Political Issue Attention and the Role of Legacy Media. Analyzing the Trends, Topics and Dynamics in a Hybrid Media Environment

In a hybrid media environment, politicians have gained new ways of reaching their electorate through social media, allegedly circumventing established media gatekeepers. While this story is hardly new, it remains unclear if these digital transformations fundamentally change established media logics and gate-keeping mechanisms. Do social media afford new agenda setting power to politicians, or is their issue attention still defined by thematic agendas driven by traditional media? Do politicians and news media even share the same agenda, or can we witness an increased disconnect of what is deemed important by the different actors, giving rise to potential fragmentation? Not only are these questions still out for debate and beg further empirical evaluation. Particularly little attention has been paid to the different temporalities afforded by social and legacy media, begging the question of how they shape attention dynamics and define the mechanisms unique to hybrid media environments. In order to answer these questions, the proposed paper utilizes the Twitter communication of politicians as a proxy for their issue attention. Two novel datasets of 502.525 Tweets by German politicians elected on a state, federal and European level, as well as 133.554 articles of established German news outlets, both collected between January and July 2021, are analyzed and systematically compared. Using a dynamic graph-based methodology, we identify trending topics in politicians’ tweets and news outlets’ articles. By generating word co-occurrence networks from the documents, we can identify clusters of densely connected words as topics and changes in their network centrality as trends. By employing vector autoregression models on these topics in social and news media, we quantify potential agenda-setting effects between established media and politicians’ issue attention. Of special interest are the timeframes in which politicians and media outlets divert attention to a given issue. These are analyzed by selecting the optimal lag length of the vector autoregression models via statistical criteria, shedding light on the distribution of timeframes in the agenda setting effects between social and news media. Preliminary results show that, rather than uniformly changing the gate-keeping structure to the benefit of politicians, a hybrid media environment affords divergent temporalities in the communication of political issues. While not independent of the established media discourse, politicians aim to divert their electorate’s attention to specific issues in shorter cycles than the established media outlets by using digital communication tools such as Twitter. Not only does this paper introduce a novel graph-based approach towards the study of political issue attention. It also contributes to the ongoing debate on the transformation of mediated publics with a case study on the issue attention and news coverage of Germany’s hybrid media environment.