Keywords

Introductory Comments

We now examine how false election information violates democratic values and explore the extent to which and the manner in which false election information undermines the legitimacy of Australian (and other authentic) elections. We also argue that, because of the perverse incentives to produce false election information coupled with the significant social costs it entails, the problem should be approached as a collective action problem rather than as an issue of individualised rights. It therefore calls for state action to resolve it. This is consistent with public attitudes: the fact that there is very strong support among the public, politicians and the legal profession for regulation in Australia suggests that the majority of Australians also regard false election information as a collective action problem in need of state action. Exploring the democratic justification for why false election communication should be regulated illuminates the risks of over-regulating, on the one hand, and failing to regulate, on the other.

Democratic Criteria

A standard normative framework for assessing any practice or regulation around the delegation of power in democracy is Robert Dahl’s classic theory of procedural legitimacy. Dahl formulates criteria for determining the legitimacy of processes that generate the decisive moment of representative democracy: elections. According to Dahl’s account, certain conditions must be met in evaluating election procedures and he identifies five such conditions:

  1. 1.

    Political equality (i.e. that every vote should count for one and only one)

  2. 2.

    Effective participation

  3. 3.

    Enlightened understanding

  4. 4.

    Final control of the agenda by the demos (i.e. that the procedures should deliver an outcome that confirms the people as sovereign) and

  5. 5.

    Inclusiveness (i.e. that the procedures are universalistic and are not unjustifiably ‘exclusivistic’. Universal suffrage is key here).

If a system’s processes can satisfy these five criteria it can be regarded as a full ‘procedural democracy’ and the outcome generated by the procedure (elections) can be seen as legitimate (Dahl, 1979). It will therefore be accepted as the ‘correct’ outcome. For the purposes of the present discussion, criteria 2 and 3 are most relevant and so we focus on those.

‘Effective Participation’

On this criterion, procedures are judged according to ‘the adequacy of the opportunities they provide for, and the relative costs they impose on, expression and participation by the demos in making binding decisions’ (Dahl, 1979: 102–103). This means that electoral participation should be relatively easy and not costly to individual voters. In particular, information costs should be low, and everyone should be able to be equally informed. When false information is at play, this capacity is seriously eroded because the information costs of becoming informed and knowing who to vote for go up. This is because much more work is required in attempting to sift the wheat from the chaff or the facts from the falsehoods. We need reasonably accurate information to make our voting decision, but that information is elusive when the market is flooded with inaccurate information. This dynamic disproportionately affects the poor and less educated because the socially disadvantaged face more challenges in gathering information about politics and voting and are typically less ‘knowledgeable about politics’. This impacts negatively on their electoral participation (Gallego, 2010) as well as their attitudes to democracy in general. This, in turn, further delegitimises the election procedures and outcomes because it also violates the ‘political equality’ criterion mentioned above (Hill in Brennan & Hill, 2014).

‘Enlightened Understanding’

Enlightened understanding is closely related to the ‘effective participation’ criterion and requires that ‘in order to express his or her preferences accurately, each citizen ought to have adequate and equal opportunities for discovering and validating…what his or her preferences are on the matter to be decided’ (Dahl, 1979: 98–108). According to this criterion, every citizen should be attentive and engaged and it is unacceptable to ‘cut off’ or suppress information that could affect decisions made, or afford some citizens easier access to information of crucial importance. This makes it sound as though we should tolerate false information at election time and not ‘cut it off’. But the situation is actually the opposite because with false election information, the false information is doing the cutting off of the accurate information. It is clogging the free speech condition that needs to exist in order for people to make their voting decision.

The Free Speech Condition

What exactly is the ‘free speech condition’? We, along with many others, think it might be better to think of free speech, not as an individualised act exercised as a purely individual right, but as a social condition that we all inhabit and own and to which we are all subject and to which we also contribute. That condition is dependent on other things, including its proper use; imagine how quickly the condition would deteriorate in a universe full of irresponsible and untruthful speech. Pretty soon no one would bother to speak or listen and speech itself would lose all value. So, the free speech condition is synonymous with a kind of efficiency in the ‘marketplace of ideas’. Such efficiency does not simply refer to free and unfettered flow, since such a flow could lead to a speech universe in which anything goes (including ‘alternative facts’) but in which speech eventually becomes meaningless.

Free speech only exists when certain social and political conditions are present to enable the speech to flow freely, enabling different ideas to compete with each other and thereby promote the truth. This involves commitments like resolving never to mislead others intentionally and treating others with equal respect so that their good faith speech can be heard and heeded (Hill, 2007). It is worth emphasising that the marketplace of ideas underlying the social condition does not have to be completely unregulated in order to serve this condition; the term ‘market-place’ can be misleading because it conjures up the image of ‘laissez-fairism’. In fact, a completely unregulated market will introduce distortions that are hurtful to both the free flow and the truth. Regulation of the market is therefore, in many cases, warranted: for example, it is justified to verify and police the content of articles published in scientific journals in order to serve the ‘truth’ while laws requiring that poor people be offered legal aid in criminal proceedings seek to address the speech power asymmetries to which an unregulated market for legal services gives rise (Goldman & Cox, 1996).

So, the social condition of free speech is not anarchic; rather, it is a balanced environment that offers, not only negative liberties like toleration of dissent and freedom of religious belief but positive freedoms that promote the condition. These can be brought about by legislation and might include the provision and defence of rights like the right to an education so that everyone can learn to communicate, hear and understand each other effectively or the right to equal employment opportunities and other institutions and laws that seek to address the social and economic barriers to equal speech access. The free speech condition is also dependent upon a culture of tolerance and respect, and this is also something that can also be enhanced by laws.

If we adopt the ‘free speech as a social condition’ approach, then any speech act that fails to promote, or which negatively affects the efficiency of the marketplace of ideas will fail to qualify as an act of free speech (Braddon-Mitchell & West, 2004: 460). It is therefore unworthy of the same protection as speech that promotes the overall condition. Lies and hate speech, for example, will generally fail to meet the test. But so too does false election information, especially if it comes from leaders who speak from extraordinarily privileged speech platforms and therefore have a vastly magnified capacity to clog up the speech condition and distort election processes. This power is especially amplified in the case of political incumbents in Australia; not only is their speech protected by parliamentary privilege within the house, but politicians are exempt from laws which constrain the rest of us, including the Spam Act 2003 (Cth), the Privacy Act 1998 (Cth) and the Do Not Call Register. A case that underlined these extensive freedoms occurred in 2021 when the Hon. Craig Kelly repeatedly bombarded millions of Australians with spam election messaging. Although Kelly had been temporarily suspended from a number of social media sites due to his posting of COVID-19 disinformation, not all platforms had done so because the main mechanism for regulating digital speech—the DIGI code of conduct (see below)—is voluntary. Therefore, Kelly’s actions were perfectly legal and the relevant regulator (ACMA) was powerless to act.

Markets as Regulators

It is worth pausing at this point to problematise the idea of markets as the natural (or even effective) regulators of bad information. For many, ‘fake news’ is a normal part of the marketplace of ideas that liberal democratic society is supposed to tolerate. On this view, consumer demand will cause the best ideas to ‘rise to the top’ while bad or faulty ideas will be rejected by rational consumers. However, there are major problems with this optimistic faith in information markets: first, as we have shown, consumer demand unconsciously tends towards the consumption of news that confirms existing prejudices therefore there are perverse incentives within the market to keep producing it; second, consumers are not always equipped to sift the information wheat from the chaff. A recent study undertaken in the United States of over 8000 subjects found that those ‘who are most confident’ of their ability to distinguish between legitimate and ‘fake news’ were also the most likely to both believe, consume and share it. The study authors also rated over 70% of subjects ‘overconfident’ (Grover, 2021). The idea that consumers will rationalistically regulate the market is therefore based on the faulty premises that consumers (a) have the time and energy to sift and assess every piece of information and (b) can and will always act in the rational manner predicted. Historically a source of security against threats to democracy, in the digital age, the marketplace of speech has become ‘a principle weaponised against the ideals from which it sprang’ (Stone & Schauer, 2021).

Market processes are therefore ineffective sifters of inaccurate political speech and in many cases are even the source of the problem (given that there is money to be made in disinformation). In any case, it is debatable whether markets should be expected to regulate facts in the first place: after all, as Ari Waldman points out ‘[t]he marketplace of ideas was always meant to be a marketplace of ideas, not facts. There is no marketplace in facts. Indeed, no area of law permits a market in facts… Facts, like gravity, are not up for debate’ (Waldman, 2018: 869). In other words, there is no such a thing as ‘alternative facts’ and it is incoherent to insist that false or inaccurate information can ever be a fact and therefore deserving of the same legal protection.

There are, and should always be, limitations on what can be said in the public domain because there are other competing values at stake such as the right to privacy, equality, dignity and to be free from speech that damages our social and civil standing. Markets are not particularly adept at respecting such values and even when market actors have signed up to codes of conduct in relation to them, they still fail.

Regulating false election speech by more definitive means is therefore warranted; this approach is hardly unprecedented because no society has ever existed where speech has not been limited to some extent in order to protect individuals and the public interest (Fish, 1994). Blackmail, coercive threats, invasion of privacy, sexual or racial intimidation and harassment, conspiracy, extortion, libel, discriminatory job advertisements, perjury, fraud and misrepresentation are all areas where we would probably agree that speech cannot always claim exemption from legal regulation or redress. So, despite how crucial free-flowing political speech is to the functioning of democracies, it cannot be assumed to be exempt from any constraint. Indeed, some constraint is needed to ensure authentic free flow.

But this is tricky to regulate. What if, for example, votes and entire elections really are being stolen? It is vital that citizens can draw attention to that publicly if it is really happening and so should any good faith whistle-blowing be protected. Therefore, any remedial regulations for ‘bad’ election speech will need to be rather conservative with high thresholds for proof and inbuilt redress mechanisms and they would have to carry with them safeguards and a high degree of epistemological scepticism. The regulator would also have to be trusted. There also needs to be a strong justification for regulation such as we have sought to provide here by arguing (a) that false election information inflicts significant harm and (b) that the market not only fails to remedy the situation but is often the cause of it.

Public Support

In Australia TIPA legislation has a good chance of meeting with bipartisan support, as occurred in South Australia and the ACT (Renwick & Palese, 2019). Further, a substantial majority of Australians want it to be enacted. Polling conducted by The Australia Institute found that 84 per cent of Australians support TIPA laws; significantly, such support is independent of party preference, with only 0.5 per cent separating Coalition and Labor voters on the issue (Browne, 2019a). Subsequent to its popular poll, the Australia Institute conducted an online petition for stronger TIPA laws; it received support from 29 prominent Australians including Supreme Court judges, politicians of various stripes and business leaders (The Australia Institute, 2020).1 These sentiments are corroborated by a national ReachTEL poll conducted after the 2016 federal election which found that 87.7 per cent of Australian respondents want tougher TIPA laws (Australia Institute, 2016). Therefore a key justification for legal regulation is that there is broad support for it at all levels.

A State Duty in an Atypical Setting

A final justification for regulating false election speech is found in the fact that voting is compulsory in Australia. Although a clean information environment for elections is, for any democratic government, a matter that falls squarely within the purview of government responsibility, that duty is magnified in the Australian setting. Requiring citizens to vote places a special duty on the state to ensure that the imposed duty does not place an unreasonable burden on citizens; this is why, for example, Australians are exempt from the requirement to vote if they live more than 8 km from the nearest polling station or if some circumstance beyond their control makes it difficult for them to vote. Voting in Australia is a comparatively painless affair for voters because the Australian state meets almost all of the opportunity and transaction costs involved. The Australian state (via its state and commonwealth electoral commissions) assumes a high degree of responsibility for making feasible what it legally requires of electors. Our electoral commissions go to extraordinary lengths in order to accommodate ageing and immobile people, people experiencing homelessness, those living in remote regions, incarcerated citizens, people who have a disability, are ill or infirm, housebound, living abroad, approaching maternity, hospitalised, have literacy and/or numeracy problems or are from a culturally and linguistically diverse background. There are special provisions for ‘silent enrolment’ (for those who believe that having their name on a public roll endangers either themselves or their families) and itinerant enrolment (‘for people experiencing homelessness, or people who travel constantly and have no permanent fixed address’). In any federal election hundreds of mobile teams visit special hospital locations, remote outback locations and prisons; there will be easily accessible pre-poll voting centres and overseas polling places to which tonnes of election-related and staff training material will be air-freighted immediately prior to polling. To ease access, election day is always on a Saturday and Australian electoral commissions are extremely proactive in ensuring that everyone is enrolled to vote. They also provide electoral education; offer absent voting, early voting and postal voting; and ensure that polling booths are numerous and close at hand. All of this means that electors don’t have to sacrifice much in terms of cost and lost opportunities for work or leisure in order to vote. No one in Australia, no matter how disadvantaged, isolated or immobile, is expected to meet the potentially high transaction and opportunity costs of voting which citizens in most other settings have to put up with because voting is voluntary in most other settings; apparently, voluntary voting states do not consider they are under the same obligation to make voting this easy (Hill, 2002, 2010).

False election information is now a major barrier to ease of voting because it pushes up the information costs of voting, making it harder and harder to know who is the best candidate to trust to represent our interests. Yet, however strong is the need for a legal solution for reducing these costs, we are mindful that great care needs to be taken in order to prevent regulatory measures around political speech from damaging the democratic conditions they are supposed to protect. After all, we know that some governments misuse anti-disinformation laws to suppress opposition. But there does seem to be a case for them. Notably, the Australian public agrees.

All such regulation encounters what may be referred to as the ‘freedom-fairness’ trade-off and the precise calculus of this trade-off varies considerably between established liberal democracies. But by tipping the scales towards ‘fairness’ (Tham & Ewing, 2021), TIPA laws, as constraints on political speech, can be designed to find this balance. When considering such laws in light of the implied freedom of political communication, the Australian High Court has struck a unique balance between speech ‘libertarianism’ and ‘fairness’ and is unusually accommodating to TIPA-type speech regulation, so long as certain requirements are met. In Chapter 5 we show how this plays out in Australia and explain why it is an especially congenial setting for the accommodation of TIPA laws. But before doing so, we canvass how the problem is currently being managed here.

Note

  1. 1.

    These figures included Dr. John Hewson, Cheryl Kernot and Michael Beahan; former Supreme Court judges, The Hon. Anthony Whealy QC, The Hon Paul Stein AM QC and The Hon David Harper AM QC. See An Open Letter to the Parliament of Australia calling for Truth in Political Advertising 2020, The Australia Institute, viewed 22 November 2021,  https://australiainstitute.org.au/wp-content/uploads/2020/12/Open-Letter-Truth-in-Political-Advertising-WEB.pdf.