did buy and sell Tesla shares in response to the information he had disclosed. This case is interesting because the grounds for the SEC’s claim was the harm to the proper functioning of the market itself. While individual investors could, and did, avail themselves of the opportunity to sue Musk for their individual losses^[11] – the SEC’s action was one designed to punish a more diffuse harm to trust in the market itself and its dependence on the honesty of relevant statements and disclosures.

4 Honesty Regulations and Information Asymmetry

The foregoing examples illustrate just some of the different ways in which markets are regulated for honesty, whether by private lawsuits or civil or criminal regulatory enforcement actions. The question which necessarily follows is why we have this type of regulation. One obvious explanation, mirroring the above arguments regarding free speech by Milton, Mill, and Holmes above, is that this is about ensuring the discovery of the truth for its own sake. This explanation has some intuitive plausibility, as dishonest and misleading communications are necessarily designed to induce false understanding in their recipients. However, such regulations cover a variety of situations where the truth will become known without regulation. In the Gordon and Autonomy cases the buyers discovered the issues themselves shortly after their ill-fated purchases.^[12] There are also a variety of alternative mechanisms to help ensure that the truth will come out eventually. Free discussion allows people to write reviews of goods and sellers, and so purchasers may often have a wealth of information to hand relating to both products and their sellers before they purchase them.^[13] In cases such as property or company acquisitions buyers can, and do, instruct their own experts to verify the condition of goods before they buy them. Therefore, while honesty requirements may cover some cases where the truth may not be discovered in the normal course of the market, they also cover many instances where truth discovery is to be expected through other means. Truth discovery is therefore not a sufficient explanation for honesty regulations.

Rather than truth per se, it is more appropriate to see the purpose of such regulation as an attempt to solve the problem of information asymmetry. The economics literature has long recognised that information asymmetry between market participants can cause market failure (Mas-Colell, Whinston, and Green 1995, 436–76). Akerlof’s (1970) famous paper illustrated that in second-hand car markets without credible ex ante ways to differentiate good from bad products prices would fall to accommodate this uncertainty and sellers of good quality cars would be effectively driven from the market as a result. Conversely, given cheap and credible means to signal quality goods a different cascading effect can occur – known as unravelling – whereby sellers are incentivised to display their quality credentials and those who do not are swiftly identified as being of lower quality (Viscusi 1978).^[14] Suitably enforced honesty regimes are key to giving market participants the confidence to transact (Darke and Ritchie 2007; Grossman 1981; Milgrom 1981). Without such regulations distrust and uncertainty act as a drag on market activity.

The economics literature cited above is particularly focused on seller disclosure; following Akerlof it is now well understood that requiring honesty facilitates buyer-seller relationships. The example of Musk provides a slightly different example. In this instance the SEC fined Musk even though he and Tesla were not directly transacting with people who bought or sold Tesla shares on the basis of his statements. What he did do was cause those persons to trade with others on the basis of misinformation. Within the EU framework on financial market abuse this offence is known as ‘dissemination’. Under Article 12(1)(c) of the Market Abuse Regulations this offence is constituted by:

disseminating information through the media, including the internet, or by any other means, which gives, or is likely to give, false or misleading signals as to the supply of, demand for, or price of, a financial instrument … including the dissemination of rumours, where the person who made the dissemination knew, or ought to have known, that the information was false or misleading.

The issue with false and misleading statements such as Musk’s is that they pollute the discursive pathways which enable people to communicate and disseminate information in reliable ways. Individuals in financial markets need to incorporate information from a range of sources, including statements by the issuers of the securities they want to trade in. If market participants are faced with the prospect of dishonest or misleading information this hinders their ability and willingness to trade as they have to account for the risk that they are being misled. Across all three cases honesty regulations act as a disincentive to engage in deceptive speech and thus allow for parties to trust and rely on one another’s statements and representations with less reserve.^[15]

5 An Updated Analogy

The foregoing analysis has focused on the first two claims in the MoIM. Drawing on the above legal cases and economics literature to revise and add more specificity to the first two steps in the traditional MoIM produces the following:

1. The market is an honesty-regulated space.

2. This honesty regulation is conducive to the success of the market.

3. The dynamics of the public forum and its exchange of ideas and speech are, in some relevant sense, analogous to the market.

4. The public forum should therefore also be subject to honesty regulation.

Both Akerlof’s information asymmetry findings and the offence of dissemination suggest market failure occurs because individuals cannot trust or rely on information they are receiving from other actors. Similar dynamics can be observed in the public sphere, albeit with significantly higher stakes. Our understanding of the world itself – ranging from global affairs to scientific discoveries to local news – rests in large part on the testimony of others. What issues we pay attention to, and how we treat them, is heavily influenced by what we hear from others. This has important consequences if beliefs based on falsehoods are translated into specific policies or, alternatively, styles of politics which undermine public trust in one another and civic institutions (Hundley 2017; Schroeder 2018; Waldman 2017). Of course, this is hardly an entirely novel concern. It has long been suggested that the case for government intervention in the market for ideas is stronger than in the economic market given people’s superior adeptness at assessing products they consume compared to evaluating social, political and economic policy (Coase 1974, 389–90; Schumpeter [1942] 2010, 235–46).

The revised MoIM proposed here therefore functions to draw our attention to the common reliance on trustworthy communication in both economic and public spheres. The main function of this paper has been to reinterpret the MoIM in order to uncover this analogic relationship. There are a variety of directions in which subsequent research could flow – for example, analysing subspheres such as speech in courtrooms, scientific journals, or political speech, the practical challenges of how such regulations might function in the public sphere or, contra my approach, identifying potential disanalogies between the two spheres with respect to honesty regulation. The revised MoIM now directs our attention to a new conceptual space in comparing how certain types of public speech regulation might profitably mirror those in the economic sphere. In the spirit of identifying productive similarities between otherwise differing domains, I will indicate one potentially auspicious analogue in the study of free speech – the work of Shiffrin and her thinker-based account of free speech protection (2014). Although the economics literature on information asymmetry works within a normative paradigm of welfarism and pareto equilibria and Shiffrin’s view is explicitly deontological, the two lead us towards surprisingly similar destinations in their respective fields. In both instances it is imperative that we, as separate human beings, are capable of overcoming the communicative barriers between us in a reliable manner in order to suitably coordinate our behaviour.

Shiffrin argues that our mental separateness from one another requires sincere communication if it is to be overcome. Reducing the informational barriers between us through reliable mutual communication enables complex social organisation, which in turn is necessary to enable us to pursue our respective moral duties. This coordination ranges from our direct interpersonal relationships to cooperative social endeavours such as scientific inquiry or democratic governance. Shiffrin therefore argues that sincerity is a deontological requirement and lying should therefore be morally prohibited and, in principle, should be the subject of legal regulation, notwithstanding freedom of speech concerns (Shiffrin 2014, 116–56). Although Shiffrin’s case is deontological and not consequentialist – she is concerned that lying undermines the rational basis for trust (Shiffrin 2019, 557–8) – considering the consequences of widescale lying may help drive her point home. For example, authoritarian disinformation campaigns which through sheer volume of lies and half-truths try to induce a sense of ‘informational nihilism’ whereby individuals come to doubt everything they hear.^[16] Were such strategies to prevail, society would come to resemble Oceania from Orwell’s novel Nineteen Eighty-Four – each individual fragmented and cut off from any real way to meaningfully relate to one another or to situate themselves with respect to real events happening in the world.^[17] This is, of course, an extreme fictional case.^[18] Yet it illustrates the principle that a proliferation of lying in the public sphere is ultimately a threat to our ability to meaningfully relate to our fellow citizens and to both maintain solidarity and cooperation with one another. Less extreme examples of the corrosive nature of widespread dishonest speech, particularly with respect to politics and civic cohesion, abound in our contemporary societies – facilitated by modern technology-enabled methods of communication (McIntyre 2018; O’Connor and Weatherall 2019; Woolley 2020). Just as the apparatus of state-enforced honesty regulations came about as markets moved from more primitive to modern types, it seems plausible that the modern technology-enabled public square may also require revised rules of conduct.

A pertinent question at this stage is what the regulation of such revised rules of conduct would look like.^[19] For example, what substantive standards would agents be held to, who would be targeted by the regulations, and what particular mechanisms would be used to enforce said regulations. Given this paper’s emphasis on reinterpreting the MoIM rather than developing regulatory proposals, what follows will necessarily be quite brief and highly tentative. The first and most important point to re-emphasise is that the public sphere includes a vast array of different contexts and circumstances. Therefore it will likely be the case that any potential regulation would have to vary substantially due to particular vulnerabilities, the presence of self-correcting mechanisms, or other idiosyncratic factors.^[20]

When it comes to the relevant substantive standards, it is notable that Shiffrin is careful to specifically limit her arguments to sincerity and lying – understood as asserting a proposition one knowingly does not believe which is intended for interlocutors to treat as an accurate representation of the speaker’s belief (Shiffrin 2014, 12, 2019, 579–80).^[21] This would potentially leave untouched cases where agents are reckless as to the truth of their statements or disseminate falsehoods in good faith, perhaps relying on trust in the sources who conveyed it to them. The focus suggested by Shiffrin is narrower than economic regulation but one must recall that market transactions usually form relatively discrete elements of an agent’s activities, whereas discourse in the public sphere potentially encompasses everything from messaging with friends and idle chatter at the school gates to political speech. Demanding self-scrutiny of all public statements risks an unwelcome chilling effect on our ability to carry on the interactions underpinning our everyday lives (Mill [1859] 1974, ch. 1).^[22] Therefore, both the targets of regulation and the scrutiny levels of any such standards should vary according to the reach and potential impact of statements. For example, see how the UK’s media regulator Ofcom applies particular standards of regulation to broadcasts regarding elections and referendums given the particularly high stakes of such communications for democracy (Ofcom 2024). There cannot be a one-size fits all model here, but prominence and reach of the speaker as well as sensitivity and importance of the context and content will all be relevant features which must be accounted for in any such regulations. As the three example legal cases hint at, the mechanisms for enforcing any such regulations would also likely vary due to particular circumstances and contexts. The Gordon case involved a straightforward civil case, Autonomy involved both civil and criminal penalties, and the driving force behind the Tesla settlement was a regulator’s intervention. It is therefore perhaps best to think of the market regulatory framework for honesty as a patchwork of different mechanisms which together help maintain baseline confidence and trust in one another’s representations. Given the even wider diversity of communications and actors in the public sphere, the MoIM would suggest that, while the law and courts would inevitably feature as part of regulation, we should expect the specific mechanics, actors bringing and being the subject of claims, and types of penalties all to vary. Of particular concern would be the risk of government overreach or malicious action,^[23] which would be attenuated by having a varied range of actors able to bring regulatory action through a diverse range of channels.

6 Conclusions

One should not overstate what one can argue for by metaphor, the purpose of which is, after all, to highlight similarities between different contexts (Best 1984, 165). Critics of the orthodox approach to the MoIM have focused their attention on the ways in which the economic marketplace and the public forum are not analogous. My updated version of the MoIM does not rely on weaker elements of the analogy, for example, trying to claim that market dynamics of supply and demand or pricing function in some kind of analogous way in the public forum. Instead, my revised approach to the MoIM focuses on a clear commonality between both domains – the need for honest communication for participants to achieve the goods internal to both spheres. In doing so, my revised MoIM helps shift our perspective away from the idea that there is some kind of ‘wild west’ unregulated market sphere to which public speech can be compared as an argument against regulation. On the contrary, along the particular axis of honest speech, markets are more stringently regulated than the public sphere, and for reasons which seem broadly applicable across both domains.^[24] By resetting our perspective on the influential MoIM, I aim to help open the space for further work on how this type of regulation could be conceptualised and implemented.

In conclusion, in this paper I began by drawing on Milton, Mill, and Holmes to outline an orthodox understanding of the MoIM focused on the substantive regulation of economic markets and speech in the public sphere. I then narrowed the focus from regulation in general to regulation of speech, and in particular that of honest speech. Drawing on three legal cases as illustrative examples and the relevant economics literature, I identified that the best justification for honesty regulations in the market is reliably reducing information asymmetry between market participants to promote better outcomes. Incorporating these insights into the MoIM produces a revised metaphor that suggests the public sphere would also benefit from similar regulation. In particular, the work of Shiffrin highlights that fulfilling our duties as individuals relies on collective human endeavours which depend on reliable channels of communication, which in turn are threatened by dishonesty. Notwithstanding the disclaimers as to the limitations of arguing by analogy – and the practical challenges of regulating public speech for honesty – the conclusion suggested by the revised MoIM is that honesty regulation could prove beneficial for the public sphere as it does in economic markets.