Document Type
Article
Publication Date
9-7-2026
JEL Codes
I23; I28; K23; D83
Abstract
This paper asks two questions. Which institutional characteristics are associated with stronger or weaker campus speech climates? And do institutions take corrective action when they are publicly ranked near the bottom? A university faces two costs of restricting speech: a legal cost that binds only public institutions, and a reputational cost is paid when its climate is publicly observed and private institutions also bear this cost. The evidence comes from five waves of the FIRE College Free Speech Rankings, 1,112 institution-waves covering 257 colleges from 2022 to 2026, merged with IPEDS. On the first question, being a public institution is the strongest institutional correlate of a strong speech climate among a plethora of institutional characteristics such as size, selectivity, wealth, religious affiliation and student ideology. On the second, institutions do respond, and the response grows with how conspicuous the bad rank was. Those placed in the bottom quintile of a wave are four times as likely than the best-ranked quintile to upgrade FIRE’s written speech-code rating in the next wave as the low ranking causes reputational damage. Private institutions respond at least as strongly as public ones. Reputational accountability therefore reaches institutions the First Amendment does not. This suggests that broader and more salient disclosure is a market-based alternative to government regulation of campus speech.
DOI
https://doi.org/10.2139/ssrn.7418658
Recommended Citation
Oh, Town and Sellers, Abbylin, What Determines Free Speech on College Campuses? (August 29, 2026). Available at SSRN: https://ssrn.com/abstract=7418658 or http://dx.doi.org/10.2139/ssrn.7418658
Comments
We thank DePauw University’s Asher fellowship that allowed us to initially undertake this project. We are grateful to the Janet Prindle Institute for Ethics and to the Civil Discourse Project seminar at Duke University, in which this collaboration began. We also thank the Foundation for Individual Rights and Expression for providing the anonymised records of policy-revision requests analysed in Section 6.5.