Written by Philip Boucher,
With the Cambridge Analytica scandal, it became clear how technologies such as social media and techniques such as psychological profiling can be combined in election campaigns with worrying effects. Digital forms of personalised political messaging can be highly automated. They start and end with social media, which provides both the data for categorising users and the medium for targeting them with personalised messages. Messages might be designed to favour a particular candidate or to encourage widespread discord and mistrust. In either case, it could lead to more polarised societies in which citizens share less common ground and are less understanding of those with different political ideologies, attitudes to populism, or perspectives on specific topics such as immigration.
These same technologies and techniques also shape trends in news production and consumption. As newspaper sales dwindle, outlets increasingly rely upon advertising revenue generated by clicks, making extensive use of social media platforms and user profiling. Public debate increasingly occurs via these social media platforms in which citizens, politicians, companies and bots communicate directly to each other without the traditional filters of journalistic standards and editorial oversight. It has been suggested that, where citizens increasingly rely on such platforms for news, they risk entering ‘filter bubbles’ in which they are exposed to a narrow range of perspectives oriented around their own profiles, shielded from contrasting views, in a broad trend that could also lead to more polarised societies. In this context, STOA launched two studies to explore the mechanisms by which these technologies and techniques may foster polarisation in Europe, and published an accompanying Options Brief.
One study, conducted by Richard Fletcher and Joy Jenkins of the Reuters Institute for the Study of Journalism at the University of Oxford, considered the effects of technology on news production and consumption across Europe and their potential to lead to more polarised societies. One of its key messages is how little we understand about the mechanisms that link news production and social polarisation. The internet has created more consumer choice, to the point where most people select their own news sources based on their ideologies and preferences. Yet, the review found little evidence to support the ‘filter bubble’ thesis, or that exposure to populist material has a significant effect on citizens with mainstream views. However, there are key exceptions to these findings at the fringes, with evidence that people who already hold extreme ideological views or attitudes to populism tend to develop even stronger perspectives when exposed to news with which they either strongly agree or strongly disagree. The authors suggest that individuals’ basic interest in current affairs is a key factor as – in today ‘s high-choice media environment – some users may opt-out of news consumption entirely. Such news aversion could be a worrying trend if healthy democracies rely upon citizens understanding their political system.
The other study was conducted by Lisa Maria Neudert and Nahema Marchal of the University of Oxford, and focused on trends in political campaigning and communication strategies. It highlighted a trend towards more emotionally charged content – particularly negative material that provokes fear, hatred or disgust – in political communications. While such highly charged and targeted messages may be effective, they can also escalate mistrust and tensions between groups with different perspectives and, thus, foster social polarisation. The review also highlighted that some ‘clickbait’ based on political issues may be designed for purely financial purposes, but have the side-effect of increased polarisation. In other cases, polarisation has been the deliberate aim of manipulative political campaigns by hostile foreign and domestic political actors, making use of automated bots and ‘dark ads’ to amplify disagreement, provoke hostility between different groups, and undermine social cohesion.
Hasty policy action that attempts to control communications directly – for example by restricting some media content or political expression – could do more harm than good, and could even have ‘chilling effects’ on democracy. However, both studies present policy options that could help to foster healthier digital environments and mitigate trends towards social polarisation. These are combined and further developed in the STOA Options Brief, which includes options targeting news consumption, digital divides, political communications, news producers and governance institutions.
The authors of both studies presented their work during the STOA Panel meeting on 14 March 2019, which can be viewed here.