Towards Unfair Political Practices Law: Learning lessons from the regulation of unfair commercial practices for online political advertising

Prof. Dr. Natali Helberger
Tom Dobber
Prof. Dr. Claes de Vreese

Abstract

Online political advertising operates in a tense forcefield between political and commercial elements and thus presents regulators with a difficult conundrum: because online political advertising is political rather than commercial speech, it is destined to follow a different regulatory tradition than commercial advertising. And yet many of the tools used, players involved and concerns triggered by modern online political advertising strategies very much resemble the tools, players and concerns in online commercial targeting. Commercial advertising is subject to consumer law and unfair advertising regulation, including rules about unfair commercial practices. Unfair commercial practices law and other rules about commercial advertising, however, are explicitly not applicable to forms of non-commercial political or ideological advertising. An important reason why this is so is the different level of protection of political and commercial speech under fundamental rights law standards. And yet with the ongoing commercial turn in advertising, the traditional division between forms of commercial and political advertising is no longer that self-evident. Also, it cannot be denied that commercial advertising law has a long tradition of thinking of where and how to draw the line between lawful advertising and unlawful persuasion through withholding or misleading consumers about the information they need to take informed decisions, or abusing superior knowledge, exerting undue psychological pressure and engaging in other forms of unfair behaviour. The question this article explores is whether there are lessons to be learned from the regulation of commercial advertising for the pending initiatives at the national and the European level to regulate online political advertising, and online political targeting in specific.

Keywords

fundamental rights
online and commercial political targeting
platforms
regulation
unfair commercial practices

URN: urn:nbn:de:0009-29-53386

Untitled Document

urn:nbn:de:0009-29-53386

1. Introduction*

“Hold political ads to the same standard as other ads” was the first recommendation made by hundreds of Facebook employees in an open letter to the Facebook leadership. [1] The letter criticised Facebook’s policy of excluding political ads from its fact-checking programme that is intended to identify false or misleading content, including advertising. Facebook cites freedom of expression concerns and respect for the democratic process as reasons for the different treatment of political versus commercial ads. [2] The fundamental right to freedom of expression and the importance of political speech for the democratic process are more generally important reasons why commercial advertising also in law follows a different path than the regulation of political advertising. And yet with the advent of digital technology, social media and new forms of political advertising, elements of political and commercial advertising are increasingly intertwined. Online political targeting in particular raises new issues of voter protection and challenges a number of regulatory assumptions and path-dependencies that we have taken for granted for too long. [3]

Political campaigns still rely on the mass media to send campaign messages that appeal to a large part of the electorate, [4] but in addition to that, digital technology has enabled new forms of personalised political advertising, whereby political campaigns can target increasingly small segments of the electorate with tailored messages. Students, for example, no longer see political ads about pensions; instead, they see ads about student debt or student housing in the city where their university is located. Political campaigns can personalise these messages to a considerable degree, as long as the political campaign has 1) vast amounts of data about the electorate, 2) the skills and tools to analyse the data and to make meaningful advertisements and 3) the infrastructure (and money) required to spread those ads. [5] The advent of online political targeting has given rise to both new hopes and concerns about the fairness and governance of these practices. [6] There are concerns about the opacity and lack of accountability of these practices, [7] the danger of polarisation, [8] filter bubbles [9] and the ability of voters to engage in a shared discourse. [10] Related to this are more general concerns about the way political power is shifting from political parties to platforms, [11] and instances of voter exclusion, discrimination and the ability ‘to sidestep less sympathetic audiences’ or invest time in voters who are unlikely to vote. [12] In response, countries around the world are increasingly devising ways to regulate political microtargeting. Devising new rules for online political targeting is also a priority for the European Commission (EC). The Commission’s European Democracy Action Plan announced legislative proposals on the transparency of political advertising and possible further restrictions on microtargeting and forms of psychological profiling. [13] The regulation of online political targeting, however, presents regulators with a difficult conundrum.

The existing rules on political advertising are intended to strike a careful balance between respecting the status of political advertising as the highest protected form of speech and the need to lay down some ground rules in the interest of fair elections and the protection of voters. [14] Most of these rules focus on the traditional mass media that have long been the primary vehicle to disseminate political advertising to voters. [15] Online political targeting is different from the traditional forms of advertising via the mass media. There is, first of all, the far more central role of data, in combination with powerful data analytics tools that allow for predictive modelling and the increasingly precise targeting of content and delivery of political messages, than in the traditional mass media. [16] The combination of detailed knowledge about voters, their behaviours, fears and preferences with data-driven profiling (i.e. adjusting message and distribution strategy to individual or group profiles) provides entirely new levels of persuasion knowledge and therefore has heightened concerns about voter manipulation and unfair forms of subconsciously undermining voter autonomy. [17] Data-driven tools provide advertisers and platforms with a much more detailed view of the target audience than traditional forms of advertising do (information asymmetries). The advertisers and platforms learn information about the citizen, while the citizen has a limited understanding of the data machinery operating behind the scenes, leading to their exposure to a (micro)targeted political ad.

A second important difference is the prominent role of new players, primarily social media platforms that serve as both new sources of data (both disclosed and inferred, e.g. in the form of look-a-like audience matching and data modelling) and new advertising infrastructure. Unlike traditional mass media, social network sites with their highly connected structure allow entirely new and far more interactive means of communication with individual voters. As powerful controllers of both economic and communication power, their ability to change power balances and affect fair competition in the marketplace of ideas is the source of much scholarly concern, [18] especially because these actors fall outside the scope of traditional media regulation and the applicable rules in e-commerce and consumer protection law are ill suited to deal with their commercial and political power. [19]

This leads us to a third major difference between traditional forms of political advertising and online advertising, and the one that is most central to this article: the degree of professionalisation and commercialisation of political advertising. As political campaigns increasingly rely on the tools developed for commercial targeting practices and the same commercial parties (in particular the Google and Facebook duopoly) to spread their messages, [20] commercial strategies and motives are increasingly shaping political campaigning strategies. The consequence is that political advertising is turning, at least from the perspective of platforms, into 'just another form of advertising', and it is becoming difficult to distinguish the citizen from the consumer. Or in the words of Brad Parscale, digital director of the former Trump campaign: ‘It’s the same shit we use in commercial, just has fancier names.’ [21]

It is this tension between the political and the commercial that creates new challenges for the regulation of political advertising, an issue that this article is particularly interested in. Because online political advertising is political and not commercial speech, it is destined to follow an entirely different regulatory tradition than commercial advertising. Commercial advertising is subject to consumer law and unfair advertising regulations, including rules about unfair commercial practices. The provisions about unfair commercial practices are intended to protect consumer autonomy and fairness in the commercial marketplace, and to find the right balance between legitimate and illegitimate forms of persuasion. [22] Increasingly, the rules about unfair commercial practices are also discussed in the context of behavioural commercial targeting, as a potential response to concerns about data-driven forms of commercial advertising. [23] Unfair commercial practices law, and other rules about commercial advertising (e.g. rules about unfair comparative advertising), however, are explicitly not applicable to forms of non-commercial political or ideological advertising. [24] An important reason why this is so are differences in the level of protection of political and commercial speech under fundamental rights law standards. And yet, with the ongoing commercial turn in advertising, the traditional division between forms of commercial and political advertising is no longer that self-evident. Also, it cannot be denied that commercial advertising law has a long tradition of thinking of where and how to draw the line between lawful advertising and unlawful persuasion through withholding or misleading consumers about the information they need to make informed decisions, or by exploiting information asymmetries, exerting undue psychological pressure and engaging in other forms of unfair behaviour.

The question that this article therefore explores is: “Are there valuable lessons to learn from the way the law approaches fairness in commercial advertising for the future regulation of political advertising?” It is explicitly not the goal to discuss a possible extension of unfair commercial advertising regulation (as most notably laid down in the Unfair Commercial Practices Directive) [25] to online political advertising. This article also does not explore to what extent data protection law imposes regulatory constraints on online political advertising, a question that has been discussed extensively elsewhere. [26] Instead, we explore the nexus between online commercial and political advertising, and possible inspiration for regulatory tools or instruments that can inform the future regulation of online political advertising.

2. The commercialisation of political advertising

In the following section, we scrutinise in more depth the ongoing commercialisation of political advertising and of voters, and the main factors that drive it, namely data and data-driven platforms.

2.1. Merging data on voters and consumers

Collecting personal data on voters is not new to political campaigning and political parties were collecting such data long before the widespread proliferation of the internet, for example in the form of public voter registries and data that political parties collect directly from their voters. In their history of political data in the United States, Kreiss and Howard (2010) pinpoint the origins of campaign data practices in the 1960s, but also show how the arrival of the internet offered political campaigners new ways to use the data they had collected to directly interact with voters and amplify their messages. Already then they were aided in their efforts by early commercial platform services as well as the use of commercial data brokers (such as "Adobe, Oracle, Salesforce, Nielsen, and IBM") [27] and other sources of commercial data about the behaviour of voters, as consumers, online. In both the United States and Europe, data brokers gather data from public sources, through surveys, promotional actions, purchased data sets (also from offline behaviour, such as magazine subscriptions or loyalty card programmes), and they add value by cleaning the data, combining datasets and keeping them up to date. The arrival of social media platforms in the late 1990s unlocked another wealth of personal data, as well as the ability to purchase data that users disclose on these platforms and the data that social media platforms inferred from the behaviour of users (often in their role as consumers), as well as look-a-like audience matching and custom audience services. [28]

2.2. Data-driven advertising as a business model

Today, social media platforms (such as Google and Facebook) are the most important actors in online advertising, because of their size and infrastructure and the wealth of new data sources that they unlock. The size of the bigger platforms allows them to collect a lot of information about users and to subsequently use that information to infer or predict behaviour. The platforms' easy-to-use infrastructure then allows advertisers to cheaply microtarget voters. Social media platforms offer their services to commercial and political advertisers alike. Facebook, for example, offers its advertising services to commercial and political campaigners via a centralised ads manager, [29] as do Google [30] and Twitter. [31] Commercial and political advertisers can even compete with each other by placing a bid into the platforms' auction systems in the hope of being allowed to show their ad to a specific audience. Oftentimes, there are many different parties—political and commercial—seeking to display an ad to the same specific audience.

More recently, and in response to scandals such as Cambridge Analytica and increasing concerns about the role of social media in elections, social media platforms have been adjusting their service offers. For example, Twitter banned the promotion of political content altogether, based on a belief that ‘political message reach should be earned, not bought’ [32], while Google limited ‘election ads audience targeting’ to some more general categories, not offering more granular microtargeting and committed to more transparency. [33] Facebook suspended running ads about social issues, elections and politics only temporarily in the run-up to the United States 2020 elections, [34] and continued to offer outside the United States the ability to target ads at custom audiences and look-a-like audiences or to define an audience “based on criteria such as age, interests, geography and more”, including interest and behaviour. [35] The more recent adjustments to the range or reach of their advertisement services, however, do not change the general business proposition. As Witzleb and Paterson observe, “the same personal data gathered by online platforms is as valuable to platforms and other businesses seeking to sell goods and services, as it is for political parties and political interest groups seeking to ‘sell their programs, ideas and ideologies.’” [36] Thus, social media platforms are important drivers behind the increasing commercialisation of political advertising and are blurring the lines between commercial and political advertising. [37]

2.3. The same tools and strategies to rule them all

Social media platforms also sell their sophisticated skills and tools for data analysis. Advertisers do not necessarily have the in-house knowledge and tools to turn vast amounts of data into something meaningful. Platforms, therefore, actively offer their services to political campaigns in the United States [38] and in Europe. [39] Additionally, commercial and political advertisers can outsource their big data analysis to consultancies. [40] The 'meaningful information' resulting from such analyses can just as easily be employed for political as for commercial purposes, and is the source of a range of new forms of online and political advertising, ranging from programmatic advertising and targeting across different devices, through targeting based on location (geolocation targeting), demographic or personal information, to forms of psychographic targeting or neuromarketing that are driven by intimate insights into the emotions, desires, personalities, attitudes and behavioural biases of users and informed by the insights of cognitive psychology. [41]

2.4. Similar concerns of users and voters

The use of online targeting strategies and psychological targeting strategies in commercial advertising has given rise to a number of concerns about the rights of consumers. For commercial targeting, the Dutch Consumer Authority observed that as a result of profiling strategies, “businesses can steer consumers’ behavior very effectively, potentially affecting the autonomy of consumers”. [42] The European Consumer Protection Organisation (BEUC) states that under certain conditions behavioural advertising can have “undue influence” in the sense of the Unfair Commercial Practices Directive, notably if there is a situation of power due to information asymmetries, and targeting strategies are used to exert pressure on the consumer or ‘prevent the display of other advertisements and reduce consumer choice’. [43] In its Guidance on the application of the Unfair Commercial Practices Directive, the EC concedes that when profiling strategies violate the data protection rights of consumers, doing so can also constitute an unfair commercial practice, particularly if that practice is not transparent or hides the commercial intent, [44] or is designed to exert undue influence through psychological pressure. [45] Scholars have also pointed to the possibilities to identify and target individual vulnerabilities and more generally influence the taking of autonomous decisions. [46]

Some of the concerns regarding the use of commercial targeting are echoed in the literature about online political targeting. An example are concerns related to the inability of users to judge political advertising on its value and take well-informed, autonomous decisions. This can be because of the deceptive or misleading content of the political message itself, [47] a lack of transparency [48] or using microtargeting to make divergent promises to different voters. [49] The information asymmetry – where the political advertiser has a detailed profile of the voter, while the voter has no idea about the mechanics and information behind the targeted advertisement she receives [50] – enables the political advertiser to not only stay under the radar, but also to lie, mislead, pressure or leverage fears more effectively. And as in behavioural commercial targeting, also for political targeting practices the use of ‘psychographics’ or persuasion profiling and knowledge of biases and political concerns and views on particular political topics to exercise undue influence over voters is another key concern in the discussions about online political advertising, and microtargeting in particular. [51] Other concerns therefore relate to the way the political message is delivered, for example by developing rich voter profiles that reveal preferences, fears, beliefs and other characteristics and combining them with psychological insights to tailor content and form of the message, [52] or identify and exploit individual vulnerabilities and biases. [53]

To conclude, the fusion of political and commercial players, along with tools and data sources is accompanied by a number of important implications for political advertising, as well as the protection of users thereof. Both commercial and political advertisers use similar data, similar tools and similar infrastructures to target their audiences. As the tools and strategies are the same, it stands to reason that also some of the concerns regarding the commercial use of some profiling strategies (unfair forms of manipulation, loss of autonomy, data protection and surveillance, the potential to exploit individual vulnerabilities) [54] arise in the context of political targeting. Users for their part are potential voters and consumers alike and are confronted with the difficult task of having to process and distinguish between commercial and political messages. Perhaps one of the most obvious consequences is the central role of and dependency on social media platforms that can leverage the data, tools and infrastructure that they developed to both political and commercial advertisers. Unlike political parties, these are commercial players that are essentially driven by commercial interests to increase revenues and are accountable not to voters but to shareholders. If political advertising is yet another form of advertising, should we not offer users the same level of protection vis-à-vis unfair forms of commercial and political advertising? This is the question that the next section investigates.

3. Regulation of commercial and political advertising – different regulatory traditions

So far, the regulation of commercial speech and that of political speech have followed separate paths. An important reason why this is so lies in fundamental rights law, and the differences in the margin of appreciation that national governments have to regulate commercial vs political speech. From the perspective of fundamental rights law, commercial and political speech are not the same, though both enjoy freedom of expression protection. [55] Government restrictions on political speech receive a far higher level of scrutiny regarding their compatibility with Art. 10 ECHR. The European Court of Human Rights has indeed consistently held that the margin of appreciation that states have in deciding whether or not to regulate speech is “is essential in commercial matters and, in particular, in an area as complex and fluctuating as that of unfair competition,” [56] which gives states more room to interfere with commercial speech than political speech. Elsewhere, the Court explained: “For the citizen, advertising is a means of discovering the characteristics of services and goods offered to him. Nevertheless, it may sometimes be restricted, especially to prevent unfair competition and untruthful or misleading advertising. In some contexts, the publication of even objective, truthful advertisements might be restricted in order to ensure respect for the rights of others or owing to the special circumstances of particular business activities and professions.” [57] As a result, commercial advertising is subject to a range of advertising regulations that can include scrutiny of both the fairness of the message (e.g. whether or not it is misleading) and the way the message is delivered (e.g. in a way that amounts to exerting pressure on consumers). [58]

Commercial advertising regulation serves at least three goals: (1) the protection of consumers and their ability to make informed, rational choices, (2) the protection of competitors against unfair competition and (3) the protection of a broader public interest in information, [59] on the one hand, and a fair and functioning marketplace on the other hand. [60] Over the course of time, the legal order has developed a range of instruments to concretise these objectives, including rules intended to: [61]

Protect consumers against particular products (the regulation concerning tobacco advertising is an example) [62] or protect particular groups of consumers (e.g. the rules with regards to the protection of minors in the AVMSD). [63]
Protect consumers (and indirectly public information interests and fair competition) against misleading or otherwise unfair advertising (and here, in particular the provisions of the Unfair Commercial Practices Directive and its implementation into national laws). [64]
Protect competitors (and fair competition) against particular forms of unfair advertising, including comparative and denigrating advertising (the Directive on misleading and comparative advertising), [65] as well as under national, non-harmonised rules on torts, libel and defamation [66].

At the heart of the regulation of commercial advertising is the standard of fairness and good faith in advertising. [67] Under the Unfair Commercial Practices Directive, for example, commercial practices are unfair where they are either contrary to the requirements of professional diligence, or can or do “distort the economic behaviour” of consumers (Art. 5 (2) UCPD), [68] through misleading or aggressive practices. The main objective behind the ban on misleading practices is to provide consumers with the correct information they need to take informed and autonomous decisions. [69] The provisions about aggressive practices go beyond transparency and are concerned with forms of exerting pressure or other forms of undue influence on the actual decision-making, as well as on consumers’ fundamental rights, such as privacy. [70]

While the function of commercial advertising is primarily linked to the economic marketplace, political advertising is associated with the marketplace of ideas. Paid advertising can be a means for political parties to convey a message to the public, and particularly for smaller political parties it can even be a means to compensate for the relative lack of media coverage compared to what larger political parties might receive. [71] According to the European Court of Human Rights, ‘[f]ree elections and freedom of expression, particularly freedom of political debate, together form the bedrock of any democratic system’. [72] Moreover, as the Court has stated elsewhere, “[t]here is nothing to prohibit a political party or wealthy individual or organisation from spending money on publicity in support or opposition to a political party or tendency generally, at national or regional level, provided that there is no intention to promote or prejudice the electoral chances of any particular candidate in any particular constituency’. [73] Accordingly, “there is little scope under Art. 10 (2) of the Convention for restrictions on political speech or on debate on questions of public interest”. [74] Freedom of expression protection also applies to contributions to the public debate that represent a minority opinion and are not a generally accepted idea, at least in “a sphere in which it is unlikely that any certainty exists”, [75] as well as to information that offends and shocks. [76] Contributing to the high level of protection for political speech is the fact that the regulation of political advertising affects not only individual freedom of expression rights, but also—and even more so—the integrity of the political process and societal interest in political debate and fair elections. [77] In VgT, for example, the court made explicit that the margin of appreciation of whether regulatory interventions are permissible can be further reduced in situations in which what is at stake is not an individual’s purely commercial interests, but their participation in a debate that affects the general interest. [78]

This is not to say that it is impermissible to regulate political advertising, at least in Europe. [79] The ECHR acknowledged also in cases concerning political speech that the rights and freedoms granted by Art. 10 ECHR can be subject to exceptions, provided those exceptions are “construed strictly, and the need for any restrictions must be established convincingly, particularly where the nature of the speech is political rather than commercial”. [80] At times, the right to freedom of expression and that to free elections can also conflict. In such a situation, restrictions on free speech rights that are normally inacceptable can be justified if such restrictions are necessary to “secure the free expression of the opinion of the people in the choice of the legislature.” [81] For example, the Court acknowledged that a public interest in protecting the democratic debate during election times from distortion and unfair competition between candidates can be a legitimate reason to restrict political speech. [82] The Court also considered legitimate “certain formalities, restrictions or penalties … during an election period, for instance to ensure a level playing field, for example by way of regulating and controlling campaign expenditure.” [83] The same is true for rules regarding the transparency of campaign finances, and “enforcing the voters’ right to impartial, truthful and balanced information via mass media outlets and the formation of their informed choices in an election” are legitimate aims that can justify regulatory interference, [84] as are spending limits and rules with the goal of “securing equality between candidates”. [85] Similarly, regulation of political speech to protect the diversity and inclusivity of the public debate was considered a legitimate interest to restrict political speech under certain circumstance. [86] Moreover, in situations in which there was not yet a European consensus on how to regulate political advertising, states can enjoy a greater margin of appreciation. [87]

In response to the conditions for interference with political speech as defined by the ECHR, the existing rules that regulate political advertising in Europe [88] have as an important objective the creation of a level playing field between political parties—for example in terms of campaign financing rules, spending limits and transparency obligations—as well as the regulation of the role of the mass media (predominantly public broadcasting) in disseminating information and party standpoints while serving the ‘voter’s right to impartial, truthful information’. [89] Examples are the regulation of allocating equal time for political parties or even free airtime: political parties can buy broadcasting time or sponsor political ads, but each political party should be entitled to an equal share of broadcasting time. Other countries have banned paid political advertising in the media altogether, coupled with exceptions in election times or the entitlement to free airtime. Similarly, the obligations to provide fair, balanced and impartial coverage in the media, to exercise restraint in the publication of opinion polls or to enforce quiet periods, all depart from the idea of the media as a central actor whose task is to guarantee fairness in political advertising, with the national media authorities responsible for enforcing the rules. Importantly, unlike in commercial advertising law, and flowing directly from the reduced margin of appreciation of states to regulate political speech, common to all the regulations is it that it is not so much the message itself as the conditions of its placement (e.g. amount of funding, bans on funding from particular actors, reflection days, fair and balanced coverage, etc.) that are subject to regulation. Having said so, it is also worth noting that in response to the digitally enhanced proliferation of dis- and misinformation and the growing entanglement of the issues of dis- and misinformation and political advertising, more recent pieces of legislation have also opened the door to scrutiny of the political message itself (more about this later). [90]

4. Political advertising on social media platforms – between commercial and political speech

In the following we argue that from the point of view of law and freedom of expression, (paid) online political advertising on social media platforms is a special case because of the way commercial and political elements and interests are entangled (see above). Accordingly, the regulation of paid online political advertising cannot easily be dealt with under either the commercial or the political speech paradigm. To discuss the extent to which the regulation of political advertising law can learn lessons from the way commercial advertising is regulated, we therefore need more clarity about the possible margin of appreciation that states have in regulating online political advertising.

The fusion of commercial and political elements in advertising is in itself not new. On a number of occasions, the ECHR has had to decide on the margin of appreciation of states to regulate speech that included both commercial and political elements. On these occasions, the court highlighted that the mere fact that the speech originates from a commercial for-profit company does not in itself exclude its protection as political speech. [91] An important factor in the considerations of the court is whether the commercial interests of the individual advertiser outweigh the advertiser’s interest in “participation in a debate affecting the general interest” [92] and the rights of the public to receive such information. [93] In other words, speech, even if it is uttered by a commercial player and to commercial ends, can enjoy Art. 10 ECHR protection, but states may have a larger margin of appreciation in regulating it, particularly if commercial ends are overweighted. The Court has also had to decide on cases in which political and commercial interests conflicted, and where regulatory interference was necessary to “protect public opinion from the pressures of powerful financial groups and from undue commercial influence; to provide for a certain equality of opportunity among the different forces of society; to ensure the independence of broadcasters in editorial matters from powerful sponsors; and to support the press.” [94] In VgT, the Court explicitly acknowledged that a competitive advantage of ‘powerful financial groups’ in the realm of commercial advertising can ultimately impact the realisation of freedom of expression and media pluralism (albeit for the case of TV advertising). [95]

What are the possible implications of this case law for the regulation of online political targeting? Where the goal of online political advertising is to contribute to matters of public interest and debate, online political targeting will fall under the qualification of political speech, with the consequence that states are limited in their ability to regulate it, similar to political advertising in the mass media. However, a number of distinguishing features of online political advertising, as opposed to political advertising in the mass media, that we identified earlier can be expected to also affect its evaluation from the perspective of Art. 10 ECHR. One is the ability to target advertising messages at smaller segments of the population, or even individual users, based on various forms of profiling, including psychographic profiling as a practice that, so far, we know only from the realm of commercial advertising. We explained earlier that certain forms of psychological online political advertising could have a more pervasive or even manipulative effect and therefore could impinge on the fundamental rights of citizens to freedom of expression and free elections. [96] This pervasive or manipulative effect of online political advertising could justify a larger margin of appreciation for states to protect voters from unfair manipulations of their political choices, [97] particularly if that effect can be accredited to the means of dissemination of a political message, rather than its content. [98] Indeed, the ‘pervasive effect’ of particular forms of media (here, audio-visual media) has been cited repeatedly by the Court as an argument that can justify government intervention in Art. 10 ECHR. [99]

Another side effect of the more targeted nature of political ads on social media platforms is that they are, unlike political ads in the mass media, more difficult for public watchdogs to scrutinise, [100] putting more responsibilities on individual users to recognise false and misleading political ad strategies. [101] To the extent that these concerns counter the goal of promoting public debate and free elections, one could argue that there is more room for regulation to strengthen the position of users (voters). [102]

The third aspect is the commercialisation and platformisation of online political advertising that we discussed above. In the grey area between commercial and political speech, the court has so far had to decide whether the commercial or public interest contribution of the speakers themselves was overweighted. The situation of online political advertising on social media platforms is different insofar as it is a commercial party that offers, as part of a commercial service, political speakers the opportunity to use its communication infrastructure and insights into the personal and political preferences of its users. Though Facebook and Google, for example, have some additional authorisation requirements for political and issue advertising, [103] both commercial and political ads are managed via the same business manager. The sale of online political advertising as a service by platforms favours the emergence of new practices, but also endangers the fairness and integrity of the democratic process by, for example, making it easier for foreign entities to buy political advertising, parties other than political parties to buy ads under false or misleading identities, etc. [104] The distinct roles and interests of, on the one hand, political advertisers and, on the other hand, online platforms suggest the need for further differentiation, including from an Art. 10 ECHR perspective, particularly in situations in which the selling of political advertising is “just another form of advertising”.

5. Existing initiatives to regulate online political targeting

The current EU approach to the regulation of online political targeting rests on three pillars, namely protecting the personal data of voters against unfair forms of processing, increasing transparency and regulating disinformation, with enhancing transparency again an important priority. [105] Regarding the last-mentioned, so far the main regulatory instrument to deal with disinformation, and in that context also with online political microtargeting on social media platforms, is the EU Code of Practice on Disinformation: a co-regulatory initiative to set some standards regarding transparency, cooperation with authorities and academics, fact-checking and automated content moderation. [106] Signatories promise to, among other things, make efforts to explain to users why they have been targeted and who is behind the targeting. As a result of the Code, and other initiatives to exert public pressure, the major platforms have also created so-called ad archives to complement their more user-facing transparency measures. [107] The importance of transparency requirements is also underlined in the EC’s recommendation from 2018 and in statement 2/2019 by the EDPB. [108] First evaluations of the Code by ERGA [109] and the EC itself [110] have revealed a number of shortcomings in implementation and compliance with the Code, prompting the EC to announce additional legislation on transparency in political advertising as part of the European Democracy Action Plan and an update of the Code. [111] In addition, the proposed Digital Service Act (DSA) includes mandatory provisions on online advertising transparency [112] and ad archives for platforms. [113] It is worth noting that the proposed rules in the DSA make no distinction between online commercial and online political advertising transparency.

More specifically in the context of EU elections, the EC also issued a number of recommendations, again with a strong focus on awareness- and transparency-enhancing measures. [114] The issue of data protection in political campaigns has also received some regulatory attention. The EDPB has stated that significant effects can occur in the context of microtargeting when it significantly affects the circumstances, behaviour or choices of the individual. [115] Building further upon this opinion, the EC has stated that political microtargeting, given the significance of the exercise of the right to vote, has the effect of stopping people from voting or making people vote in a specific way, could be a significant effect in the sense of Art. 22 GDPR. [116] This would make Art. 22 GDPR applicable in the case of political microtargeting.

In Strasbourg, the Committee of Ministers of the Council of Europe (CoE) recommended also applying its recommendation on measures concerning election campaigns [117] to non-linear audio-visual media services. [118] Though it does not specifically mention online political advertising, the recommendation more generally advises extending national rules on the fair, impartial and balanced reporting of elections to on-demand and similar services. [119] More specifically geared towards online political advertising, the CoE’s Declaration on the manipulative capabilities of algorithmic processes emphasises the need to assess the applicability of existing regulatory frameworks on political communication also to the online world, and declares that “it should be ensured that voters have access to comparable levels of information across the political spectrum, that voters are aware of the dangers of political redlining, which occurs when political campaigning is limited to those most likely to be influenced, and that voters are protected effectively against unfair practices and manipulation.” [120]

At the level of member states, the few existing initiatives to regulate online political targeting can be divided into four types. First, some focus on the application and enforcement of data protection rules, such as the call by the Information Commissioner’s Office (ICO) for a statutory code on the use of personal information in targeted political advertising (which, at the time of writing, has not yet led to concrete legislator proposals). [121] Then there are regulations that mandate more user- or public-facing transparency. The French Law of 22 December 2018, for example, obliges online platforms to inform its users about the identity of the entity behind the advertisement, the amount paid for the advertisement and the use of the user’s data in the advertisement campaign during election times. [122] Article 7 of the Slovenian Law on Election and Referendum Campaigns has similar transparency requirements for all types of ‘media publishers.’ [123] The UK has announced an open consultation on proposals for transparency requirements for online political campaigns. This includes the obligation of advertiser identification. [124] Ireland [125] and the Netherlands [126] are debating similar initiatives.

Some countries follow the recommendation of the Council of Europe and consider the application of existing rules on paid political advertising in the mass media to online advertising, such as in the UK, where the Electoral Commission has stated that spending limits imposed on political advertisements apply to advertising of any kind, including advertising on online platforms. [127] Similarly, France extended its Electoral Code with a prohibition on online political advertising during election periods. [128]

Then there are initiatives that address more generally the online distribution of false or misleading information. An example is the controversial French Law Against the Manipulation of Information, which will be discussed in more detail in a moment. [129] In addition to transparency obligations (including the operation of ad archives), the law stipulates that during the three months preceding an election, judges can, upon the request of a public prosecutor, political candidate or party, or another interested person, decide about “inaccurate or misleading allegations or imputations of a fact likely to alter the sincerity of the upcoming ballot [and that] are disseminated in a deliberate, artificial or automated and massive manner by means of an online communication service to the public.” [130] In a similar fashion, the French Media Authority (Conseil Supérior de l’Audiovisuel or CSA) is entitled to act against the dissemination by foreign state actors of false information that is likely to alter the fairness of a ballot. [131] This last example of rules that target the dissemination of false or misleading information in political communication echoes a growing array of national rules to counter the spread of mis- and disinformation, also spurred by the Covid crisis. [132]

While most of the regulatory initiatives so far are either in the realm of data protection law or follow the tradition of regulating political advertising in the mass media, some of the new regulatory approaches can be argued to show elements that are better known from the realm of consumer law. Examples are the requirements to inform users that a message is an advertising message and to provide the identity of the issuer of that message, [133] to clearly separate editorial from commercial content [134] and to protect users from unfair and misleading advertising. [135]

The explicit reference to possible lessons to learn from the way commercial advertising practices are regulated is more pronounced outside Europe, notably in Australia and Canada. In Australia, the Legal, Constitutional and Administrative Review Committee of the Queensland Parliament suggested truth in advertising rules that were explicitly inspired by the rules and methods developed to deal with misleading or deceptive advertising under section 52 of Australia’s Trade Practices Act. [136] In the United States, where the Fair Trade Act does not apply directly to political advertising, a number of states have adopted laws against misleading political advertising, inspired by, inter alia, the way commercial advertising has been regulated. [137] Some of these laws have been struck down by courts because of First Amendment concerns, pointing again to the difficult tension between the constitutional protection granted to political speech and the use of advertising practices better known from commercial advertising. [138] And yet, as we have argued in the previous sections, in online political advertising, and political targeting on social media platforms in particular, commercial and political elements of advertising are merged in ways that seem to broaden the margin for states to draw lessons from a long tradition of protecting users against unfair marketing practices in commercial advertising law. This is not to say that commercial advertising is or should be applied to online political advertising. However, some of the approaches and instruments developed under commercial advertising law, we argue, can usefully inspire our thinking about future approaches to the regulation of political microtargeting (within the limits of Art. 10 ECHR). This is what we try to do in the next section.

6. Possible takeaways from the regulation of commercial advertising for political advertising regulation

The previous sections have demonstrated that although the regulation of online commercial and online political advertising and targeting have followed very different paths and were born out of different regulatory traditions, in practice both types of advertising have many elements in common. An important common element is the use of data-driven persuasion strategies that trigger new concerns about the ability of consumers, aka voters, to protect themselves from unfair forms of advertising. We have also demonstrated that regulating fairness in advertising in the advertiser–consumer relationship has a long tradition in the regulation of commercial advertising. This section explores whether there are possible takeaways from the regulation of commercial advertising and, if so, how they could inspire the future regulation of political advertising.

6.1. Takeaway 1: The need for a pragmatic and flexible definition of the scope of regulation

One of the difficulties of regulating online political advertising is that of defining what a political advertisement is, to what extent also issue-based advertising is covered and exactly what acts fall under the notion of political advertising—in other words, the scope of the regulation. [139] The EU Code of Practice on disinformation defines political advertising as “advertisements advocating for or against the election of a candidate or passage of referenda in national and European elections”, while issue-based advertising is not defined. In European member states, the definitions of online political advertising vary greatly between actor-based approaches (who is the advertiser), whether the advertising is paid/not paid for and purpose-driven approaches (to promote a political party or political end) and are typically geared towards banning or restricting certain practices from the onset. [140] Similarly, there are huge differences in the definitions that platforms handle. [141]

The ambiguity of any definition of ‘advertising’ is a problem that the regulation of political advertising shares with commercial advertising regulation. Unfair commercial practices law opted for a broad definition: ‘any act, omission, course of conduct or representation, commercial communication including advertising and marketing, by a trader, directly connected with the promotion, sale or supply of a product to consumers’. [142] But, and this is important, for the rules to apply, practices must ‘materially distort the economic behaviour of consumers’. [143] In other words, when regulating fairness in commercial advertising, the law acknowledges that commercial persuasion can take many forms. The proof lies in the potential effect that a particular act of communication has on the consumers’ decision-making process and whether that effect is achieved by fair or unfair means.

This could be a first valuable lesson for the regulation of online political targeting. Instead of trying to define upfront in much detail what is or is not an impermissible political ad, an alternative approach would be to opt for a fairly broad and inclusive definition, and subsequently make the assessment of the lawfulness of the political ad dependent upon the potential effect on different voters and the electoral competition. [144] This is because from a citizen’s perspective, political advertising comes in different shapes and fairness has a different meaning for each type of ad. One subset of ads are concerned with maximising engagement and turnout, or dampening it by suggesting that elections are foregone conclusions. A second subset focus on candidates, either an incumbent or an opposing candidate. A third subset focus on the issue and policy standpoints of parties. Each type of ad comes with a different set of considerations about what fairness entails. For the first, for example, cueing citizens against turning out to vote is potentially an infringement of electoral fairness. For the second type mentioned above, spreading dis- or misinformation about opposing candidates might be an infringement of electoral fairness.

A relevant political practice could be then defined as ‘any act, conduct, representation or advertising of political issues and standpoints, candidates, party programmes or part of such programmes that is directly connected with the promotion of a political party, political programme or candidate to citizens, or the engagement in the act of voting.’ Such a broad definition would also acknowledge that political advertising messages themselves could not only potentially constitute unfair behaviour, but so too could the sponsoring of certain political events, websites or Facebook groups, or the creation of persuasion profiles of particularly vulnerable citizens (see below)—as long as it has the potential to affect voting decisions. For the same reason, acts by non-party political actors would be covered, as long as the primary aim is to directly influence voter decision-making. It would exclude instances of mere journalistic reporting about political events to the extent that this reporting is, in conformity with journalistic ethics of objectivity and unbiased reporting, not directed at having a particular effect on voters’ behaviour. The advantage of such an approach is that it would be flexible enough to include current and future forms of online political advertising and account for the fact that the process of political opinion and preference forming can be influenced in many different ways and by many actors (including non-party actors). The legislator could then qualify under which conditions such practices have an non-permittable effect on voters or elections, for example because they are misleading or dissuading voters from voting. The drawback of such an approach is that it would be very inclusive and ultimately would require an authority that, similar to the judge in unfair commercial practices law, is authorised and competent to assess practices upon their fairness. The advantage of such an approach would be that it is the judge, bound by fundamental rights law, and not a social media platform that decides about the permissibly of a political ad.

6.2. Takeaway 2 – The information that users need to assess the fairness of targeted advertising depends on the situation and the concrete targeting strategy

One of the key principles of unfair commercial practices law is that for commercial practices to be fair, they need to provide the consumer with all the relevant information that she needs to take an informed decision in a particular situation. On the contrary, practices that omit relevant information or contain false information or truthful information that is presented in a way that can still deceive the consumer, are considered misleading and are thus banned. [145] A necessary precondition is that the provision of misleading or the omission of relevant information has caused or is likely to cause the consumer to take a decision that she otherwise would not have taken. The reason for this qualification is that unfair commercial practices law protects not truth in advertising in abstract, but the ability of consumers to make autonomous and well-informed decisions.

Using political targeting strategies to mislead the voter is also a key concern in the discussion around online political targeting (see section 2), but what information voters need to assess the fairness of a practice depends on the practice. Earlier we distinguished between political ads that are aimed at maximising engagement and turnout, focus on candidates or focus on the issue and policy standpoints of parties. Regarding the latter category, Zuiderveen Borgesisus et al. (2018), for example, warn of a situation in which online political targeting can be used in such a way that a party presents itself falsely as a one-issue party so that each individual receives only information on the issue that she is likely to be most interested in, while omitting information on other issues. [146] Arguably, for a voter to take an informed decision in such a situation she would need to have an idea of the broader set of issues a party stands for. This information is different from the information a voter might need for ads that fall into the second category and cue citizens against turning out to vote. Here, information about the party that commissioned the ad is relevant to assess the ad upon its value. And regarding the first category, ads that are concerned with maximising turnout, information about the strategies used might be the most relevant information for voters. For example, empirical research in the United States [147] found that 86% of respondents thought it was not okay to be targeted with political ads (as compared to 61% being uncomfortable with commercial targeting). Similar research in Europe has demonstrated that many people are concerned about online political targeting. [148] Turow et al. found that “between 57% and 70% of Americans do say it would decrease the likelihood of voting for their candidate either a lot or somewhat”. [149] In other words, it can actually matter for the decision of a voter what techniques are used to maximise engagement, and hence having that information is necessary to take an adequately informed decision. Similarly, one could also argue that to be able to take an informed decision, voters should learn whether they are subject to A/B testing (meaning the message has been optimised for resonance rather than political content [150]), whether a political message has been automatically generated by AI or a bot to respond to individual profiles (rather than by a human campaigner) whether an ad is based on custom audiences, or whether it is paid for or not.

How is that approach distinct from current calls about the need for more voter transparency? The proposed measures at the national or European level require that voters should be informed about a number of items. For instance, the EU Code of Practice calls for transparency “also with a view to enabling users to understand why they have been targeted by a given advertisement”. [151] The Council of Europe recommends revealing to users the “advertising purpose, the methods by which they are targeted to citizens, and their funding”. [152] Draft Art. 24 DSA requires an advertisement to be labelled as such and the provision of the name of the person on whose behalf the advertisement is displayed and meaningful information about the main parameters used to determine the recipient. And the recent French Law relating to the manipulation of information requires consumers to be explicitly informed about the identity of the political advertiser as well as the way personal data is being used. [153] The approach suggested here is less deterministic from the onset. It is a flexible approach that leaves room to take into account the concrete informational needs of users by asking: what kind of information and in which form do voters need to take informed decisions in this particular advertising context? In other words, this is a user-centric approach that is oriented a particular situation, as opposed to the ‘long list-approach’ that can be found in many of the current rules (and proposals) for regulating online political advertising. Not only misrepresenting such information, but also leaving out necessary information should be considered unfair. Such a more flexible approach would also allow to interpret the concrete information needs of voters in the light of the insights of the most recent empirical findings on users’ perceptions and information requirements for taking informed decisions. [154]

6.3. Takeaway 3 – It should be up to judges, not platforms, to assess whether claims made are false or misleading

A more controversial issue than transparency is the evaluation of truth in advertising in the message itself. Much has been written on the topic of disinformation and the way it could threaten the democratic process, [155] as well as the risk of regulatory intervention interfering with fundamental rights, including freedom of expression interests. [156]

Engaging in commercial communication that is false or deceptive can be considered an unfair commercial practice, and thus be banned provided it causes or is likely to cause the consumer to take a decision that the consumer would otherwise not have taken (e.g. not only the message as such but also the potential effect on the consumer matters). [157] Arguably, in such a situation, the public interest in trust in a fair and functioning marketplace, and the protection of the autonomy of consumers, carries more weight than potential interferences with the freedom of expression interests of commercial advertisers. Ultimately, however, it is the judge who is tasked with this decision.

In the case of political advertising, this balance can tip, at least because of the higher level of protection under Article 10 ECHR and the reduced margin of appreciation of public authorities (see also section 3). This is arguably less true for political advertising that is clearly unlawful (e.g. because it is defamatory) or false and purposefully harmful (so-called disinformation), including false deepfakes. [158] In all the cases the real difficulty lies in the grey zone of communication that is neither clearly false nor intentionally harmful. Any regulation or standards on unfair political practices would need to avoid a situation in which the scrutiny of such practices results in prohibited censorship or interference with political speech rights.

Interestingly, in Google’s announcement of the changes to its political advertising rules, the company bans practices that essentially echo the principles of unfair commercial practices law, including “misleading claims about the census process, and ads or destinations making demonstrably false claims that could significantly undermine participation or trust in an electoral or democratic process”. [159] The question is: do we want Google to be the arbiter that decides whether political claims are misleading? The heavy criticism about the ‘private censorship’ of platforms was exactly why Facebook refused to do exactly that, namely assess the content of political messages. [160] Concerns about possible interference with free speech rights is one side of the coin, and the fact that each platform sets a different standard for what it considers fairness in advertising is the other side.

One possible takeaway from the approach under unfair commercial practices law is that it should ultimately be up to a judge (or another authority) to make this decision, and this authority must be bound by fundamental rights law and procedural fairness guarantees. Legislators and judges (or similar authorities), not platforms, should evaluate in which situations the fundamental right to speak is outbalanced by the fundamental rights of citizens to form their political opinion free from deceit and false propaganda. Only in this way can a shared and transparent standard of fairness in political advertising develop. Another potential lesson is that with commercial speech, false or potentially misleading claims are never prohibited without also considering their potential effect on the ability of users to take autonomous and informed decisions. This is different from the approach that some member states have taken lately by outright banning or even criminalising certain forms of alleged disinformation in online targeting. [161] Making the effect of a political advertising on the ability of voters to take autonomous decision central could add an extra level of protection against arbitrary decision making and politically motivated censorship.

6.4. Takeaway 4 – Walking the fine line between regulating content and the conditions of delivery

Distinct from questions about the fairness or legality of a message are questions about the fairness of the way the message is delivered. So far, the predominant approach to protecting voters against unfair forms of delivery of online political advertising focuses on the lawfulness of the way users’ personal data are used, or transparency approaches (see section 5). A question that the existing approaches discussed in section 5 are less well prepared to tackle is under which conditions do data-driven targeting political messages exploit structural power imbalances, individual vulnerabilities and advantages in persuasion power. In the European Democracy Action Plan, the EC hints at the possible necessity of “further restricting micro-targeting and psychological profiling in the political context”, without being more specific about how this could be done. [162] Again, the approach to the regulation of unfair commercial advertising in general and so-called aggressive practices, in particular, can provide useful inspiration.

Perhaps one of the key concerns regarding online political targeting is the risk of voter manipulation and distortion of the democratic process. [163] This is a concern that debates around online political targeting share with discussions on consumer law. Also in consumer law, the use of data analytics and ‘persuasion profiles’ [164] has raised concerns regarding the protection of the autonomy of consumers, and the potential unfairness of these practices. [165] In its last guidance on the application of the Unfair Commercial Practices Directive, the European Commission made clear that certain forms of data-driven targeting—notably targeting that exerts undue influence or constitutes an aggressive practice—can constitute an unfair commercial practice. [166] And yet, although there is a shared perception that manipulating [167] users is potentially wrong, it is difficult to actually pinpoint the conditions under which doing so is unethical or unlawful. After all, there is also agreement that each form of advertising, commercial or political, is essentially an attempt to persuade and, ultimately, to manipulate users. The challenge is to define the conditions that distinguish between lawful persuasion and unlawful manipulation.

Unfair commercial practice law has a long tradition of doing exactly that, namely defining the conditions of unlawful manipulation vis-à-vis lawful persuasion. While the principles of misleading advertising address the existence and abuse of information asymmetries between advertiser and user, the rules on aggressive practices focus on situations where physical or psychological influence is applied in such a way as to reduce a user’s freedom of choice, where an advertiser takes advantage of the specific situation of a user, [168] or uses mental or physical force (coercion) or harassment (causing emotional distress while not serving a legitimate purpose). [169] Unfair commercial practices law thereby makes an, important distinction (for our context) between scrutiny of the commercial message itself (i.e. false information about the price or the product itself) and the conditions surrounding the way the message is delivered (i.e. by omitting critical information the consumer needs to be able to assess the message adequately, or by using force, undue influence, etc. to reduce users’ actual information choice or autonomy in responding to the message). Under unfair commercial practice law, a message can be aggressive if “in its factual context, taking account of all its features and circumstances, by harassment, coercion, including the use of physical force or undue influence, it significantly impairs or is likely to impair the average consumer’s freedom of choice or conduct.” [170]

The provisions about unfair and aggressive practices in commercial advertising law can be criticised, and rightfully so, for example because of their relative vagueness. [171] And yet the way the law deals with these practices includes a deeper truth, namely that advertising that exploits knowledge of individual biases and susceptibility to persuasion, invades an individual’s personal space, [172] as well as forms of economic and intellectual domination [173] or forms of advertising that exert pressure by abusing fears or emotions, [174] are examples of practices that have crossed that precarious line between acceptable persuasion and unacceptable manipulation, particularly when they do so for commercial gain. [175] Unfair commercial practices law for the case of commercial advertising touches on concerns that are also echoed in the literature around online political advertising, particularly in context of so-called psychographic profiling practices (see section 2.4.).

Therefore, another important lesson from unfair commercial practices law could also be that the particular messaging strategies can under circumstances have (by design or circumstance) an adverse effect on the ability of consumers to take autonomous decisions and, if they do so, deserve legal scrutiny. [176] Arguably, in the context of online political targeting that distinction is even more relevant because those who formulate the message (and thus engage in political speech) are often distinct from those that distribute it (social media platforms as part of a commercial service) (see section 2). This is a subtle but important difference: while political advertisers enjoy, according to the case law of the European Court of Human rights, a high level of protection under Art. 10 ECHR, for social media platforms the selling of advertising services is first and foremost a commercial service. As we have argued above, this arguably leaves states with a larger margin of appreciation to regulate targeting strategies, commercial and political, by social media platforms, and because of the potential effects of psychographic profiling on users’ autonomous decision making, there is also a clear public interest in doing so. The concept of ‘aggressive practices’ and the long experience of national courts in identifying the conditions under which advertisers engage in strategies to exert undue influence in the sense of unfair commercial practice law could provide useful inspiration and therefore deserves further exploration.

6.5. Takeaway 5 – Some persuasion strategies are simply unacceptable and should be banned altogether

Much of the current regulatory discourse is focused on the question of how to govern online political targeting practices, beginning with the question of how to make them transparent and observable in the first place. [177] Nevertheless, as important as more transparency in this area may be, another, even more important question remains unanswered: once we are in a position to observe all instances of political advertising (e.g. in the form of ad archives), how do we decide which practices are acceptable in a democratic society, and how should judges or regulatory authorities respond? The above discussion has already pinpointed a number of possible criteria, inspired by insights from a long history in the law of identifying unfair forms of advertising. The experience with regulating unfair commercial advertising, however, also teaches us that, in addition to the more ambiguous cases, there are instances of advertising that are simply unacceptable in a just society.

In unfair commercial practices law, these instances of unacceptable advertising practices are listed in the Annex to the Unfair Commercial Practices Directive. The Annex includes a wide range of forms of commercial communication that are always considered unfair. As arbitrary as this list may be, the message is clear: under certain circumstances there is a role for the regulator to ban practices that conflict with the idea of a functioning and fair marketplace. Do we need a similar list for political advertising and, if so, which instances should be included in such a list? The following section offers a number of suggestions.

One example to consider in that context could be targeted messages that are directed at demobilising voters, [178] such as the example of messages targeted at African-American voters with advertisements that recalled Hillary Clinton’s earlier remarks about calling African-American males ‘super predators’, thereby using microtargeting to suppress voter turnout for their opponents. [179] Arguably, such a practice is in conflict with key principles of electoral fairness, such as the principle of equality of opportunity discussed above, and also triggers concerns about manipulation of the public discourse—thereby questioning their protection under fundamental rights law. In such a situation, is transparency enough to address the potentially anti-democratic effects? Or should society take a stance and ban this form of political targeting?

The reverse practice—that is, targeting voters without them being aware that they are being targeted with specific political messages (so-called dark posts)—could be another example of a practice that deserves critical discussion. Political messages like these potentially bypass the broader public discourse, [180] and thus are in conflict with established criteria of fair elections. [181] Declaring dark posts unfair could build on long-standing experience in the realm of commercial advertising, namely that advertising practices that are invisible to users (because they are unmarked or camouflaged as editorial messages) make informed autonomous decision-making essentially impossible, [182] and because of that are considered unfair.

Another, related potential contender for a political targeting practice that is potentially always unfair is that of redlining, [183] in the sense of focusing on ‘profitable’ sections of the voting population and ignoring others, either because they are unlikely to vote for a candidate or because they are seen as secure votes. Kreiss warns that political advertisers would routinely “redline the electorate, ignoring individuals they model as unlikely to vote, such as unregistered, uneducated, and poor voters”. [184] Communication scientist Joseph Turow points to another concern in this context, one that he calls “rhetorical redlining”, namely the practice of presenting voters with “ads from candidates based on what the campaign’s statisticians believe they want to hear—shutting them off from messages that the statisticians determined might make them waver in their support”. [185] On the one hand, it could be argued that targeting political messages at particular groups in society can be a way to involve those who have shown less interest in politics, and thereby increase engagement. On the other hand, however, redlining can and in practice most likely will be used in such a way as to exclude them. [186] Such a practice seems very much at odds with the key principles of a democratic society, which requires a sphere of mutual shared values and equality: “The dynamics of deliberative democracy are characterised by the norms of equality and symmetry; everyone is to have an equal chance of participation”. [187] As Bayer argues, practices like these potentially also conflict with voters’ right to receive information. [188] Reasons enough to at least question their desirability in a democratic ‘marketplace of ideas’.

Potentially, there are also other ways of delivering political advertising messages that are so problematic from the perspective of voter autonomy and fundamental rights, including the right to receive information, that they should be banned. A practice that is being very critically discussed in that context is again that of psychographic profiling. The main focus of psychographic profiling is not so much the political message itself as figuring out ways to affect how users internalise and respond to the message. As Burkell and Regan explain, in such situations it is far more difficult for voters to detect and counteract a message, particularly if it speaks to unconscious biases. [189] In such situations, constitutional concerns about interfering with the political speech of political advertisers or the commercial interests of platforms are more easily outweighed by the concerns of voters to receive information and fair elections than when more ‘simple’ targeting strategies are concerned with matching the right content with the right people. In a similar vein, a suggestion exists to limit the types of data that may be used for targeting. For example, Jaursch suggests operating a set list of data that may be used for ad targeting, such as electoral district or age and gender, combined with a ban on using certain other kinds of data, such as inferred data or purchased consumer data. [190] Again, doing so would amount to a restriction not so much on political speech itself, as on the way it is delivered.

This list is far from complete and merely serves as an argument that it may be time to develop clearer guidance as to what practices are acceptable or inacceptable in a democratic society, and that for the sake of respect for fundamental rights, such guidance needs to be transparent and prescribed by law.

7. Conclusion

At the heart of our proposal is the argument that data-driven targeted political advertising can not only distort the conditions for fair competition of ideas and opinions between political parties, but also be a threat to democracy because such a practice can impact the ability of citizens to make free, autonomous and informed political decisions. Arguably, and as in the commercial market, the marketplace of ideas can only function if citizens can take free and autonomous decisions and are adequately protected against deception, manipulation and other unfair and misleading practices.

So far, evaluating the fairness of political microtargeting practices has very much been a process driven by individual platforms with Google, Twitter and Facebook all developing their own standards of what they consider fair or unfair advertising. As we saw earlier, these standards differ considerably, also over time, with Twitter imposing a general and very broad ban on all paid political advertising, Google banning certain forms of microtargeting and Facebook essentially adopting a liberal approach. The lack of any benchmarks or commonly agreed upon procedures to assess fairness in political advertising is in the best case confusing for voters, political advertisers and regulators, a situation that is not healthy for the political debate. In the worst case, this is a situation that promotes “platform shopping” and migration to the least strict and responsible platforms, including some less trustworthy ones. [191] Interestingly, when reviewing some of the suggestions made in recent policy initiatives, reports and documents, a trend towards identifying certain elements of fair or unfair online political advertising practices is already observable. However, this is very much an ad hoc process, without any clear conceptual approach. What is needed right now is a method of identifying evaluation criteria or standard benchmarks regarding which online political advertising practices are potentially unfair, also beyond the ambit of one particular platform. We have argued that the experiences with unfair commercial practices could serve as a useful conceptual frame to build on.

As this article has demonstrated, there are a number of lessons to be learned from a long legal tradition of dealing with unfair commercial advertising, and of unfair commercial practices law in particular, including:

The need for a pragmatic and flexible definition of the scope of political advertising regulation.
The information that users need to assess the fairness of targeted advertising depends on the situation and the concrete targeting strategy.
It should be up to judges, not platforms, to assess whether claims made are false or misleading.
Political persuasion can exert undue forms of influence and could thus be unfair.
Some persuasion strategies are simply unacceptable and should be banned altogether.

Political advertisers are learning from the experience of commercial advertisers with online targeting strategies. It is time that those making policies and rules for political advertising learn from a long tradition of evaluating fairness in commercial advertising law.

* Prof. Natali Helberger, Institute for Information Law (IViR), University of Amsterdam; Tom Dobber, Amsterdam School of Communication (ASCoR), University of Amsterdam; Prof. Claes de Vreese, Amsterdam School of Communication (ASCoR), University of Amsterdam.

The authors thank Sander Kruit and Ljubisa Metikos for valuable research assistance. This project was funded by the Research Priority Area Information & Communication in the DataSociety of the University of Amsterdam and the Dutch Organisation for Scientific Research, grant no. MVI.19.019 (Safeguarding democratic values in digital political practices).

[1] The New York Times, ‘Read the Letter Facebook Employees Sent to Mark Zuckerberg About Political Ads’ The New York Times (28 October 2019) <https://www.nytimes.com/2019/10/28/technology/facebook-mark-zuckerberg-letter.html> accessed 19 March 2021.

[2] https://www.facebook.com/business/help/315131736305613 (last visited on 2 March 2021).

[3] The focus of this article is on online political targeting as a specific form of online political advertising. Since online political targeting is a form of political advertising, at some places in the article we used the notions interchangeably. We depart from the definition of Zuiderveen et al that describe political targeting as “a type of personalized communication that involves collecting information about people, and using that information to show them targeted political advertisements, Frederik J Zuiderveen Borgesius and others, ‘Online Political Microtargeting: Promises and Threats for Democracy’ (2018) 14 Utrecht Law Review 82, 82.

[4] John R Petrocik, William L Benoit and Glenn J Hansen, ‘Issue Ownership and Presidential Campaigning, 1952-2000’ (2003) 118 Political Science Quarterly 599; Lynn Vavreck, The Message Matters: The Economy and Presidential Campaigns (STU-Student edition, Princeton University Press 2009) <https://www.jstor.org/stable/j.ctt7t1g4> accessed 19 March 2021.

[5] Zeynep Tufekci, ‘Engineering the Public: Big Data, Surveillance and Computational Politics’ [2014] First Monday <https://journals.uic.edu/ojs/index.php/fm/article/view/4901> accessed 19 March 2021; Daniel Kreiss and Christopher Jasinski, ‘The Tech Industry Meets Presidential Politics: Explaining the Democratic Party’s Technological Advantage in Electoral Campaigning, 2004–2012’ (2016) 33 Political Communication 544; Daniel Kreiss, Taking Our Country Back: The Crafting of Networked Politics from Howard Dean to Barack Obama (Oxford University Press) <https://oxford.universitypressscholarship.com/view/10.1093/acprof:oso/9780199782536.001.0001/acprof-9780199782536> accessed 19 March 2021; Bruce Bimber, ‘Digital Media in the Obama Campaigns of 2008 and 2012: Adaptation to the Personalized Political Communication Environment’ (2014) 11 Journal of Information Technology & Politics 130; Colin J Bennett, ‘Voter Databases, Micro-Targeting, and Data Protection Law: Can Political Parties Campaign in Europe as They Do in North America?’ (2016) 6 International Data Privacy Law 261.

[6] Normann Witzleb and Moira Paterson, ‘Micro-Targeting in Political Campaigns: Political Promise and Democratic Risk’ in Uta Kohl and Jacob Eisler (eds), Data-Driven Personalisation in Markets, Politics and Law (CUP, Forthcoming 2021) <https://papers.ssrn.com/abstract=3717561> accessed 19 March 2021; Frederik J Zuiderveen Borgesius and others, ‘Online Political Microtargeting: Promises and Threats for Democracy’ (2018) 14 Utrecht Law Review 82; Jeff Chester and Kathryn C Montgomery, ‘The Role of Digital Marketing in Political Campaigns’ (2017) 6 Internet Policy Review; Solon Barocas, ‘The Price of Precision: Voter Microtargeting and Its Potential Harms to the Democratic Process’, Proceedings of the first edition workshop on Politics, elections and data (Association for Computing Machinery 2012) <https://doi.org/10.1145/2389661.2389671> accessed 19 March 2021; European Commission for Democracy Through Law, ‘Joint Report of the Venice Commission and of the Directorate of Information Society and Action Against Crime of the Directorate General of Human Rights and Rule of Law (Dgi) on Digital Technologies and Elections’’ (CoE 2019) CDL-AD(2019)016 <https://www.venice.coe.int/webforms/documents/?pdf=CDL-AD(2019)016-e> accessed 19 March 2021>.

[7] Witzleb and Paterson (n 6); Varoon Bashyakarla and others, ‘Personal Data: Political Persuasion – inside the Influence Industry’ (Tactical Tech 2019) <https://cdn.ttc.io/s/tacticaltech.org/methods_guidebook_A4_spread_web_Ed2.pdf>; Kathleen Jamieson, ‘Messages, Micro-Targeting, and New Media Technologies’ (2013) 11 The Forum.

[8] Judit Bayer, ‘Double Harm to Voters: Data-Driven Micro-Targeting and Democratic Public Discourse’ (2020) 9 Internet policy review; Daniel Kreiss, ‘Yes We Can (Profile You)’ (2012) 64 Stanford Law Review <https://www.stanfordlawreview.org/online/privacy-paradox-yes-we-can-profile-you/> accessed 19 March 2021; Dobber, Ó Fathaigh and Zuiderveen Borgesius (n 45).

[9] Borgesius and others (n 6); Axel Bruns, ‘Filter Bubble’ (2019) 8 Internet Policy Review.

[10] Ira Rubinstein, ‘Voter Privacy in the Age of Big Data’ (2014) 2014 SSRN Electronic Journal; European Commission, ‘Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions - Securing Free and Fair European Elections’ (2018) COM(2018) 637 final <https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52018DC0637> accessed 19 March 2021; Committee of Ministers, ‘Declaration by the Committee of Ministers on the Manipulative Capabilities of Algorithmic Processes’ (Council of Europe 2019) Decl(13/02/2019)1 <https://search.coe.int/cm/pages/result_details.aspx?ObjectId=090000168092dd4b> accessed 19 March 2021.

[11] Witzleb and Paterson (n 6).

[12] Barocas (n 6) 33.

[13] European Commission, ‘Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions – European Democracy Action Plan’ Ares(2020)3624828 <https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=PI_COM%3AAres%282020%293624828>.

[14] Jacquelyn Burkell and Priscilla M Regan, ‘Voter Preferences, Voter Manipulation, Voter Analytics: Policy Options for Less Surveillance and More Autonomy’ (2019) 8 Internet Policy Review 3.

[15] Lynda Lee Kaid and Christina Holtz-Bacha, The SAGE Handbook of Political Advertising (2006) <http://sk.sagepub.com/reference/hdbk_politicaladvert> accessed 19 March 2021.

[16] Zuiderveen Borgesius and others (n 6); see also the extensive comparison in Julian Jaursch, ‘Rules for Fair Digital Campaigning, What Risks Are Associated with Online Political Advertising and What Reforms Are Necessary in Germany’ (Stiftung Neue Verantwortung 2020) <https://www.stiftung-nv.de/sites/default/files/rules_for_fair_digital_campaigning.pdf>; Katharine Dommett, ‘Data-Driven Political Campaigns in Practice: Understanding and Regulating Diverse Data-Driven Campaigns’ (2019) 8 Internet Policy Review.

[17] Burkell and Regan (n 14); Daniel Susser, Beate Roessler and Helen Nissenbaum, ‘Technology, Autonomy, and Manipulation’ (2019) 8 Internet Policy Review <https://policyreview.info/node/1410> accessed 27 November 2020; William A Gorton, ‘Manipulating Citizens: How Political Campaigns’ Use of Behavioral Social Science Harms Democracy’ (2016) 38 New Political Science 61.

[18] Martin Moore, ‘Tech Giants and Civic Power’ <https://kclpure.kcl.ac.uk/portal/en/publications/tech-giants-and-civic-power(b8e837ec-abd8-4838-b8e7-f0059f0de550).html> accessed 19 March 2021; Natali Helberger, ‘The Political Power of Platforms: How Current Attempts to Regulate Misinformation Amplify Opinion Power’ (2020) 8 Digital Journalism 842; Julie E Cohen, Between Truth and Power: The Legal Constructions of Informational Capitalism (OXFORD UNIV PR 2020); Urs Gasser and Wolfgang Schulz, ‘Governance of Online Intermediaries: Observations from a Series of National Case Studies’ (Social Science Research Network 2015) SSRN Scholarly Paper ID 2566364 <https://papers.ssrn.com/abstract=2566364> accessed 19 March 2021.

[19] Helberger (n18); Victor Pickard, ‘Restructuring Democratic Infrastructures: A Policy Approach to the Journalism Crisis’ (2020) 8 Digital Journalism 704.

[20] Joseph Turow, The Daily You: How the New Advertising Industry Is Defining Your Identity and Your Worth (Yale Univ Press 2011); Michael Trusov, Liye Ma and Zainab Jamal, ‘Crumbs of the Cookie: User Profiling in Customer-Base Analysis and Behavioral Targeting’ (2016) 35 Marketing Science 405; Anastasia Siapka, ‘The Ethical and Legal Challenges of Artificial Intelligence: The EU Response to Biased and Discriminatory AI’ (Social Science Research Network 2018) SSRN Scholarly Paper ID 3408773 <https://papers.ssrn.com/abstract=3408773> accessed 26 February 2021.

[21] Cited in: John Miglautsch, ‘Did Direct Marketing Swing the Election?’ (LinkedIn, 22 December 2016) <https://www.linkedin.com/pulse/bigdata-fail-vs-trump-win-john-miglautsch/> accessed 19 March 2021.

[22] Geraint Howells, ‘Aggressive Commercial Practices’ in Hans-W Micklitz and Thomas Wilhelmsson (eds), European fair trading law: the unfair commercial practices directive (Ashgate 2006); Hans-W Micklitz, ‘The General Clause on Unfair Practices’ in Geraint Howells and Thomas Wilhelmsson (eds), European fair trading law: the unfair commercial practices directive (Ashgate 2006).

[23] Marijn Sax, Natali Helberger and Nadine Bol, ‘Health as a Means Towards Profitable Ends: MHealth Apps, User Autonomy, and Unfair Commercial Practices’ (2018) 41 Journal of Consumer Policy 103; Ryan Calo, ‘Digital Market Manipulation’ (2013) 82 University of Washington School of Law Research Paper nr 2013-27 995; Autoriteit Consument & Markt (ACM), ‘Guidelines on the Protection of the Online Consumer’ (11 February 2020) <https://www.acm.nl/en/publications/guidelines-protection-online-consumer> accessed 19 March 2021; BEUC, ‘The Report of the Consumer Law Enforcement Forum CLEF and of the Consumer Justice Enforcement Foirmum COJEF’ <https://www.beuc.eu/general/consumer-justice-enforcement-forum-cojef>.

[24] Art. 2 (d), art. 3 (1) of Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (‘Unfair Commercial Practices Directive’) 2005 [32005L0029].

[25] Ibid.

[26] ICO, ‘Investigation into Data Analytics for Political Purposes’ (6 October 2020) <https://ico.org.uk/action-weve-taken/investigation-into-data-analytics-for-political-purposes/> accessed 19 March 2021; Bennett (n 5); Witzleb and Paterson (n 6).

[27] Daniel Kreiss and Philip N. Howard, 'New Challenges to Political Privacy: Lessons from the First U.S. Presidential Race in the Web 2.0 Era', (2010) 4 Int’l J. Comm. 1032.

[28] Dommett (n 16).

[29] https://www.facebook.com/business/ads/pricing

[30] https://ads.google.com/home/

[31] https://ads.twitter.com/onboarding/18ce5478v4d/welcome

[32] https://business.twitter.com/en/help/ads-policies/ads-content-policies/political-content.html

[33] https://blog.google/technology/ads/update-our-political-ads-policy/

[34] https://www.facebook.com/business/help/167836590566506?id=288762101909005

[35] https://www.facebook.com/business/ads/ad-targeting

[36] Witzleb and Paterson (n 6).

[37] Chester and Montgomery (n 6) 4; Daniel Kreiss and Shanon McGrogor, ‘Technology Firms Shape Political Communication: The Work of Microsoft, Facebook, Twitter, and Google With Campaigns During the 2016 U.S. Presidential Cycle’ (2018) 35 Political Communication 155, 155–177.

[38] Kreiss and McGregor (n 35).

[39] From personal talk with Facebook and interview with the Dutch party D66’s campaign leader; ‘Facebook In Person Marketing Training’ (Facebook for Business) <https://www.facebook.com/business/learn/in-person>.

[40] Barbara Thau, ‘Consumer Goods Industry Trends: How Companies are Driving Product’ (IBM Big Data & Analytics Hub) <https://www.ibmbigdatahub.com/blog/consumer-goods-industry-trends-how-companies-are-driving-product-sales-big-data>; ‘Capgemini Invent UK: BRINGING TO LIFE WHAT’S NEXT.’ (Capgemini UK, 10 September 2018) <https://www.capgemini.com/gb-en/service/invent/>.

[41] Burkell and Regan (n 14) 3; Chester and Montgomery (n 6).

[42] Autoriteit Consument & Markt (ACM) (n 23).

[43] Emilie Barrau, ‘DATA COLLECTION, TARGETING AND PROFILING OF CONSUMERS ONLINE BEUC Discussion Paper’ (BEUC 2010) <https://www.beuc.eu/publications/2010-00101-01-e.pdf>.

[44] European Commission, ‘COMMISSION STAFF WORKING DOCUMENT GUIDANCE ON THE IMPLEMENTATION/APPLICATION OF DIRECTIVE 2005/29/EC ON UNFAIR COMMERCIAL PRACTICES Accompanying the Document COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS A Comprehensive Approach to Stimulating Cross-Border e-Commerce for Europe’s Citizens and Businesses, SWD/2016/0163 Final’.

[45] Calo (n 23); Sax, Helberger and Bol (n 23); Natali Helberger, ‘Profiling and Targeting Consumers in the Internet of Things – A New Challenge for Consumer Law’ in Reiner Schulze and Dirk Staudenmayer (eds), Digital Revolution: Challenges for Contract Law in Practice (Nomos 2016) <http://www.nomos-elibrary.de/index.php?doi=10.5771/9783845273488-135> accessed 27 November 2020; Bram van Duivenvoorde, The Consumer Benchmarks in the Unfair Commercial Practices Directive (2014) <https://dare.uva.nl/search?identifier=1519cbfb-a08a-4132-a207-af6355e53bcd> accessed 19 March 2021; European Commission, ‘COMMISSION STAFF WORKING DOCUMENT GUIDANCE ON THE IMPLEMENTATION/APPLICATION OF DIRECTIVE 2005/29/EC ON UNFAIR COMMERCIAL PRACTICES Accompanying the Document COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS A Comprehensive Approach to Stimulating Cross-Border e-Commerce for Europe’s Citizens and Businesses, SWD/2016/0163 Final’ (n 42).

[46] Susser, Roessler and Nissenbaum (n 17).

[47] Tom Dobber, Ronan Ó Fathaigh and Frederik J Zuiderveen Borgesius, ‘The Regulation of Online Political Micro-Targeting in Europe’ (2019) 8 Internet policy review; Borgesius and others (n 6); Witzleb and Paterson (n 6).

[48] Barocas (n 6) 34 , pointing to the fact that secrecy of the campaign is often considered an important success factor, limiting the incentives for political advertisers to share campaign strategies with voters or third parties.

[49] Julian Jaursch (n 16) 22.

[50] Tufekci (n 5).

[51] Martin Moore and Damian Tambini (eds), Digital Dominance: The Power of Google, Amazon, Facebook, and Apple (Oxford University Press 2018); Gorton (n 17); Chester and Montgomery (n 6).

[52] Tal Z Zarsky, ‘Privacy and Manipulation in the Digital Age’ (2019) 20 Theoretical Inquiries in Law 157; Julian Jaursch (n 16); Burkell and Regan (n 10) 9; Susser, Roessler and Nissenbaum (n 17).

[53] Shaun B. Spencer, ‘The Problem of Online Manipulation’ [2020] Illinois Law Review <https://illinoislawreview.org/print/vol-2020-no-3/the-problem-of-online-manipulation/> accessed 19 March 2021; Muhammad Ali and others, ‘Discrimination through Optimization: How Facebook’s Ad Delivery Can Lead to Biased Outcomes’ (2019) 3 Proceedings of the ACM on Human-Computer Interaction 199:1; Lisa Maria Neudert and Nahema Marchal, ‘Polarisation and the Use of Technology in Political Campaigns and Communication.’ (European Parliament Directorate General for Parliamentary Research Services 2019) <https://data.europa.eu/doi/10.2861/167110> accessed 19 March 2021.

[54] Sax, Helberger and Bol (n 23); Calo (n 23).

[55] CASE OF MARKT INTERN VERLAG GMBH AND KLAUS BEERMANN v GERMANY [1989] ECHR 10572/83 [26], stipulating that information of a commercial nature cannot be excluded from the scope of Art. 10 ECHR.

[56] ibid 33; X and CHURCH OF SCIENTOLOGY VS SWEDEN [1979] ECHR 7805/7 7.

[57] CASE OF CASADO COCA v SPAIN [1994] ECHR 15450/89 [51].

[58] CASE OF MARKT INTERN VERLAG GMBH AND KLAUS BEERMANN v. GERMANY (n 55) para 35; observing that even the publication of items that are true may under certain circumstances be prohibited, e.g. if they fail to respect the privacy of others, the duty to respect confidentiality, but also regarding any false impressions that a message can invoke and that these are factors that national courts can take into account to decide whether statements are permissible or not.

[59] Roger A Shiner, Freedom of Commercial Expression (Oxford University Press) <https://oxford.universitypressscholarship.com/view/10.1093/acprof:oso/9780198262619.001.0001/acprof-9780198262619> accessed 19 March 2021; Reto M Hilty and Frauke Henning-Bodewig (eds), Law Against Unfair Competition: Towards a New Paradigm in Europe? (Springer-Verlag 2007) <https://www.springer.com/gp/book/9783540718819> accessed 19 March 2021.

[60] Rogier de Vrey, Towards a European Unfair Competition Law: A Clash Between Legal Families (Brill Nijhoff 2005) <https://brill.com/view/title/12739> accessed 19 March 2021.

[61] Since advertising regulation in Europe has to a large extent been harmonised, we will concentrate in the following on the relevant European regulatory acts.

[62] Directive 2003/33/EC of the European Parliament and of the Council of 26 May 2003 on the approximation of the laws, regulations and administrative provisions of the Member States relating to the advertising and sponsorship of tobacco products [2003] OJ L 152.

[63] See article 6a of the Directive (EU) 2018/1808 of the European Parliament and of the Council of 14 November 2018 amending Directive 2010/13/EU on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the provision of audio-visual media services (Audiovisual Media Services Directive; AVMSD) in view of changing market realities, OJ L 303, 28.11.2018, p. 69–92.

[64] In particular, note that the national provisions on unfair competition in a number of Member States have traditionally had a double function of protecting consumers as well as competitors (e.g. in the context of the German UWG), but also that, as Henning-Bodewig has pointed out, over the course of time and under the influence of European law a shift in focus on consumer protection has taken place, Henning-Bodewig, 2007.

[65] Directive 2006/114/EC of the European Parliament and of the Council of 12 December 2006 concerning misleading and comparative advertising (codified version) (text with EEA relevance) (henceforth: Directive on misleading and comparative advertising) [2006] OJ L 376/21.

[66] Rules that again make a distinction between truthful and untruthful, fact and opinion, and typically include the possibility for competitors to lodge a complaint, file for an injunction (stop or prevent from doing so in the future) or damages, De Very, 2005, 287 (n 60).

[67] Hugh Collins, ‘The Unfair Commercial Practices Directive’ (2005) 1 417.

[68] Such economic behaviour of consumers can include a broad range of activities along the entire lifecycle of a commercial relationship, from processing advertising and deciding to buy or not buy a product, to using and ceasing to use it, or exercising any contractual rights a user may have, such as compliance with contractual agreements, maintenance, and after sales services.

[69] Thomas Wilhelmsson, ‘Misleading Practices’ in Geraint Howells and Hans-W Micklitz (eds), European fair trading law: the unfair commercial practices directive (Ashgate 2006).

[70] Howells (n 22) 200.

[71] TV VEST AS & ROGALAND PENSJONISTPARTI v NORWAY [2008] ECHR 21132/05 [73].

[72] BOWMAN V THE UNITED KINGDOM [1998] ECHR 141/1996/760/961) [42].

[73] ibid 47.

[74] TV Vest AS & Rogaland Pensjonisparti v Norway [2008] ECHR 21132/05 (11 December 2008).

An argument that, of course, can also be abused. As Turow et al. report, political marketers consistently invoke the US First Amendment as an argument why they fall outside existing regulations about data protection and commercial regulation (Turow et al. (n 72); Ravel (n 59).); Dellebeke 2000, 20.

[75] CASE OF HERTEL v SWITZERLAND [1998] ECHR 59/1997/843/1049 [50].

[76] HANDYSIDE v THE UNITED KINGDDOM [1976] ECHR 5493/72 [49].

[77] Justice Oftedal Broch of the Norwegian Supreme Court, cited in para. 20 of the TV Vest decision.

[78] VEREIN GEGEN TIERFABRIKEN SCHWEIZ (VgT) v SWITZERLAND [2009] ECHR 32772/02.

[79] In the United States, under the First Amendment the barriers to regulation are arguably higher, see e.g. Cohen 2020 (n 18), p. 51: “To regulate those activities would go to the core of the free speech guarantee, by establishing regulations that control viewpoint and are unduly burdensome. Moreover, it would defeat the point of political discussion.” From a US First Amendment perspective,

[80] VEREIN GEGEN TIERFABRIKEN SCHWEIZ (VgT) v. SWITZERLAND (n 78) para 66.

[81] BOWMAN V. THE UNITED KINGDOM (n 72) para 43.

[82] Erdoğan Gökçe v Turkey [2014] ECHR 31736/04 [40]. In a similar vein, Burkell and Regan argue that there are arguments to be made to convey less freedom of expression protection for manipulative speech. Maybe one could argue that also less protection for commercial-political speech, see ECHR Verein gegen Tierfabrieken.

[83] CASE OF ORLOVSKAYA ISKRA v RUSSIA [2017] ECHR 42911/08 [102] (emphasis added).

[84] Ibid, para. 104.

[85] BOWMAN V. THE UNITED KINGDOM (n 72) para 38.

[86] CASE OF DEMUTH v SWITZERLAND [2003] ECHR 38743/97 [45]. Interestingly, the Swiss Federal Council justified their decision to not grant a licence with the need to protect pluralism and the interest of an inclusive general debate: “The result may be the formation of public opinion, influenced by the media by way of specific content, and no longer primarily by way of broadly based, full programs. Such a development would indubitably have consequences for the culture of communication. Communicative integration via the electronic media would be impaired, and would lead to a society increasingly shaped by segmentation and atomisation.”, cited in para. 12.

[87] CASE OF MURPHY v IRELAND [2003] ECHR 44179/98 [2].

[88] Note that unlike the rules on commercial advertising, the regulation of political advertising is largely unharmonised, though the Recommendations of the Council of Europe, Article 3 of Protocol No. 1 to the ECHR and Article 25 (b) of the International Covenant on Civil and Political Rights as well as the Code of Good Practices in Electoral Matters from the Venice Commission have probably had a certain harmonising influence.

[89] A comparative analysis of the rules on political advertising would have gone beyond the scope of this study and would also not have contributed much to the already existing comparative studies. Instead, this paragraph is the result of a review of a number of comparative studies, including Apa et al. (n 37); IRIS, ‘Media coverage of elections: the legal framework in Europe’, (European Audiovisual Observatory 2017) < https://www.ivir.nl/publicaties/download/IRIS_Special_2017_1.pdf>, Raphaël Honoré, ‘ERGA : Report on the Implementation of the European Code of Practice on Disinformation’ (Conseil supérieur de l’audiovisuel 2019) 10:1/6 <http://merlin.obs.coe.int/iris/2019/10/article6.en.html> ; Davor Glavaš, ‘Political Advertising and Media Campaign during the Pre-Election Period: A Comparative Study’ (OSCE Mission to Montenegro 2017); ‘Regulation of Paid Political Advertising: A Survey’ (Centre for Law and Democracy 2012).

[90] One example is France with its Loi relative à la lutte contre la manipulation de l’information. 2018.

[91] CASE OF CASADO COCA v. SPAIN (n 57) para 35; CASE OF DEMUTH v. SWITZERLAND (n 86) para 41.

[92] CASE OF DEMUTH v. SWITZERLAND (n 86) para 41; CASE OF HERTEL v. SWITZERLAND (n 75) para 47; VEREIN GEGEN TIERFABRIKEN SCHWEIZ (VgT) v. SWITZERLAND (n 78) para 71.

[93] Ibid, 73.

[94] VEREIN GEGEN TIERFABRIKEN SCHWEIZ (VgT) v. SWITZERLAND (n 78) para 72.

[95] Ibid, 73.

[96] Maja Brkan, ‘Artificial Intelligence and Democracy: The Impact of Disinformation, Social Bots and Political Targeting’ (2019) 2 Delphi - Interdisciplinary Review of Emerging Technologies 73 <https://cris.maastrichtuniversity.nl/en/publications/artificial-intelligence-and-democracy-the-impact-of-disinformatio> accessed 19 March 2021.

[97] In this sense also Burkell and Regan (n 14).

[98] Bayer (n 8)..

[99] This observation is especially valid in relation to audio-visual media, whose programmes are often broadcast very widely, confirmed in Informationsverein Lentia and Others v Austria [1993] ECtHR 13914/88; 15041/89; 15717/89; 15779/89; 17207/90 [38]; VEREIN GEGEN TIERFABRIKEN SCHWEIZ (VgT) v. SWITZERLAND (n 78) para 73.

[100] Saikat Guha, Bin Cheng and Paul Francis, ‘Challenges in Measuring Online Advertising Systems’, Proceedings of the 10th annual conference on Internet measurement - IMC ’10 (ACM Press 2010) <http://portal.acm.org/citation.cfm?doid=1879141.1879152> accessed 19 March 2021; Balázs Bodó, Natali Helberger and Claes H de Vreese, ‘Political Micro-Targeting: A Manchurian Candidate or Just a Dark Horse?’ (2017) 6 Internet Policy Review <https://policyreview.info/articles/analysis/political-micro-targeting-manchurian-candidate-or-just-dark-horse> accessed 19 March 2021;

Facebook said in October that all content posted by politicians and political candidates, including paid advertising, would be exempt from any of the fact checking for intentionally misleading content, exception: voter suppression: Kate Cox, ‘Misleading Political Ads Are the User’s Problem to Avoid, Facebook Says’ (Ars Technica, 27 January 2021) <https://arstechnica.com/tech-policy/2020/01/misleading-political-ads-are-the-users-problem-to-avoid-facebook-says/ 2/4>.

[101] Burkell and Regan (n 14); Paddy Leerssen and others, ‘Platform Ad Archives: Promises and Pitfalls’ (2019) 8 Internet Policy Review <https://policyreview.info/articles/analysis/platform-ad-archives-promises-and-pitfalls> accessed 19 March 2021.

[102] See also CASE OF DEMUTH v. SWITZERLAND (n 86).

[103] https://www.facebook.com/business/help/975570072950669?recommended_by=167836590566506 ; https://support.google.com/adspolicy/answer/6014595?hl=en

[104] Leerssen and others (n 101).

[105] Iva Nenadić, ‘Unpacking the “European Approach” to Tackling Challenges of Disinformation and Political Manipulation’ (2019) 8 Internet Policy Review <https://policyreview.info/articles/analysis/unpacking-european-approach-tackling-challenges-disinformation-and-political> accessed 25 February 2021; European Commission, ‘European Democracy Action Plan’ Text 2 <https://ec.europa.eu/commission/presscorner/detail/en/ip_20_2250> accessed 19 March 2021.

[106] European Commission, ‘Code of Practice on Disinformation’ (2018) <https://ec.europa.eu/digital-single-market/en/code-practice-disinformation> accessed 19 March 2021.

[107] Leerssen and others (n 101).

[108] European Commission, Recommendation of 12.9.2018 on election cooperation networks, online transparency, protection against cybersecurity incidents and fighting disinformation campaigns in the context of elections to the European Parliament,

https://ec.europa.eu/info/sites/info/files/soteu2018-cybersecurity-elections-recommendation-5949_en.pdf
EDPB, statement 2/2019 on the use of personal data in the course of political campaigns, adopted on 13 March 2019, article 5.

https://edpb.europa.eu/sites/edpb/files/files/file1/edpb-2019-03-13-statement-on-elections_en.pdf

[109] European Regulatory Group for Audiovisual Media Services (ERGA), ‘Report of the Activities Carried out to Assist the European Commission in the Intermediate Monitoring of the Code of Practice on Disinformation (ERGA Report)’ (2019) <https://erga-online.eu/wp-content/uploads/2019/06/ERGA-2019-06_Report-intermediate-monitoring-Code-of-Practice-on-disinformation.pdf>.

[110] European Commission, ‘Assessment of the Code of Practice on Disinformation – Achievements and Areas for Further Improvement’ (2020) Text SWD(2020) 180 final <https://ec.europa.eu/digital-single-market/en/news/assessment-code-practice-disinformation-achievements-and-areas-further-improvement> accessed 19 March 2021.

[111] European Democracy Action Plan: Remarks by Vice-Pres-ident Vera Jourová’ (Brussels, 12 March 2020) <https:// ec.europa.eu/commission/presscorner/detail/en/ SPEECH_20_2308> accessed 19 March 2021.

[112] European Commission, Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on a Single Market For Digital Services (Digital Services Act-DSA) and amending Directive 2000/31/EC, Brussels, 15.12.2020 COM(2020) 825 final.

[113] Proposed Art. 30 DSA, applicable only to so-called Very Large Online Platforms.

[114] European Commission, ‘COMMISSION RECOMMENDATION of 12.9.2018 on Election Cooperation Networks, Online Transparency, Protection against Cybersecurity Incidents and Fighting Disinformation Campaigns in the Context of Elections to the European Parliament’ (2018) C(2018) 5949 final <https://ec.europa.eu/info/sites/info/files/soteu2018-cybersecurity-elections-recommendation-5949_en.pdf>.

[115] Article 29 Working Party, ‘Guidelines For Automated Decision Making and Profiling for the Application of Regulation (EU) 2016-679), finalised on 3 October 2017, last edited and adopted on 6 February 2018 WP251rev.01.

[116] European Commission, Commission guidance on the application of Union data protection law in the electoral context https://eur-lex.europa.eu/legal-content/en/ALL/?uri=CELEX%3A52018DC0638.

[117] Recommendation No. R (99) 15 of the Committee of Ministers to Member States on measures concerning media coverage of election campaigns, 9 September 1999, https://search.coe.int/cm/Pages/result_details.aspx?ObjectID=09000016805e3c6b

[118] Recommendation CM/Rec(2007)15 of the Committee of Ministers to Member States on measures concerning media coverage of election campaigns, https://search.coe.int/cm/Pages/result_details.aspx?ObjectId=09000016805d4a3d.

[119] Ibid.

[120] Council of Europe, Declaration on manipulative capabilities of algorithmic processes (n 10).

[121] ICO, ‘Democracy Disrupted? Personal Information and Political Influence’ (2018) <https://ico.org.uk/media/action-weve-taken/2259369/democracy-disrupted-110718.pdf>.

[122] Art. 11, Titre III : DEVOIR DE COOPÉRATION DES OPÉRATEURS DE PLATEFORME EN LIGNE EN MATIÈRE DE LUTTE CONTRE LA DIFFUSION DE FAUSSES INFORMATIONS (Articles 11 à 15), LOI n° 2018-1202 du 22 décembre 2018 relative à la lutte contre la manipulation de l'information, https://www.legifrance.gouv.fr/loda/id/JORFTEXT000037847559/.

[123] The Law on election and Referendum, http://www.pisrs.si/Pis.web/pregledPredpisa?id=ZAKO4749.

[124] Government of UK (August 2020): Open consultation – Transparency in digital campaigning: technical consultation on digital imprints, https://www.gov.uk/government/consultations/transparency-in-digital-campaigning-technical-consultation-on-digital-imprints.

[125] Government of Ireland (November 2019): Interdepartmental Group on Security of Ireland’s Electoral Process and Disinformation – Progress Report, https://assets.gov.ie/39188/8c7b6bc1d0d046be915963abfe427e90.pdf.

[126] In the Netherlands, the Staatscommissie Hervorming Parlamentair Stelsel did signal the potential positive but also negative consequences of political microtargeting as well as the fact that so far online political microtargeting is unregulated. The committee hence argued in favour of a new law on political parties that would, among other things, tackle political microtargeting,Staatscommissie Parlementair Stelsel and J Remkes, Lage drempels, hoge dijken: democratie en rechtsstaat in balans  : eindrapport (2018).

[127] The Electoral Commission, ‘Digital Campaigning Increasing Transparency for Voters’ (2018) <https://www.electoralcommission.org.uk/sites/default/files/pdf_file/Digital-campaigning-improving-transparency-for-voters.pdf>.

[128] Art. 52 Electoral Code, https://www.legifrance.gouv.fr/codes/texte_lc/LEGITEXT000006070239/.

[129] PROPOSITION DE LOI relative à la lutte contre la manipulation de l’information. 2018.

[130] Art. 1 ibid, amending L. 163-2.-I. of the Electoral Code.

[131] Art. 6 ibid, amending article 33-1 of the law n ° 86-1067 of September 30, 1986 relating to the freedom of communication.

[132] For a comparative overview, see European Regulatory Group for Audiovisual Media Services (ERGA), ‘NOTIONS OF DISINFORMATION AND RELATED CONCEPTS (ERGA Report)’ (2019) <https://erga-online.eu/wp-content/uploads/2021/03/ERGA-SG2-Report-2020-Notions-of-disinformation-and-related-concepts-final.pdf>.

[133] Art. L. 163-1. of the French Electoral Code, to give but one example.

[134] HLEG, A Multi-Dimensional Approach to Disinformation: Report of the Independent High Level Group on Fake News and Online Disinformation (Publications Office of the European Union 2018) <https://ec.europa.eu/digital-single-market/en/news/final-report-high-level-expert-group-fake-news-and-online-disinformation>.

[135] France, L. 163-2.-I. of the Electoral Code.

[136] George Wiliams, ‘Truth in Political Advertising Legislation in Australia’ (1996) Research Paper 13 1996-97 2.

[137] For an overview, Campaign Fair Practices Law (Is There a Right to Lie?', 2014, <https://www.ncsl.org/research/elections-and-campaigns/campaign-fair-practice-laws-is-there-a-right-to-lie.aspx> accessed on 19 March 2021.

[138] Matt Vasilogambros, Political Candidates Don’t Always Tell the Truth (And You Can’t Make Them), 2019 Pew Research < https://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2019/03/21/political-candidates-dont-always-tell-the-truth > accessed on 19 March 2021.

[139] Leerssen and others (n 101); Tom Cardoso, ‘Google to Ban Political Ads Ahead of Federal Election, Citing New Transparency Rules’ The Globe and Mail (4 March 2019) <https://www.theglobeandmail.com/politics/article-google-to-ban-political-ads-ahead-of-federal-election-citing-new/> accessed 19 March 2021.

[140] European Regulatory Group for Audiovisual Media Services (ERGA) (n 132) 41–42.

[141] For an excellent overview, see Center for Information, Technology and Public Life, Platform Advertising, 2020, <https://citapdigitalpolitics.com/wp-content/uploads/2020/05/Platform-comparison-tables.001.jpg> accessed on 19 March 2021.

[142] Art. 2 (d) UCP

[143] Art. 2 (e) UCP.

[144] As hinted at also in the EU Code on Disinformation, calling for the need to develop a working definition on "issue-based advertising" “which does not limit reporting on political discussion and the publishing of political opinion and excludes commercial advertising” European Commission 2021 (n 106), highlight by the author).

[145] Arts. 6 and 7 Unfair Commercial Practices Directive.

[146] Borgesius and others (n 6).

[147] Joseph Turow and others, ‘Americans Roundly Reject Tailored Political Advertising’ (2012) 30 Annenberg School for Communication, University of Pennsylvania.

[148] Tom Dobber and others, ‘Spiraling Downward: The Reciprocal Relation between Attitude toward Political Behavioral Targeting and Privacy Concerns’ (2019) 21 New Media & Society 1212.

[149] Interestingly, the researchers also found that the percentage of voters saying that being targeted would decrease their likelihood of voting for that particular candidate was the highest in the context of social media, and 85% agreed or strongly agreed that they would be angry if they found out that Facebook was sending them ads for political candidates based on profile information they had set to private (Turow et al. (n 147)) (note that the survey took place even before the Cambridge Analytics scandal).

[150] Jamie Bartlett, Josh Smith, and Rose Acton, ‘The Future of Political Campaigning’ (ICO 2018) <https://ico.org.uk/media/2259365/the-future-of-political-campaigning.pdf> accessed on 19 March 2021.

[151] European Commission, ‘Code of Practice on Disinformation’ (n 106).

[152] Council of Europe, Conclusions of the Council and of the Member States on securing free and fair European elections Brussels, 19 February 2019 6573/19

[153] Art. L. 163-1 of Loi n° 2018-1202 du 22 décembre 2018 relative à la lutte contre la manipulation de l'information.

[154] See for the case of commercial practices: Chris Willett, ‘Fairness and Consumer Decision Making under the Unfair Commercial Practices Directive’ (2010) 33 Journal of Consumer Policy 247.

[155] Claire Wardle and Hossein Derakhshan, ‘INFORMATION DISORDER: Toward an Interdisciplinary Framework for Research and Policy Making’ (Council of Europe 2017) DGI(2017)09; HLEG (n 138).

[156] Council of Europe, Recommendation CM/Rec(2018)2 of the Committee of Ministers to member States

on the roles and responsibilities of internet intermediaries (Council of Europe 2018).

[157] Art. 6 (1) Unfair Commercial Practice Directive: ‘A commercial practice shall be regarded as misleading if it contains false information and is therefore untruthful or in any way, including overall presentation, deceives or is likely to deceive the average consumer, even if the information is factually correct, in relation to one or more of the following elements, and in either case causes or is likely to cause him to take a transactional decision that he would not have taken otherwise’.

[158] Tarlach McGonagle, ‘“Fake News”: False Fears or Real Concerns?’ (2017) 35 Netherlands Quarterly of Human Rights 203.

[159] ‘An Update on Our Political Ads Policy’ (n 33).

[160] Associated Press, ‘Facebook Refuses to Restrict Untruthful Political Ads and Micro-Targeting’ (the Guardian, 9 January 2020) <http://www.theguardian.com/technology/2020/jan/09/facebook-political-ads-micro-targeting-us-election> accessed 19 March 2021.

[161] European Regulatory Group for Audiovisual Media Services (ERGA) (n 132).

[162] European Democracy Action Plan 2020 (n 112), p. 5.

[163] Jean-Baptiste Jeangène Vilme and others, ‘Information Manipulation: A Challenge for Our Democracies’ (Policy Planning Staff (CAPS, Ministry for Europe and Foreign Affairs), Institute for Strategic Research (IRSEM, Ministry for the Armed Forces)) <Information Manipulation: A Challenge for Our Democracies> see also section B. IV .

[164] William D Wells, ‘Psychographics: A Critical Review’ (1975) 12 Journal of Marketing Research 196; MC Kaptein, Persuasion Profiling: How the Internet Knows What Makes You Tick (Business Contact Publishers 2015); William A Gorton, ‘Manipulating Citizens: How Political Campaigns’ Use of Behavioral Social Science Harms Democracy’ (2016) 38 New Political Science 61; Burkell and Regan (n 14).

[165] Karen Yeung, ‘“Hypernudge”: Big Data as a Mode of Regulation by Design’’ (Social Science Research Network 2016) SSRN Scholarly Paper ID 2807574 <https://papers.ssrn.com/abstract=2807574> accessed 19 March 2021.

[166] European Commission, ‘COMMISSION STAFF WORKING DOCUMENT GUIDANCE ON THE IMPLEMENTATION/APPLICATION OF DIRECTIVE 2005/29/EC ON UNFAIR COMMERCIAL PRACTICES Accompanying the Document COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS A Comprehensive Approach to Stimulating Cross-Border e-Commerce for Europe’s Citizens and Businesses, SWD/2016/0163 Final’ (n 42).

[167] The authors are well aware that manipulation is a very complex notion and that it would go far beyond the scope of this article to provide a more in-depth discussion, for a possible interpretation in the sense of unfair commercial practice law, based on insights from philosophy, see Sax, Helberger and Bol (n 23). More generally: Susser, Roessler and Nissenbaum (n 17).

[168] Howells (n 22).

[169] Damian Clifford, ‘Citizen-Consumers in a Personalised Galaxy: Emotion Influenced Decision-Making, a True Path to the Dark Side?’ (Social Science Research Network 2017) SSRN Scholarly Paper ID 3037425 <https://papers.ssrn.com/abstract=3037425> accessed 19 March 2021.

[170] Art. 8 Unfair Commercial Practice Directive.

[171] Howells (n 22).

[172] Geraint Howells, Hans‐W Micklitz and Thomas Wilhelmsson, ‘Towards a Better Understanding of Unfair Commercial Practices’ (2009) 51 International Journal of Law and Management 69, 76.

[173] ibid 77.

[174] Willett (n 154) 260.

[175] See Dutch Consumer Authority ACM, ‘Concept Consultatiedocument Leidraad Bescherming van de Online Consument. Grenzen Aan Online Beïnvloeding’ (2019) <https://www.acm.nl/sites/default/files/documents/leidraad-bescherming-online-consument.pdf>.

[176] UK Electoral Reform Committee 2018, ‘Digital Campaigning: increasing transparency for voters’ (Electoral Reform Committee, 2018), 31.

[177] Leerssen and others (n 105).

[178] Borgesius and others (n 6).

[179] Example described in: Joshua Green and Sasha Issenberg, ‘Inside the Trump Bunker, With Days to Go’ Bloomberg (27 October 2016) <https://www.bloomberg.com/news/articles/2016-10-27/inside-the-trump-bunker-with-12-days-to-go> accessed 19 March 2021.

[180] D Sunshine Hillygus and Todd G Shields, The Persuadable Voter: Wedge Issues in Presidential Campaigns (62308th edition, Princeton University Press 2009).

[181] As Jess Garland, director of policy at the Electoral Reforms Society asserts: “These techniques as well as playing into tribalism and polarisation in politics are also moving democratic life outside of our shared public space.” 'With Facebook election ad spend soaring, who is controlling our elections?', 2019, Shropshire Star, <https://www.shropshirestar.com/news/politics/2019/05/21/who-is-controlling-our-elections/> accessed on 19 March 2021.

[182] Compare Unfair Commercial Practices Directive, Annex, No. 11.

[183] Political communication expert Prof. Phil Howard defines political redlining as “the process of restricting our future supply of political information with assumptions about our demographics and present or past opinions”. Philip N Howard, New Media Campaigns and the Managed Citizen (Cambridge University Press 2005) <https://www.cambridge.org/core/books/new-media-campaigns-and-the-managed-citizen/6D88539C6FD25C7026A721DF9C9AC09D> accessed 19 March 2021.

[184] Daniel Kreiss (n 8).

[185] Turow et al. (n 147) 72.

[186] Gorton (n 164); Howard (n 183) 131.

[187] Peter Dahlgren, ‘Doing Citizenship: The Cultural Origins of Civic Agency in the Public Sphere’ (2006) 9 European Journal of Cultural Studies 267; Bernard Manin, Elly Stein and Jane Mansbridge, ‘On Legitimacy and Political Deliberation’ (1987) 15 Political Theory 338.

These practices are potentially also at odds with the principles of inclusive and equal elections, as codified e.g. in the Venice Code of Good Practices in Electoral Matters.

[188] Bayer (n 8).

[189] Burkell and Regan (n 14) 8.

[190] Julian Jaursch (n 16) 29. Google, for example, has already adopted this approach by limiting targeting options to age, gender and postal code.

[191] Lauren Feiner, ‘Google Changed Its Targeting Policies to Shine a Light on Political Ads, but Campaigns Are Now Eyeing More Opaque Platforms’ (CNBC, 8 December 2019) <https://www.cnbc.com/2019/12/08/google-policy-change-has-political-advertisers-looking-elsewhere.html> accessed 19 March 2021.