Research Group > Audiences & Mediated life > News

New research countering the existence of echo chambers

While echo chambers have long been seen as a threat to democratic participation and the informed citizen in today’s high choice media environment, Peter M. Dahlgren finds little evidence of such chambers. Jannie M?ller Hartley and Niklas Alexander Chimirri report from his PhD thesis defense and ask: what does this mean for audience research?

The new digital media landscape has created a high-choice media environment that makes it easier for people to find news and information that support their own political beliefs and attitudes, and to avoid news and information that may challenge those beliefs or attitudes. They may wind up in so called filter bubbles or echo chambers. How do these new possibilities of consuming news affect people’s selection of content and their political polarization in the long run?

This question had been posed by Peter M. Dahlgren in his PhD thesis “Media echo chambers: Selective exposure and confirmation bias in media use, and its consequences for political polarization”, which he defended on January 22, 2021, at the University of Gothenburg, Sweden.

The argument has long been that the current development of the media systems is to blame for people increasingly seeking confirmation of their already existing beliefs. But according to Dahlgren, this phenomenon has to be understood and studied in a much broader perspective:

“Selective exposure is highly connected to the broader political system and development in the country, rather than merely on media consumption patterns,” Dahlgren said in his conversation with opponent Prof. Michaela Meier from the University of Koblenz-Landau, Germany.

Dahlgren draws this conclusion by looking at the relationship between different political preferences (political party, political interest, and ideological leaning) and selective media use over time among the Swedish population, and whether this selectivity leads to political polarization (ideological polarization and affective polarization).

The thesis’ study used longitudinal surveys with a cross-section and panel design, and also a survey experiment. The results suggest that selectivity has not increased to the point that people in general only select information that supports their beliefs or attitudes, nor that people in general necessarily avoid information that challenge those beliefs or attitudes.

Political interest indeed plays an important role in the selection process. But other than initially expected, it is one key motivator for people to select news and information that challenge rather than support them politically: “The more confident you are in your own political beliefs the more likely you are to seek out diverse political views”, Dahlgren underlined.

This and other of his study’s longitudinal insights were also of surprise to Dahlgren himself, as much political communication research had assumed that people engage with news primarily on the grounds of their political interest, and that moreover, people with very selective exposure are also very set in their respective views.

Yet some years ago, Axel Bruns from the Queensland University of Technology similarly concluded that so far, there is little evidence of filter bubbles or echo chambers:

”While such claims may appear intuitively true to politicians and journalists – who have themselves been accused of living in filter bubbles (Bradshaw, 2016) –, the evidence that ordinary users experience their everyday social media environments as echo chambers is far more limited,” Bruns wrote in a study from 2017.

Dahlgren’s empirical work now importantly adds to this evidence, and contributes to a questioning of the somehow inherent bias in current research, as he emphasized during the defense:

“There seems to be a confirmation bias present in political communication research. In the new media environment, people are not necessarily engaging just because they are politically interested. De facto or inadvertent selective exposure is not uncommon.” In a nutshell: selective exposure is far more cognitively and socially complex than hitherto thought of.

Dahlgren accordingly concludes his thesis with an implicit invitation to audience researchers, given that he regards selective exposure as mere first step in a chain of socially embedded selection processes. People’s selective interpretations, as he calls them, should therefore be studied in social groups. For instance, he suggests that “it might be the case that the many choices that the internet gives people do not necessarily cause people to insulate themselves from opposing viewpoints, but instead help organize people into like-minded groups where they can collectively respond to opposing viewpoints” (Dahlgren, 2020, p. 56; see also Dahlgren’s comment in his department’s official news piece: https://www.gu.se/en/news/few-indications-that-the-new-media-landscape-leads-to-increasing-polarisation-in-society). It may be these opposing viewpoints that are selectively interpreted by specific groups, and that these selective interpretations may in turn lead to political polarization.

Thanks to its roots in Cultural Studies, Audience Research has a prominent history of studying interpretations in social groups (e.g., Radway, 1984), and of furthermore connecting these to broader political and cultural structures and processes, such as frameworks of knowledge, relations of production, and technical infrastructures (Hall, 1973). Exploring audience readings of political news in terms of the balance between the forces of hegemony and emancipation, as for instance suggested by Schr?der (2013), may assist in better understanding the media’s old and new roles in the current expressions of political polarization in the US and many other places around the world.