The political effects of X's feed algorithm.

Area of Science:

• Social Sciences
• Computational Social Science
• Political Science

Background:

• Social media algorithms are suspected to influence political attitudes.
• Previous research on Meta platforms showed no political effects from algorithm changes.
• The impact of algorithms on political attitudes remains a key research question.

Purpose of the Study:

• To investigate the political effects of algorithmic content curation on the social media platform X.
• To compare the effects of algorithmic versus chronological feeds on user attitudes and behavior.
• To identify the mechanisms driving any observed political effects.

Main Methods:

• A field experiment was conducted on X with active US users over 7 weeks.
• Users were randomly assigned to either an algorithmic or a chronological feed.
• Political attitudes, online behavior, feed content, and user engagement were measured.

Main Results:

• Switching to an algorithmic feed increased engagement and shifted political opinions towards conservative stances.
• Algorithmic feeds promoted conservative content and demoted traditional media posts.
• Users exposed to the algorithm followed conservative accounts, showing persistent behavioral changes.

Conclusions:

• Initial exposure to X's algorithm has persistent effects on political attitudes and following behavior.
• The algorithm's promotion of conservative content and activist accounts explains the observed effects.
• Unlike previous findings, this study demonstrates a significant political impact of social media feed algorithms.