Dr Luke Temple – written evidence (DAD0048)

This consultation response is written by Dr Ana Langer (University of Glasgow) and Luke Temple (University of Sheffield). Dr Langer is an expert in political communication, with focus on how the use of media impacts on the quality of the democratic process, especially during elections. She has also done work on the use of electronic petitions of different kinds by civil society actors trying to influence policy. Dr Temple is an expert in political participation and electoral geography. The response is primarily focused on the section on technology and engagement and other related questions, and partly draws on a current project funded by a British Academy/Leverhulme grant about how non-party organisations use digital technologies for campaigning

General

Q1. How has digital technology changed the way that democracy works in the UK and has this been a net positive or negative effect?

The ways in which digital technologies have changed how democracy works are countless and often in the background of how organisations (government, civil society and media) work. The most impactful tools are often overlooked in these debates: email, Customer Relations Management Systems (CRM), collaborative tools (especially G suite) and messaging systems such as WhatsApp and Telegram. Social media comes a clear second in many cases.

It is not possible to speak overall of a net positive or negative effect, and it’s probably not all that helpful to think this way; often the same digital tool can be used to aid different agendas in different ways. Generalising, and focusing on social media, the most obvious effect is that they have lowered the cost of engagement and mobilisation for civil society actors both in positive (democratic causes of all kinds) and negative ways (hate groups, incivility, threats to privacy, and misinformation). This allows for a sense of a near-permanent election campaign – however, this is also influenced by the current heightened state of politics.

One major concern is how digital technologies, and especially the role of quasi-monopolies such as Google and Facebook, threaten news organisation business models and therefore their sustainability, especially at the local level. Part of this space has been filled on social media by citizenship journalism organisations with an anti-mainstream media agenda that have considerable biases of their own. At the positive end, the use of electronic petitions (by Parliament but also by other organisations such as change.org and 38degrees) have enabled some citizens to engage with Parliament, the policy process and encourage debate.

Q2. How have the design of algorithms used by social media platforms shaped democratic debate? To what extent should there be greater accountability for the design of these algorithms?

The short answer: enormously and therefore yes, there should be more accountability. Democratic debate is enabled by the production and distribution of high-quality information. An increasing proportion (especially among younger generations) of people access news online, especially via social media and search engines. These platforms dominate the advertising market and hence news organisations’ major source of income (with exception of the BBC). A change in Facebook’s (or Google’s) algorithm, even if minor, has a massive impact on how much and which news people consume as well as, crucially, on the news organisations’ business model.
For these reasons, there should be greater accountability regarding these algorithms. As these are private corporations there are limits to what it is viable, but they are also providing an essentially public service of vital influence in the democratic process. Facebook, whilst originally being resistant, has seemingly come to acknowledge that is a publisher of information and not just a ‘neutral’ platform. This suggests that algorithms act as editors and require thinking about in terms of responsibility and public accountability. However, a possible consequence of this (suggested to us by campaigners we’ve spoken to) is that in response we can expect Facebook to shift emphasis from Public Pages to Closed Groups – therefore moving these debating forums into less visible spaces which are harder to regulate.

Education

Q3. What role should every stage of education play in helping to create a healthy, active, digitally literature democracy?

Fast-paced change in this area can leave citizens unprepared and unaware. We have spoken to non-partisan organisations who focus on political education and technology. A key message from them is shock at the lack of teaching about our political system in primary and secondary schools. Some organisations run schemes going into schools and providing resources to staff to teach about the system more generally, as well as how to navigate the digital side of politics. These organisations provide crucial services but suffer from funding gaps as, despite being non-partisan, they can be viewed as ‘political’ and therefore risky. Therefore, they are established as companies or social enterprises instead (see below on funding). They have also highlighted the low level of digital literacy among teaching staff, which makes these kinds of initiatives harder.

It is important however to note that these issues effect all ages, for instance studies have suggested those most likely to share fake news articles are over the age of 65. Teaching digital literacy to groups outside the education system will be an ongoing challenge as technology continues to develop.

Online campaigning

Q4. Would greater transparency in the online spending and campaigning of political groups improve the electoral process in the UK by ensuring accountability, and if so what should this transparency look like?

Yes BUT rules designed to increase transparency can also have detrimental effects on small organisations, which are crucial for the vitality of civil society. It is clear from interviews we have done as part of our project that these organisations do not have access to the lawyers and accountants often needed to navigate the rules, and hence might opt-out from participating in the electoral process. In contrast, well-funded organisations can afford the advice required to play the rules to their advantage.

Therefore, greater transparency must be accompanied with clear rules and extensive, free, well-resourced advice on how these rules apply to specific cases. Transparency should include:

Add a separate category of digital advertising (and sub-categories of social media, search engine, etc) in the reports to the Electoral Commission
In-house advice from platforms (i.e. someone seconded to e.g. a party during campaigns, even if part-time) to be included as spending ‘in kind’.
Re-think the distinction between constituency and national spending. Digital technologies have made them increasingly hard to distinguish. This is due to how ads are targeted (despite all the talk about micro-targeting a lot of online advertising is not that precise) but also because there are increasing (and democratically legitimate) attempts to encourage tactical voting, which cut across specific constituencies.
Platforms to be legally obliged to include the name of the entity or individual in all social media advertising, regardless of whether it is considered electoral or political. A clear-cut definition of these categories is all but impossible and hence prone to manipulation. If there is a legal obligation for every entity paying for advertising, regardless of categorisation, to be named, then this problem disappears. We acknowledge that this would make the amount of data rather vast, and harder to navigate. But this problem can be minimised using computational methods.
Making users of social media aware of that the ad is micro-targeted and which data was used for doing so – see the work of WhoTargetsMe.
Publicly available data on complaints made to digital platforms (e.g. about targeting, inflammatory speech, dubious organisations) and how it was dealt with.
Complete ad libraries. There have been positive developments in this regard, both for Facebook and Google, but there is need for this to be mandatory and require greater detail about who was targeted, any why (which data it is based on), as well as reach.
Consider restricting the categories of ‘look alike’ that could be regarded as discriminatory / inflammatory, e.g. people who are sympathetic to islamophobia or anti-Semitism.

Q5. What effect does online targeted advertising have on the political process, and what effects could it have in the future? Should there be additional regulation of political advertising?

Researchers are yet to develop a comprehensive evidence base regarding effects. It likely depends on the nature of the advert. We don’t know how effective online political advertising is when it comes to changing someone’s mind about who they might vote for. It might simply remind them to vote for who they were going to vote for anyway, in which case it acts a boost to turnout. However, advertising based on misinformation (for instance an ‘X can’t win here’ message based on false data) could depress turnout. Some organisations we have spoken to believe this is happening (but again, we don’t have evidence of any impact).

It could be beneficial that all political advertising using statistics provides a clickable link to the source – however, the problem with that is that even basic election data in the UK is fragmented and unreliable (see below).

Technology and democratic engagement

Q12. How could the Government better support the positive work of civil society organisations using technology to facilitate engagement with democratic processes?

There are several crucial issues for the sustainability of ‘tech for good’ organisations. We name each below and how they might be potentially tackled:

Funding: the perennial and hard-to-solve problem, and where the state can have only minimal involvement. However, small amounts of funding can go a long way for these types of entities, as they are often partly run by volunteers and have access to highly skilled individuals. Two specific suggestions: provide competitive funding streams, especially for running costs, to try to minimise one of the Achilles heels of start-ups: high rate of emergence but low sustainability. Secondly, make competitive funding available for organisations that provide services that enable the work of others (e.g. open data), including for network organisations. Some of these funding could be tied up with the organisations running digital literacy training.
Issues to do with legal status, which also affects their access to funding. As above, the government should try to increase transparency, clarity and support in regulation about spending as well as about legal status (e.g. what can be done in the political realm while remaining a charity).
Access to data: there has been very good work done in this respect by the government digital service, including for example in relation to Parliament petitions. More could be done, however, including enabling more access to proprietary data as discussed above. Also, encourage investment (for instance via NESTA) on open source data initiatives. A large amount of work done by civic tech groups is attempting to scrape together data (i.e. election results or candidate voting histories) and make it accessible.
Scale and digital distribution costs: as explained in Hindman’s book The Internet Trap, on the Internet ‘seemingly tiny advantages in attracting users can snowball over time’. So, organisations need well-designed and fast-loading platforms, regular content updates, and a large number of visitors. These advantages are also compounded. This makes it very hard for small players to survive and explains why we have quasi-monopolies like Google or Facebook. It also has strong implications for organisations trying to use digital for enhancing democracy. There is not a separate solution as such for this, but it makes even more urgent to deal with the above. It also highlights again the key role that platforms play in the eco-system, and how digital technologies tend to centralise rather than decentralise, as often believed.

Q14. What positive examples are there of technology being used to enhance democracy?

There is lot of important enhancing-democracy work carried out by what the sector calls ‘tech for good’ or ‘civic tech’, by both non-partisan and partisan organisations (‘third parties’ in the parlance of the Electoral Commission). We focus specifically on this type of organisation (all small, some very small) because when discussing digital there is the risk of focusing too much on the technology and not enough on those who enable the use of it in positive ways. Equally, there tends to be a lot of emphasis on what appears as spontaneous mobilisation. We mention here four different kinds to illustrate the diverse ways in which technology can be used to enhance democracy. They are of course by no means exhaustive.

a) Democracy Club (https://democracyclub.org.uk/): create tools to aggregate what they call ‘data for democracy’ (e.g. polling stations, lists of candidates including their social media addresses) and make it widely available. They coordinate a ‘club’ of about 1,000 volunteers working remotely and coordinated by digital tools such as Slack and partly motivated by a digital ‘leader board’. They work on developing databases and apps (e.g. 'Where do I vote?'). They have also created a Wiki page for all candidates. Many civic-tech organisations in the GE2015 and GE2017 used Democracy Club data, both partisan and non-partisan, and so they are crucial in the ‘tech for good’ eco-system and more generally for the campaign environment. They have received grants from the Electoral Commission and Google, among others.

b) Who Targets Me (https://whotargets.me/en/): monitors the use of political advertising on social media. Involves a browser extension on Facebook used to track social media trace data and how it is used for micro-targeting: Facebook has a lot of data about us which is monetised for use by political campaigns but we (users) do not know we are being targeted and why. Who Targets Me analyses the anonymous advertising data crowd-collected by those who install the software and provides it to partnered academics and journalists, helping to provide transparency and increase understanding of how micro-targeting works. Now it is present internationally in up to 60 countries.

c) Stop Funding Hate (https://stopfundinghate.info/): a digital campaign organisationused social media very successfully to help change advertising patterns away from hate inciting outlets, especially Daily Mail, The Express and The Sun. This is done by crowdsourcing customers to put pressure on companies who advertise on these outlets. No boycott of either media outlets or companies involved. This has led to them developing a wider agenda of ‘ethical advertising’.

d) ShoutOutUK (shoutoutuk.org/) / Simple Politics (http://simplepolitics.co.uk/). Two organisations that work to improve political literacy, including going into schools and putting on public events for children.