From activism to polarization: temporal dynamics of #MeToo politicization on Twitter/X

This study is focused on how social media activism is politicized over time, attending to diverse groups of users and their response to real-world events on social media. Extending the scholarship on politicization, we hypothesize three temporal dimensions of this process: partisan voices (1) crowd out activist voices, (2) show sustained participation, and (3) respond to events selectively through partisan references and criticisms. Situating our question in the #MeToo movement, we analyzed 83 million #MeToo-related posts on Twitter/X and comprehensive event data between 2017–2020. We conducted user cluster analysis to identify key user groups and supervised machine learning to detect different discourse categories, followed by formal time-series analysis. Results show episodic, asymmetric, and elite-cued politicization of #MeToo. Partisan voices were more dominant at the initial stage of the movement than during the entire timeframe, but they were always more persistent than activist voices. Also, liberals’ references to conservatives were spurred by conservative culprits, while conservatives’ partisan references spiked when the culprit was either liberal or conservative. Lastly, partisan participation in the discourse was more elite driven than activist participation. We discuss the theoretical and methodological implications for studying the evolution of social media activism in a polarized era.