The Credibility Cost of Scientists’ Political Expression on Social Media

Abstract Scientists are increasingly visible on social media, and much of what they post touches politically salient issues. Drawing on our recent work (Alabrese et al. 2024. “Politicized Scientists: Credibility Cost of Political Expression on Twitter.” CAGE Working Paper 735), which combines large-scale observational data on nearly 98,000 U.S. academics on Twitter/X, pre-registered experiments with the U.S. public and with international journalists, and two surveys of scientists, we highlight a trade-off between visibility and credibility. Political expression is widespread among academics – midway between ordinary users and elected politicians – and consists mostly of opinion rather than verifiable claims. Audiences penalize it: perceived credibility declines monotonically as political signals become more explicit, the largest losses arise under ideological mismatch, and a research post on a politically charged topic moves credibility about as much as a contentless partisan cue. Scientists anticipate these costs and, if anything, overestimate them, inviting a spiral of silence, while expecting professional gains in media and policy arenas. The policy question is not whether scientists should speak, but how universities, funders, and platforms shape the terms on which they do.