Germain Gauthier (Bocconi University)

The Political Effects of X’s Recommender Algorithm

Abstract: We examine the impact of the recommender algorithm of Elon Musk’s social media platform X (formerly Twitter) on political attitudes. In a field experiment conducted in 2023, we randomly assigned active U.S.-based users to either an algorithmic feed or a chronological feed for seven weeks, measuring their political attitudes and online behavior. Several key findings emerged: Switching from the chronological to the algorithmic feed increases user engagement and shifts political opinions toward more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump, and views on the war in Ukraine. Using NLP methods to analyze the content of users’ feeds, we confirm that the algorithm promotes more conservative content. In contrast, we do not find symmetrical effects among users randomized to switch from the algorithmic feed to the chronological feed, suggesting that exposure to recommender algorithms persistently affects political attitudes. We provide evidence for the mechanism behind this effect: exposure to algorithmically curated content induces users to follow new accounts, which they continue to follow even after the algorithmic feed is switched off.