Transparency in digital campaigning: response to Cabinet Office technical consultation on digital imprints

Questions 1 and 2: Do you agree or disagree with this proposal for the extension of the imprints regime to digital election material? Please provide any further detail to explain your response here. 

We strongly agree with this proposal. We have previously drawn attention to the significant implications of a lack of transparency in digital campaigning. As the consultation acknowledges, data from election and referendum spending returns show a significant growth in digital campaigning in the last decade. While digital communication can be transformative in engaging the voter with campaigning messages, our research with the public has found evidence that significant public concerns about the transparency risks overshadowing the benefits. In our survey after the 2019 UK Parliamentary general election1

we found that:

• Nearly three quarters of people (72%) agreed that it was important for them to know who produced the political information they see online
• Less than a third (29%) agreed that they can find out who has produced the political information they see online
• Nearly half (46%) agreed that they were concerned about why and how political ads were targeted at them. 

Although the majority of more established registered political parties and campaigners follow our guidance to include some form of identifying information on their digital material, there is no legal obligation at present to do so. This means that material can be published and promoted to influence voters without the campaigner ever having to identify themselves as the source. The UK Government’s proposals will help to close this gap in transparency.

The UK Government has also said that it intends to legislate so that all printed election material in Northern Ireland will need an imprint. We strongly support this proposal, as it would address a longstanding anomaly and will ensure greater consistency in the imprint rules. This change would mean that voters in Northern Ireland would be able to see who is responsible for all election or referendum campaign material that is produced by candidates, parties or other campaigners. 

Questions 3 and 4: Do you agree or disagree that this regime will improve the transparency of digital election material? Please provide any further detail to explain your response here. 

We agree that this regime will improve transparency about the source of most digital campaign material aimed at voters in the UK. However, it will not cover all digital material, and we set out our concerns about the proposal to exclude some material in our response to questions 11 to 14.

Questions 5 and 6: What do you consider to be the main benefits of the digital imprints proposal? Do you have any other comments on proposal 1? 

The main benefit of the proposal is that voters, other campaigners and those responsible for enforcing electoral law will be able to more easily identify who is spending money on campaigning at elections and referendums. We believe that this will help to improve public confidence in campaigning.

We also welcome the proposal that the regime should be broad enough to cover all relevant digital technologies and communication channels and able to keep pace with technological change. 

Questions 7 and 8: Do you agree or disagree that the regime should be extended to registered political parties, registered third party campaigners, candidates, holders of elected office and registered referendum campaigners - both paid and unpaid (or ‘organic’) material? Please provide any further detail to explain your response here.

We agree that the requirement should be extended to election or referendum material produced by the specified individuals and organisations. These campaigners are responsible for the majority of the digital campaign material that voters see during elections and referendum campaigns.

We also agree that it should cover material that is produced and distributed through paid or unpaid means by these groups of campaigners. As formally recognised participants in an election or referendum, these campaigners should have to identify themselves as the source of all of their campaign material regardless of whether they have paid to advertise or promote it.

Making this change will reassure the public that these campaigners are not able to run unofficial ‘stealth’ campaigns (alongside their official campaigns) without the law requiring them to identify themselves as the source of that material. We describe an example of recent public concerns about this possibility in our response to question 12. 

Questions 9 and 10: Do you agree or disagree that the regime should be extended to prospective office holders (both paid and unpaid, or ‘organic’, material)? Please provide any further detail to explain your response here. 

We agree that the imprints regime should apply to “prospective office holders”. This would mean that the requirement would apply to candidates, people intending to stand as candidates, and also to other campaigners who share campaign material about these people. It would be help voters to have additional transparency outside a formal campaigning period.

There are complexities to defining when someone is legally regarded as being a prospective candidate for the purpose of the imprint rules. Currently, candidacy is linked to other events in an election timetable, such as when a Parliament dissolves or when a notice of election is published. It can be relevant whether a person has self-declared or been declared by their agent as intending to stand. There are also rules on pre-candidacy spending which apply for several months before UK and Scottish Parliamentary general elections; these concepts could be considered when developing a definition of a prospective office holder. 

Question 11: Do you agree or disagree that the regime should be extended to unregistered third party campaigners promoting paid material only? 

We disagree. We support the proposal that unregistered non-party campaigners should be required to include an imprint on campaign material. However, we recommend that this should apply to any election or referendum campaign material that promotes an outcome, regardless of whether a campaigner has paid to promote or advertise it.

We have several interrelated concerns about the implications of limiting the scope as proposed. These concerns relate to the impact on voters of campaigns that are designed to be shared ‘organically’ or ‘virally’ by supporters, and the impact of algorithms. Under this proposal:

• Unregistered campaigners could coordinate unpaid ‘organic’ campaigns without identifying themselves as the original source of the campaign material.
• Material could be shared widely through organised networks of a campaigner’s supporters, without anyone paying to promote or advertise it.
• Social media platforms’ algorithms could boost distribution to a potentially large number of people, who would have no way to identify the source of that material.
• Existing registered campaigners could run unofficial organic campaigns without revealing that they were ultimately responsible for them – although this would not comply with the imprint requirement, the arrangements could be difficult to trace or prove.
• An unregistered campaigner could spend significant amounts of money producing campaign material but would not be required to include an imprint on the material if they hadn’t paid to distribute it – if we or the police had concerns that a registration threshold or spending limit might have been breached, it may not be possible to contact the campaigner to find out these production costs. 

Question 12: Please provide any further detail to explain your response here. 

We saw some of these kinds of issues at the 2019 UK Parliamentary General Election. Members of the public raised concerns with us about digital adverts and other unpaid material published on Facebook by non-party campaigners which people suspected were front organisations for a political party. However, these were independent campaigners and had no connection with any registered party. An imprint requirement would have shown clearly that a political party was not the source of the material.

At the time of the election, Facebook, Google and Snapchat had put in place their own policies to require political adverts to be labelled with information about their source. Facebook also required labelling on some ‘profile pages’ belonging to organisations. Some of the labels were useful, but others only contained a campaign slogan rather than meaningful information about who paid for an advert or was responsible for the content. This led to members of the public raising concerns with us about who was responsible for this material. The UK Government’s proposals for future elections would go some way to addressing these concerns because adverts would have to include an imprint, but other content such as profile pages would not.

These kinds of gaps in digital imprint rules are likely to reduce the positive impact that these changes have on public confidence in campaigns. There are particular risks if campaign material includes misleading or controversial material with no way for voters or regulators to identify who is responsible for it. Gaps in transparency rules could also be exploited by malicious actors who know how to ‘game’ social media platforms’ algorithms and seek to influence voters anonymously. These issues should be considered further as the Government’s proposals are developed.

Nevertheless, the regime proposed by the Government will represent a significant step forward. It will ensure greater transparency about the source of the majority of digital campaign material that previously would not have been required to include an imprint.  

Questions 13 and 14: Do you agree or disagree with the distinction made in this proposal between paid and unpaid material? Please provide any further detail to explain your response here. 

We disagree. Both kinds of material involve design and production costs, which may be regulated campaign spending under election law. The difference is that ‘paid material’ has also involved the payment of a distribution fee by a campaigner (including paying to target distribution of the advert to particular users). Applying the digital imprint regime to both kinds of material from registered and unregistered campaigners would give voters a key part of the information they need to assess the credibility of campaign information.

It is important that the public, other campaigners and regulators can all find out who is behind both paid and unpaid campaign material – they are both intended to influence the perceptions and choices of voters. It is particularly important that material that has been specifically designed to gain authenticity through organic or viral sharing is clearly labelled with its source so voters can assess its credibility. Public focus group research we published in 20182 showed that participants did not always readily identify material about political or social issues as being linked to a campaign, and that using humorous graphics or video clips made these materials more engaging.

We recognise that the UK Government has proposed this distinction between paid and unpaid material for specific reasons. The consultation paper proposes that members of the public expressing their personal political views in a private capacity should not be required to include an imprint. We agree with this aim, and a regime that does not apply to unpaid material by unregistered campaigners is one way to achieve this.

The Scottish Government has recently published legislation which adopts a different approach for digital imprints at Scottish Parliament and local council elections. These requirements will cover both paid and unpaid material by registered and unregistered campaigners, and will include an exemption for individuals expressing their personal opinion. This builds on the experience of the 2014 Scottish Independence Referendum, when imprints were required on digital materials. We will report on the experience of regulating this requirement at the May 2021 elections. 

Question 15 and 16: Do you agree or disagree that the regime should be expanded beyond what is considered election material (as set out in this proposal), to wider online political advertising? Please provide any further detail to explain your response here. 

We recognise that a wider regime covering online political advertising could support the aims of transparency and voter confidence, and agree that it is a timely moment to ask this question.

Political campaigning can have a wide variety of aims, including supporting and opposing political objectives or supporting and opposing political parties. Election campaigning is a sub-set within political campaigning. Election law covers campaigns that are intended to promote or prejudice the success of an individual or group of parties or candidates an election, and campaigns that promote an outcome in a referendum.

The well-established and clearly defined requirements in election law for printed campaign materials to include an imprint should provide a solid basis for extending the regime to digital election or referendum materials. In contrast, there is currently limited regulation of other forms of political material and advertising in the UK.

We would welcome contributing to discussion and consideration about the scope and impact of such a change, in the event the Government decides to progress development of this policy, and to defining the harm the policy change would seek to address.

We can contribute expertise from regulating the election and referendum imprint rules to any consideration of regulating other forms of political advertising. It would be important to adopt a clear definition of political advertising so that the requirement is clear and workable for all parties, including advertisers and host companies of advertising, as well as from a regulatory perspective.

Questions 17, 18 and 19: Do you agree or disagree that the digital imprints rules should apply to all forms of elections and referendums (beyond those already listed in the proposal and excluding devolved elections and referendums)? Please provide any further detail to explain your response here. Do you have any other comments on proposal 2?

We agree with the principle that voters at any election or referendum should be able to understand the source of the material which is trying to influence how they vote. We recognise that enforcing a digital imprint requirement for local elections such as parish council elections could present additional challenges (given the large number of candidates that can be nominated in these elections, for example). Enforcement of the law for these elections would be the responsibility of local police forces.

Questions 20, 21 and 22: Do you agree or disagree with the proposal on the details to be contained within the imprint i.e. the name and address of the promoter of the material and the name and address of any person on behalf of whom the material is being published? Please provide any further detail to explain your response here. Do you have any other comments on proposal 3?

We strongly agree with the Government that requiring this information will support the dual policy aims of the digital imprints regime. It will provide transparency for voters about who is trying to influence their vote. It will also help the police and the Commission to enforce the spending rules. 

Requiring the name of the promoter and any person on behalf of whom the material is being published means that campaigners cannot simply include a campaign slogan which cannot be traced to an individual. Requiring an address means that we or the police can contact campaigners if there is a need to look into anomalies or concerns with the spending rules or other relevant law, such as defamation law. An address will also help demonstrate that a campaigner is based in the UK and reduce perceptions of foreign interference.

We recognise that some candidates or campaigners who do not have an office address may not want to include their home address on campaign materials. In these cases, a PO Box or mailbox address may still enable us and the police to correspond with a campaigner or serve legal papers, provided it is a genuine address of theirs. 

Questions 23 and 24: Do you agree or disagree with the proposal for the location of the imprint - that the imprint must be located as part of the material where it is practical to do so and where it is not practical, must be accessible from the material? 

We agree that the imprint must be part of the campaign material itself so the public can easily identify who is responsible for it. However, we disagree with the proposal that this will only apply where it is reasonably practical to do so, and where it isn’t, the imprint can be located elsewhere. A ‘reasonable practicability’ provision like this could mean that: 

• Campaigners could remove or alter the imprint from a ‘bio’, ‘profile’ page or a website, at any time. Transparency would be reduced and campaigners would be more difficult to trace after the event.
• Fewer people would see the imprint because some users will be less willing to click on a hyperlink, for example because of the risk of downloading malicious software.
• Campaigners could exploit loopholes, and those responsible for enforcing the law would face uncertainty about what is reasonably practicable in many different individual scenarios.

If such an approach is included, it should be drafted very tightly to ensure that the concept of ‘reasonable practicability’ only applies in exceptional circumstances. The proposal in the consultation document would not incentivise a campaigner to include the imprint on the material itself. It would allow the imprint to be included in different places away from the campaign material, based on the campaigner’s own judgment about what was practicable.

The final proposals should seek to strengthen the onus on campaigners to take steps to include an imprint in the graphics of their material, or to use a platform that offers a clear method to add an imprint alongside an advert or other material published. This kind of approach would act as an incentive for tech companies to continue to make changes to their platforms to facilitate campaigners to include imprints on the material itself, as some have successfully done with voluntary measures to add labels to political adverts. 

Questions 25 and 26: How do you think digital platforms can facilitate campaigners to include imprints? Do you have any other comments on proposal 4? 

In recent years, some of the major digital platforms have required those running political advertising on their platforms and channels to include ’disclaimers’ on those adverts. The disclaimer is like an imprint which provides basic information about the source of political adverts. These initiatives have been a positive step in providing information to the public and are a clear demonstration that digital imprint rules could work in future on these platforms and channels.

As mentioned under Q12, the information provided in these disclaimers has varied considerably. We think more consistent guidance from the platforms, backed by new legal requirements, would facilitate campaigners to include imprints and help avoid confusion over who is responsible for campaign material.

In their voluntary efforts, the companies have devised their own solutions to facilitate political advertisers to include the necessary information about themselves in their disclaimers. Such initiatives include creating new fields in posts for the information. We support these approaches and would look to assist digital platforms with building on their existing experience to develop additional solutions that allow people to view a full imprint as part of the message itself. 

Questions 27 and 28: Do you agree or disagree with the proposal for what the imprint should look like - permanent, embedded and visible/audible, clearly readable/legible/audible and replicable? Please provide any further detail to explain your response here.

We strongly agree that the imprint should be:

• a permanent and embedded part of the campaign material itself;
• clearly readable, legible or audible regardless of the platform or device (e.g. mobile or laptop) it is being accessed on; and
• replicable when re-shared, and must not distort when re-published or shared.

This would mitigate the risk that transparency could be inadvertently reduced, even where campaigners may claim to have complied with the requirement when originally publishing or promoting the material.

Information about the source of campaign material should be accessible to the public, and available before and after an election or referendum. Many people, including journalists, academics, regulators and enforcement bodies, may need to access material and verify its source after the event. 

Questions 29 and 30: What would campaigners need from digital platforms in order to comply with the rules? Do you have any other comments on proposal 5?

Digital platforms should provide campaigners with the tools and functionality they need to easily add a full imprint onto all their campaign material. We would support technical solutions that facilitate campaigners to include full imprints on campaign material. 

Questions 31, 32 and 33: Do you agree or disagree with the proposal for the re-publishing or ‘sharing’ of material? Please provide any further detail to explain your response here. Do you have any other comments on proposal 6?

We agree with this proposal, which would address a potential loophole in the new rules and deter campaigners from publishing or recirculating campaign material without an imprint. 

It will ensure that where campaigners alter other people’s digital material to promote their own campaigns and then republish it, they will have to include their own imprints (provided they are in scope of the new rules). They won’t be able to claim their own imprint isn’t necessary because the original material already has another campaigner’s imprint, or because the rules don’t require an imprint on the original material. 

It will also mean that if a registered campaigner republishes another person’s or an unregistered campaigner’s (unaltered) unpaid material, they will have to put their own imprint on the material. They won’t be able to claim their own imprint is not necessary because the original publisher wasn’t in scope of the rules.
 

Questions 34, 35 and 36: Do you agree or disagree with the proposal that the regime will apply to all election material regardless of where it has been promoted from? Please provide any further detail to explain your response here. Do you have any other comments on proposal 7?

We agree with the principle that all election and referendum material that appears before UK voters should have an imprint, whether the material originates in the UK or abroad. If the imprint rules don’t apply to material that originates overseas, this creates an obvious loophole.

The consultation document notes the challenges of enforcing UK law on campaigners outside the UK who are targeting UK voters. It also proposes a ‘notice and take down’ rule which would give limited responsibilities to digital platforms that host material. Digital platforms will have to remove content without an imprint expeditiously once they are made aware of it. Platforms would commit an offence and could be fined if they fail to do so.  While there will be practical questions to work through with Government, police and digital platforms, we agree that the ‘notice and take down’ proposal seems to provide a practical way of helping to remove non-compliant material and limiting its impact on voters.

We recognise that platforms based in the UK or which have subsidiaries here may be more likely to comply, whilst platforms without a UK presence may refuse to comply with requests to take down material. Enforcing the law on foreign-based platforms may pose the same enforcement challenges as enforcing it on campaigners overseas. However, it is encouraging that we have heard from election regulators in other countries like Australia and Canada with digital imprint regimes that the major digital platforms have cooperated with their enforcement of the rules. We also found that overseas organisations, including international companies and state bodies, have cooperated with us as the UK regulator and complied with UK law at major electoral events. 
 

Questions 37 and 38: Do you agree or disagree that the relevant authorities are in a position to effectively enforce digital imprints? Please provide any further detail to explain your response here. 

It is important for voter confidence that campaigners understand and comply with the new digital imprint requirements, and that voters see non-compliance being addressed swiftly by the enforcement bodies. We partially agree that we are in a position to effectively enforce the digital imprint regime as set out in these proposals. We have previously highlighted concerns about the limits of our existing powers to obtain information as part of delivering our regulatory responsibilities.

As the consultation paper sets out, the police are responsible for enforcing the rules covering campaigning for a particular candidate standing in a particular electoral area. We are responsible for regulating the rules covering campaigning for one or more political parties or groups of candidates, and campaigning at referendums.

The majority of campaigners comply with the current imprint regime for printed material, and we expect that the same would be true with the digital imprint requirement; we will therefore be able to focus our regulatory activity on a minority of cases. We do not want our approach to implementing the new digital imprints regime to rely solely on using our formal investigation powers, particularly where allegations may involve an inadvertent error by a campaigner and especially in a new regime.

Consistent with our approach to regulating all other aspects of the regime, we would seek to secure compliance through support, guidance, swift and early intervention to bring campaigners into line with the rules and, only if necessary, enforcement.  To ensure that we are able to effectively ensure compliance at the earliest opportunity we would want to gather information about potential issues quickly, in order to determine what if any regulatory action is appropriate.  

For example, where we identify material without an imprint we would want to be able to contact the campaigner immediately and give them the opportunity to amend or remove the material to comply with the law. This would mean we could resolve the matter without the need to consider opening an investigation, which would often be a more effective and proportionate approach to securing compliance with the law.

But we can only take this course of action where we are able to identify and contact the campaigner. We have existing powers to obtain information from registered political parties or non-party campaigners to secure compliance with the law, separate from our powers to gather evidence for the purpose of an investigation. However these powers do not extend to third parties such as digital platforms, advertising suppliers or unregistered campaigners themselves. Currently, we can only require information from suppliers or unregistered campaigners if we have reasonable grounds to suspect an offence has been committed and open an investigation.

We will aim to make agreements on a platform by platform basis for them to share information about non-compliant campaigners with us. This would specifically include information that we would need to identify and contact campaigners responsible for producing material that should include an imprint, or companies advertising the material on their platforms. However, we would be reliant on those companies to provide us with information on a voluntary basis, which could reduce our ability to act quickly or proportionately.

To ensure that we are able to effectively secure compliance with the requirements, we would therefore welcome clear powers in the legislation for digital imprints that would enable us to obtain information from suppliers of advertising services and unregistered campaigners outside of an investigation.  Powers to obtain information from third parties (such as digital platforms) would assist many aspects of our regulatory work, but this change would be particularly important in the context of new digital imprints rules.

As mentioned above, the Government is proposing to apply a ‘notice and take down’ rule (which is briefly described in paragraphs 26 and 27 of the consultation document). It will be based on the UK’s existing liability regime for digital platforms. They will be required to remove material without an imprint once they are made aware of it. The rule will cover content originating in the UK as well as from abroad. We support the idea of making it clear what digital platforms providers’ responsibilities are for dealing with material that doesn’t include an imprint, and would welcome further information about how it would work in practice.

We note that this means the proposed regime would have two different routes to securing compliance. Non-compliant material could be removed by a digital platform under the ‘notice and take down’ rule. Alternatively, a campaigner could be investigated for publishing non-compliant material by us or the police. In some instances, both routes could be followed. However, if we are not able to obtain contact details or other information necessary to investigate campaigners, then the regime will rely more heavily on digital platforms taking down non-compliant material, which could have a low deterrent effect against future breaches of the law.
 

Questions 39, 40 and 41: Do you agree or disagree that civil sanctioning powers should be extended for use in relation to offences committed concerning election material in support of candidates? Please provide any further detail to explain your response 

We agree. We support the use of civil sanctions as a proportionate alternative tool to criminal prosecution for dealing with imprint offences. We are experienced in using a civil sanctioning regime for offences under the Political Parties, Elections and Referendums Act 2000 (PPERA).  We have long recommended that an equivalent civil sanctions approach should be put in place for some candidate offences in the RPA 1983.   For administrative or careless breaches of the rules, criminal investigation is a disproportionate step; a civil sanctions regime would be a more proportionate way of dealing with such breaches.

The civil sanctions for offences under PPERA were designed to be an alternative to criminal prosecution and has been successful in delivering transparency of political finance. This experience will be valuable in designing a civil sanctions powers to use in relation to offences committed concerning election material in support of candidates.