Regulating Social Media. A Relational Analysis of Policy Discourse in German Policy-Making

Digitalisation and the emergence of social media as new arenas of political discourse confront policy-makers with multiple challenges. One of them is law enforcement in a global and highly decentralized environment. While Western democracies refrained from regulating the Internet for a long time, a recent shift towards more state-sponsored control fosters the development and discussion of a wide range of possible governance regimes. The recently passed German “Network Enforcement Law" (NetzDG) exemplifies a governance system of regulated self-regulation by obliging social network operators to delete user-generated posts with criminal content (e.g. “Hate speech”) under the threat of a high penalty. As one of the most far-reaching regulations in this policy domain, the formulation of the law was accompanied by a lively debate, involving multiple governmental and non-governmental actors with diverging positions on issues like free speech, the general legitimacy of internet regulation and the accountability of private actors’ regulatory activities. By investigating this discourse, we aim to map the complex interests and belief constellations associated with this policy issue. More specifically, our research aims to identify (1) who is involved in the policy discourse, (2) what policy positions and arguments are used to promote interests and ideas and (3) which actor-coalitions are observable on basis of ideational and interest similarities. Furthermore, by comparing various venues of debate over multiple periods of time, this paper uncovers varying constellations of discourse and actor-coalitions (4) at different stages of policy-making and (5) in different media environments. The investigation period of the paper is set between March and September 2017, thus capturing the whole period from the first draft of the law till it entered into force. Data was collected by inductively coding each contribution to the debate in print media, social media and parliamentary text records and linking the respective statement’s general concept to its author and his organizational affiliation. This data allows to analyze the discourse within this policy process in a relational perspective, using a combination of qualitative content analysis and methods of social network analysis applied to print media, social media and parliamentary text data. This new method (discourse network analysis) traces issue and actor networks in synchronical and diachronical ways to discover structural cores as well as shifts and punctuations in policy processes. It is particularly fitted for the present policy issue as it allows to clarify informal and possibly unconscious relationships between different types of actors, to display the complex overlapping of different issues and belief-systems and to depict the multiple levels of discourse-participation and actor-identity.