10/21/25 · Communication

A study confirms the increase in ideological polarization and dissemination of biased or fake news on Facebook

This study, which is groundbreaking in terms of its extensive corpus of analysis and timeframe, examined more than 6 million links to news reports from American media which were shared on Facebook

The period studied (2017-2020) includes several elections, including those that led to the assault on Congress after Trump's defeat, and the Covid pandemic
Facebook

Beyond algorithms, the study warns that polarization is also driven by users’ own behavior, as they tend to consume content aligned with their ideology (photo: Adobe)

A study by the Universitat Oberta de Catalunya (UOC) and Pompeu Fabra University (UPF) has highlighted an increase in ideological polarization and the spread of biased and fake news on Facebook. This groundbreaking study analysed more than 6 million links from news items from 1,231 different domains in the United States which were shared on Facebook between 2017 and 2020.

It examined news reports related to major events which took place during this four-year period, including the Covid pandemic, the 2020 US presidential election (which led to the assault on Congress after Trump's defeat) and the midterm elections (2018), in which all seats in the House of Representatives and a third of those in the Senate were up for election.

“More biased content also tends to come from lower-quality sources”

The study shows how the trend towards increased ideological polarization and dissemination of fake and biased news reports took place at the same time as adjustments to the platform which led to changes in the way the information shown to users is ordered. These changes increased the weight attached to certain types of interactions in the algorithm that determines this ranking. In 2018, more weight was given to shares and comments than "likes" and in 2020, to comments rather than shares. At the same time, these changes also coincided with variations in users' patterns of engagement with content on the network.

However, the researchers point out that it is impossible to establish a direct causal relationship between these events in either of the two cases, 2018 or 2020, and this issue needs to be examined in greater depth in further studies.

The results of the study have recently been published in an open access article in the journal EPJ Data Science. Andreas Kaltenbrunner, the coordinator of the Artificial Intelligence and Data for Society (AID4So) group at the UOC-TECH centre, participated in this research.

 

Engagement, the users' ideology and the quality of sources

The research team examined the engagement level of posts, including more than 6 million links to news items. The engagement level was calculated based on a combination of metrics (clicks, shares, likes, comments and other reactions) and then correlated with users' ideology and the quality of the information sources. This enabled the team to observe the ideological divide between the news reports consumed by conservatives and liberals, as well as its evolution over time.

To facilitate understanding of the patterns of engagement, the researchers describe it as having a "U" shape: it is higher among more ideologically extreme users than among their moderate counterparts. As content with higher levels of engagement tended to attract more visitors after the changes to the algorithm in 2018 and 2020, this could explain the tendency towards an increase in extreme ideological content to suit the user's profile. More biased content also tends to come from lower-quality sources.

In addition to the possible impact of algorithms, the study notes that the polarization is also explained by the behaviour of users themselves, who tend to consume content related to their ideology. The results of the study suggest that since the changes to the platform, the differences between the news diets of liberals and conservatives have become more marked, which makes finding points in common for democratic debate difficult.

 

This study is part of the UOC's research mission covering ethical and human-centred technology and UN Sustainable Development Goal (SDG) 16, Peace, Justice and Strong Institutions

 

Related paper

Fraxanet, E., Kaltenbrunner, A., Germano, F. et al. Analyzing news engagement on Facebook: tracking ideological segregation and news quality in the Facebook URL dataset. EPJ Data Sci. 14, 73 (2025). https://doi.org/10.1140/epjds/s13688-025-00585-3

 

Transformative, impactful research

At the UOC, we see research as a strategic tool to advance towards a future society that is more critical, responsible and nonconformist. With this vision, we conduct applied research that's interdisciplinary and linked to the most important social, technological and educational challenges.

The UOC’s over 500 researchers and more than 50 research groups are working in five research units focusing on five missions: lifelong learning; ethical and human-centred technology; digital transition and sustainability; culture for a critical society, and digital health and planetary well-being.

The university's Hubbik platform fosters knowledge transfer and entrepreneurship in the UOC community.

More information: www.uoc.edu/en/research

Experts UOC

Press contact

See more on Communication