Yesterday, the UK Advertising Standards Authority’s CEO Guy Parker published a piece outlining some of the challenges of regulating online political ads in the UK (challenges which are also broadly applicable to other democracies).
We agree that it’s difficult to regulate the content of political ads. That’s why we don’t think it’s the answer to this problem. It’s simply too politicised a task for regulators or experts to ever perform in an election campaign, even if the scope of their work was narrowed to examine only the most egregious political lies.
For example, perhaps the most notorious piece of political communication in the UK in recent years — the claim that leaving the EU frees up £350m a week for the NHS — would almost certainly be outside the interest of new regulatory powers. Claims like this may stretch the truth, but they also fall within the bounds of the political speech practiced by campaigns and newspapers for centuries. They need thorough investigation and reporting and when they’re found to be false, rebuttal and defeat. Whether this happens is a wider question of the effectiveness of the system of democratic accountability, not of political advertising regulation.
Furthermore, even if an ad was found false, the act of declaring it so would immediately turn the regulator into a political weapon, as we saw in the 2019 election, when the number of fact checks the Conservatives had ‘failed’ became a key messaging point for their Labour opponents. Neutrality is impossible to sustain.
Over the last few years, we’ve examined tens of thousands of online political ads in many countries. We believe the number of claims a regulator would be able to rule on is trivial and the reach of these ads is tiny, relative to the enormous majority of ads they wouldn’t want to go near. In sum, a political content regulator would be expensive (Facebook is spending $100m to set up its “Oversight Board”) and not at all busy. Given all this, we think a system for verifying claims made in political ads just isn’t viable.
That’s not to say we’ve gone full Zuckerberg, it’s just the wrong lens through which to look at the problem of regulation. Instead, we look at practical constraints on the methods used by political advertisers in order to reduce the ways advertising can be abused. At the same time, the solution must preserve political freedom of expression and the ability to speak to (i.e. target) specific groups interested in a campaign’s message. Finally, it must deliver a system of political advertising in which voters can be confident that advertisers are who they say they are, can be properly challenged on what they say and are likely be held accountable for breaking rules.
These are our ten ideas to achieve these goals:
- Collaborate to define what is ‘political’. Alongside the existing spending-based definitions of ‘political campaigning’ (which should be reformed and retained), an independent expert panel should be convened to develop a broad set of topics and keywords, the use of which in ads would trigger the measures described below. Platforms could use the topics to pre-identify potential political advertisers, and warn them of their verification and transparency obligations (such as being included in ad libraries). This should apply to everyone, including media organisations who pay to promote their news and opinion content (they won’t like it, but can’t reasonably deny the ‘political’ role they play).
- Require maximum transparency for political advertising through excellent, detailed ad libraries, that offer complete data about ads in a standardised format, made readily available to users at source and to journalists and researchers through APIs. Making messaging, spending and targeting 100% transparent increases understanding of political advertising in the short and long term.
- Force strong verification. Voters should be confident who is behind an ad. There have been a number of examples of UK political advertisers operating as ‘fronts’ for unidentified interests and sources of money in recent years. This loophole needs to be closed. Companies accepting political advertising should publish the name and address of the ultimate entity paying for ads and feel confident that the source of their funds is clear. If they cannot establish these facts, the advertiser should be prevented from buying ads until the regulator has investigated.
- Make advertisers earn the ‘right’ to advertise. The proliferation of non-party advertisers, particularly those who didn’t exist before the election period is destabilising. Such advertisers have no reputation on which to be judged. Our data shows that a small number of brand new political Facebook pages were set up and got vast reach for their ads during the 2019 election campaign, then disappeared immediately thereafter. No-one was able to hold these sources of information accountable. New entrants (such as new Facebook pages, or those with existing audiences that change their name) should be subject to strict limits on their spending while they establish their identity, credibility and accountability.
- Allow fewer ads. Campaigns should be able to say what they like, but it’s hard to hold them accountable when they’re saying thousands of things at once. Limiting the number of ads they can run increases the chances of them being held accountable for problematic messages. It also reduces the need for campaigns to collect large quantities of data about individuals for targeting (or microtargeting) purposes. It brings ads back to a scale which the traditional accountability mechanism (journalism) can do its work. The quotas set should vary depending on the type of campaigner. For example, a national party, standing candidates in over 500 seats, might get a quota of 500 ads to run per week, whereas an individual candidate or non-party campaign group could be allocated just 10.
- Make ‘ads’ ads again. Social media advertising blurs the lines between paid and organic content. Creating inflammatory ads for the purpose of them being shared and gaining additional ‘free’ views means that political ads’ content is being led by the incentives of the platform, rather than the value of the arguments they make. No-one has ever received a piece of political direct mail at home, then shared it instantaneously with hundreds of other people. Paid political material should be permanently labelled as an ad, and share and comment features which increase reach of the ads should be disabled.
- Introduce a blackout period for political advertising. Social media ads launched on the day of an election can introduce new, untested and potentially influential claims into a campaign at the precise moment voters are making their final decision. With other sources of media having their own blackout on election day, voters have little chance of establishing whether the claims they see are true. All political advertising, other than ‘get out the vote’ advertising of the format ‘Vote X’, should be banned in the 24hrs before voters go to the polls.
- Ensure these measures are ‘always on’. Political actors seek to shape debates long before election campaigns start. Many, if not all, of the rules above should be applicable all year round, not just in the ‘regulated period’ preceding polling day.
- Enforce the rules and increase the penalties for breaking them. Fines for breaches of electoral law can be seen as a ‘cost of doing business’. This must end. If campaigns abuse the rules the fines levied should be unlimited.
- Update the rules regularly, transparently and accountably. It’s been 20 years since electoral law was last updated. An entire internet and new models of campaigning have emerged in the meantime. The proposals here won’t be perfect and, if implemented, campaigns would still look to stretch them to their advantage. They’ll therefore need to be openly and automatically tested (perhaps by political campaigners, regulators, transparency organisations), and revised on a regular basis.
These ideas would regulate political ads without making judgements on their ‘truth’. While some of the ideas have already been adopted by the platforms, it’s been a patchy, incomplete response to date. Voters and campaigners deserve clear rules, uniformly applied. The good news is we don’t believe these ideas require a significant change to the existing regulatory structure. They could be enforced by the Electoral Commission via an update to the Political Parties, Elections and Referendums Act (2000).
It’s worth noting that nothing here stops platforms going further. If they want to develop their own definition and ban political advertising, fine. If they want to limit targeting options or employ fact-checkers or deplatform campaigns, that’s fine too. Doing so may come at financial or reputational cost, but as private businesses, it’s absolutely their choice.
Together, these ten ideas raise the minimum standard expected of online political ads to increase their transparency, close loopholes that bad actors can abuse and help voters gain a better understanding of the messages that are promoted to them. As such, they’d be a significant improvement to the pre-internet laws we’re currently working under. The question, as always, is whether those in power want to update rules that have benefitted them in the past.