The EU’s Proposed Online Political Advertising Law is a Good Start, But Needs Revising – Center for Data Innovation
The European Commission proposed a new law in November 2021 to regulate online political advertising. The Commission plans to create a harmonized regulatory framework to govern rules around transparency disclosures and audience targeting for ads likely to influence political debates. The proposal is now with the European Parliament, where the Internal Market Committee is currently writing its draft report on the law.
There are three main issues that European policymakers should pay attention to as they negotiate regulations for online political advertising. First, the definition of what counts as a political ad should not encompass purely commercial marketing campaigns, which are already subject to a raft of regulations. Second, lawmakers should properly distribute responsibilities for disclosures and transparency among the different participants involved in online advertising, and not hold advertising platforms responsible for false information provided by their clients. Third, the law should not place excessive constraints on the ability of advertisers or online advertising platforms to target messages to relevant audiences in a cost-effective manner—the very feature that makes targeted online advertising useful.
Online political advertising is a sensitive area given its relationship with the democratic process, and European policymakers are justified in ensuring the rules for fairness and transparency of online political ads are on par with offline political ads. In many cases, industry already exceeds these standards. For example, Google and Facebook provide public repositories of political and issues-based ads. These databases contain all political ads published on their platforms and provide details of who bought them, how much was spent on the campaign, and its reach. The Digital Services Act makes such repositories mandatory for all ads (not just political ads) shown on large online platforms.
The proposed regulation for political ads sets out a host of information that online advertising platforms, like Google’s AdSense and Facebook Ads, would be required to show or make available to an ad’s audience, such as information on ad spending, funding sources, targeting criteria, and the basis for data processing under GDPR. Furthermore, the proposal bans targeting and amplification of political ads using sensitive personal data (as defined by GPDR). The proposed law does provide for exceptions in cases where users provide explicit consent to be shown political ads based on their personal data, or where users regularly interact with political associations or issue-based advocacy groups and non-profit organizations.
Recommendation 1: Draw a Clear Distinction Between Political Advertising and Commercial Advertising
The Commission defines political advertising broadly as the publication or promotion of messages “liable to influence the outcome of an election or referendum, a legislative or regulatory process, or voting behaviour.” This definition creates a wide scope and sweeps up issue-based advertising that may be purely commercial in intent, rather than political. For example, marketing campaigns about diversity and inclusion by commercial actors or retailers that market apparel with political slogans could potentially fall under the scope of the law. The rise of purpose-driven marketing and increasingly conscious consumer behaviour means that growing numbers of brands promote marketing messages with political themes.
Policymakers should maintain a clear distinction between political and commercial advertising and provide guidance on where to draw this line. The new proposed regulation would create extensive disclosure requirements for online political advertising. It would be unnecessarily burdensome to impose these same requirements on purely commercial advertisers. Policymakers should bear in mind that non-political digital ads are already covered by a wide range of laws and regulations. Online ads in the EU must follow rules around accurate information (E-Commerce Directive), target users based on data provided with explicit consent (General Data Protection Regulation, e-Privacy directive), follow rules around unlawful content (Audiovisual Media Services Directive), ensure adverts are accurate and truthful (Consumer Rights Directive, Unfair Commercial Practices Directive), and are not allowed to target minors or sensitive data categories (Digital Services Act).
Recommendation 2: Maintain Intermediary Liability Protection for Online Advertising Platforms Serving Political Ads
Under the proposed regulation, the entities that place a political ad must submit information about how the marketing campaign is funded, the size of its budget and reach, and its targets. As it stands, the draft requires the publishers of the ad to “make reasonable efforts to ensure the information is complete.” But there is no corresponding obligation on clients to provide truthful information. Placing the liability on platforms would allow malicious actors who deliberately submit false information with their ads to avoid facing any sanction. A better approach would be to include a limited “know your business customer” (KYBC) provision that requires online ad platforms to collect certain information about their customers to prevent bad actors from operating under a cloak of anonymity. The purpose of such a provision would be to ensure that regulators could go after bad-faith advertisers that fail to disclose correct information or breach other political advertising rules.
In addition, the regulation foresees only publishers being responsible for deciding whether an ad meets the definition of political advertising, rather than also mandating that advertisers identify whether they consider their ad to be political. This is problematic from a free speech perspective, in that ad platforms should not be alone in policing whether the ads they publish are political. The clients placing the ad should also be required to make good-faith submissions whether they are engaged in political advertising as defined by the law. Such self-declarations will encourage clients to be careful and scrupulous in submitting their campaigns. Concerns around online disinformation by outside actors cannot solely be targeted at the publishers that publish ads: the sponsor of the message, too, must be held to account. Otherwise, the law will fall short of its aim to ensure accountability across the entire political advertising ecosystem.
Recommendation 3: Ensure that the targeting advantages for online political advertisers remain
The draft regulation of political advertising comes in response to widespread concerns around disinformation and manipulation of political ads on the Internet. Yet despite some commentators claiming that online political advertising can subliminally target the innermost reaches of the human mind and thus manipulate unwitting voters, the reality of how online political advertising works is far more mundane. For all the hysteria around Cambridge Analytica’s influence on Brexit, regulators subsequently debunked the notion that it played any role in influencing the outcome of the Brexit referendum. So it is concerning that the law’s preamble refers to the “problematic targeting tactics [of the] Cambridge Analytica scandal,” when the evidence is clear that the firm did not, in fact, have any special mind-targeting tools at its disposal to influence elections. The fact is that Cambridge Analytica’s micro-targeting claims were based on junk science.
The promise of all online advertising, political or not, is that by better targeting the desired audience, an advertiser’s budget is spent more efficiently. It is neither in the interest of advertisers to show, nor their audience to see, irrelevant ads. In the context of politics and elections, this means tailoring messages to likely voters. The 2008 presidential campaign of Barack Obama, for example, made heavy use of digital ‘get out the vote’ strategies, which played a major role in helping a relatively new and inexperienced candidate reach and mobilize swing voters who ended up decisively shaping the election. There is nothing inherently sinister about using technology to fine-tune one’s political messaging or target particular potential voters.
Policymakers should give careful thought as to whether the outright prohibition of targeting audiences based on characteristics like ethnicity, political opinions, religious/philosophical beliefs, or trade union membership is not, ultimately, counter-productive. The purpose of democratic campaigning is to get one’s message out to people that care about it. A total ban on targeting would disadvantage particularly smaller political and advocacy groups that lack the same budgets and resources as large, established parties. For example, candidates that support the labour movement would be limited in their ad effectiveness if they could not target ads—should they want to—to union members.
The proposed regulation for political advertising goes a long way to ensuring that such forms of speech are properly regulated in the EU, with citizens able to discern who is promoting specific ads and why they are seeing them, without undermining the ability to target messages to specific groups. The definition of political advertising needs to be carefully drafted so as to avoid undue overlap with rules around commercial advertising. More work needs to be done in designing the law so that it creates the proper incentives for all actors in the advertising ecosystem to submit truthful and accurate information—otherwise, the law will fall short of its goal of creating fair and transparent rules for all who engage in political ads.