The Matthew Effect of Social Media Algorithms: News Avoidance and Polarisation of News Consumption

This study examines the dynamics of news consumption on social media platforms, focusing on how news avoidance motivations interact with algorithm-driven curation to potentially polarise news consumption. Using a novel experimental platform that mimics a social media news feed with a learning-based recommendation algorithm, we conducted a four-wave field experiment with repeated measures. The findings reveal that the interplay between news avoidance motivations and algorithmic influence results in a widening gap in news consumption between users with high and low levels of unintentional news avoidance, but not between users with high and low levels of intentional news avoidance. These results underline the important role of social media algorithms in generating information deficits among people with higher preference for entertainment than for news. Thereby, the study also demonstrates how experimental social media platforms can illuminate the opaque influence of algorithms, revealing otherwise hidden and inaccessible communication dynamics in today’s hybrid media environment.