Unlocking Digital Democracy in the EU

It’s become a truism that digital revolution is changing politics: and it is, of course. But few research groups have tried to measure exactly how. Yesterday, we launched some brand new research into how European citizens and MEPs are using Twitter, which illustrates both the opportunities and pitfalls of digital technology on the way our democracy works.

In e-Democracy in the EU, using Twitter’s Application Programming Interface, we collected tweets in all European languages, sent either to or from MEPs, for one month between 12 March and 12 April 2015. This includes every tweet sent by one of the 504 MEPs whose Twitter accounts we were able to identify and verify when collection commenced, and every mention of an MEP’s username in a tweet posted by any user of the platform during the month. We collected only tweets from public accounts.[i] We then subjected the data to a series of analyses turning on the metadata contained in each tweet.

In total, the 504 MEPs sent 39,556 tweets over the period – an average of 78 each, though some sent just a single tweet during the period and others sent almost a thousand.  When we included all tweets that mention an MEP’s user name (posted by any user of the platform), we found 1,074,910 tweets from 238,974 users. That means that, of those MEPs included in this dataset, each received an average of 2,132 tweets each over this one month, and many of them got significantly more. (Three MEPs received over 100,000 mentions during the period).

Because of the way the data collection works, we anticipate that millions more are expressing their political views on the platform without using the MEP username that we used to identify relevant tweets. Nevertheless, this volume of interactions suggests that Twitter has become a significant site of democratic activism in Europe. They came from every part of Europe, and in every language. What’s more, the tweets were very varied. Some were insults and jeers, of course. But plenty were insightful, thoughtful and meaningful contributions from the public directed at their elected representatives. And too many went ignored.

Add to that a wealth of attitudinal data that shows young people especially consider that the Internet is an important new space where politics happens, and we have – in social media – a new digital commons. An exciting and important one; and also just one of many. As more of us get more involved in online politics – and we fully expect they will – it will become a place that elected representatives need to be, both the listen and to speak. The Internet has transformed our social, personal, professional and economic lives, but the processes of politics and government remain remarkably similar to those of the last century. If voters disengage as a result, democracy will lose its life-blood.

However, despite the hopes of techno-optimists, there are several problems that do still need to be overcome – and we set them out in the report. Most obvious is the volume problem. Some MEPs are drowning in a digital deluge, and with no easy way of finding and acting upon relevant and important digital content. The tendency is of course to them completely ignore it all. There are legitimate worries about how to make sense of all this noise. For democracies to function, politicians have to avoid capture by special interests. New technology can make this more difficult when the intensity and scale of online debates make it hard for political actors to distinguish the signal from the noise.

Social media data of the kind we present above is particularly high in volume and complexity, making it difficult for political actors to gauge when online data reflects broader public opinion, or even just the views of most online users. A seemingly large debate with many supporters of a particular position can instead be the result of a campaign by a PR firm, or a small number of angry people who are very active online. These effects of manipulation and amplification have always affected debates, but they are harder to discern in online activity, so political actors need better ways of telling which communications really reflect the views of many citizens.

None of that takes away an important truth. There is a growing chasm between how people live their lives and how politics work: and we believe technology can bridge that gap.  But there is lots to be done. A good start would be creating a data dashboard that is freely available to all MEPs to better allow them to make sense of the new digital commons. This dashboard needs to be free and easy to use, built according to parliamentarians’ requirements and to ensure that other users’ privacy is respected.

This tool would help elected representatives to collect and analyse online conversations, distinguish genuine users from trolls and special interest groups, to identify their own constituents for priority, and sort through their comments according the topic and type of communication. These sorts of tools are readily available to many large commercial companies and advertising firms: but not to the people who need them most: those that are representing us, and should be the leaders in making sense of our digital voices: not lagging behind the pack.

———

[i] Social media data can be collected manually by copying, screen-grabbing, note-taking and saving web-pages. However, for large volumes of data, the most appropriate method is automatic collection. We used Twitter’s ‘Application Programming Interface’ (API). A portal that acts as a technical gatekeeper to all the data held by the platform. It allows researchers to use their external computer system to communicate with and acquire information directly from Twitter. APIs produce data in a consistent, structured format, and in large quantities. Twitter’s APIs also produce ‘meta-data’, that is, information about the data itself, including information about the user, their followers and profile. This meta-data is valuable to researchers because it often contains information on everything from the sender’s device type, to their account creation date, location and social media following.