The Impact of Intermediaries on Audience Fragmentation
Final Report Abstract
Contemporary media systems are facing profound changes, including a progressive transition towards digital communication channels. An important consequence is the rise of a new kind of influence on news exposure: Popular intermediaries such as Facebook and Google guide news consumption through recommendations. Since recipients have been quick to integrate these novel pathways for news in their repertoires, the consequences of their algorithmic recommendation engines have been the subject of controversial debates. On the one hand, intermediaries empower recipients to choose from a large choice of sources, selecting their individually preferred news content. This trend towards personalization, on the other hand, lends new power to intermediaries, whose recommendations may influence selectivity. This power (if wielded irresponsibly) could have harmful consequences – from discrimination across manipulation and abuse of market power to privacy infringement. Effects of algorithmic personalization have two potential opposing consequences: They may unite citizens in their use of news and thus their perception of the world around them, given that individual interests are similar, mass media retain their rallying role and/or if intermediaries recommend similar items. If, on the other hand, people have strongly diverging interests, feel less inclined to follow and trust established mass media and if intermediaries recommend highly personalized content, then the result is a growing divergence of the perception of reality. This fragments formerly connected audiences into separate parts, within which distinct opinions may become dominant. Under such a scenario, a decrease in societal cohesion would bode ill for crucial democratic institutions and mechanisms such as identifying pressing issues and resolving conflicts. In pursuit of its main goal, the project developed and implemented a large-scale empirical study to offer evidence towards the existence of both such positive (unifying) and negative (fragmenting) effects of intermediaries. The key determinant for this distinction is what appears to be the intentionality of use: Recipients who regularly and intensely rely on intermediaries see a less diverse selection of news, whereas an intermittent usage is associated with a more diverse exposure to news. The curating logics of intermediaries thus exhibit a varying influence, which is dependent not just on their immanent (and technical) design, but also on users' behavior. A second important finding relates to the network structure of German news audiences. In contrast to fears raised by, e.g., the narrative of "filter bubbles" or the trend of political polarization plaguing many western democracies, there is no large-scale split between outlets' audiences. Instead, a strong and dense core of mainstream outlets share large parts of their audience, indicating that the unifying effect of mass media remains in effect. At the same time, however, a large share of recipients are only weakly connected and thus form an apparently eroding fringe. To prevent a further erosion of audiences, media organizations, including public service broadcasters, should offer journalistic quality content tailored towards winning over audiences, which have since turned away from mainstream news. A balance clearly needs to be struck between the breadth of supply, its quality and tailoring of content to ensure that automated curation mechanisms remain in check. Future research must furthermore focus on the precise nature of selection decisions to enhance our understanding of the individual mechanisms and interplay of curation mechanisms, both algorithmically and intrinsically. Precise models of such decisions may then offer reliable guidance for regulatory measures aimed at safeguarding the diversity and openness of public debates. The project's novel findings serve as a crucial milestone in our field's understanding of the role that intermediaries play in shaping audiences and their networks. They also contribute a much-needed empirical base for the ongoing societal debate about the regulation of these corporations and their platforms, facilitating evidence-based policy making. Both existing and new-founded cooperations with public and private institutions as well as regulatory bodies provide policymakers with direct access to these valuable insights.
Publications
- (2017). The Power of Default on Reddit: A General Model to Measure the Influence of Information Intermediaries. Policy & Internet, 9(4), 395–419
Jürgens, Pascal, & Stark, Birgit
(See online at https://doi.org/10.1002/poi3.166) - (2019). Two Half-Truths Make a Whole? On Bias in Self-Reports and Tracking Data. Social Science Computer Review
Jürgens, Pascal, Stark, Birgit, & Magin, Melanie
(See online at https://doi.org/10.1177/0894439319831643) - (2020): The Centre Cannot Hold? Information Intermediaries and Audience Fragmentation. Dachschrift Kumulative Dissertation, Johannes Gutenberg-Universität Mainz
Jürgens, Pascal
- (2021). Loopholes in the echo chambers: How the echo chamber metaphor oversimplifies the effects of information gateways on opinion expression. Digital Journalism
Geiß, Stefan, Magin, Melanie, Jürgens, Pascal & Stark, Birgit
(See online at https://doi.org/10.1080/21670811.2021.1873811) - More of the same! – Assessing the role of intermediaries on exposure diversity in German news repertoires. 72. Jahrestagung der International Communication Association (ICA), 26.–30. Mai 2022, Paris
Jürgens, Pascal, Stegmann, Daniel, & Stark, Birgit
