Study finds many college students abandon their free speech ideals under ideological pressure

When it comes to free speech on college campuses, students are generally reluctant to punish objectionable statements unless the words are highly severe or target minority groups. New survey experiments reveal that most undergraduates believe marginalized communities deserve extra protection from offensive speech, though these values often waiver when students are pushed by their own strong political ideologies. The findings were published in the journal Science Advances.

Recent political protests have turned universities into highly visible battlegrounds over the limits of acceptable expression. Following the October 2023 Hamas attacks and the ensuing war in Gaza, campus debates focused heavily on whether specific slogans and statements were legitimate exercises of free speech or crossed a boundary into punishable hate. These controversies prompt questions about how young adults interpret the right to speak freely while navigating the need to prevent physical and psychological harm.

Conflicts over speech have drawn national media attention and prompted congressional hearings. Universities often function as microcosms of democratic societies, where competing visions of pluralism and tolerance are publicly contested.

Ran Abramitzky, an economist at Stanford University, led a research team to investigate this tension. Abramitzky and his colleagues wanted to understand how students draw the boundaries of protected expression in real-world scenarios. Past public opinion polls often ask broad questions about offensive speech. Such surveys rarely probe the underlying mechanics of these attitudes, such as whether a student’s judgment changes based on the target of the remarks or the context of the statements.

The debate over speech regulation often centers on two competing philosophies. The first is a universalist approach. This model holds that all individuals deserve equal protection under the law, regardless of their identity or social status. The United States Supreme Court generally follows this doctrine by rejecting hate speech laws that discriminate based on viewpoints.

Free daily newsletter

The second model relies on a particularist approach. This view recognizes that historically marginalized groups may require additional protections from harmful speech. Certain European democracies align with this stance. For instance, bans on Nazi symbols acknowledge the specific harm inflicted on vulnerable communities.

To see how these philosophies play out in a university environment, the researchers surveyed a nationally representative group of 3,065 undergraduate students in July 2024. The team designed three distinct online survey experiments. They separated attitudes toward speech from attitudes toward physical protest tactics, such as building occupations or encampments. This allowed the researchers to measure support for free expression strictly on its own merits.

The first trial was a policy experiment. The researchers asked students if they would favor or oppose a campus rule banning offensive public statements. The target of the hypothetical offensive speech was randomly varied to include white, Black, Jewish, Muslim, or transgender individuals.

The results showed a strong preference for protecting marginalized identities. Students were highly likely to support a policy banning offensive statements directed at Jewish and Muslim people compared to white individuals. The response rates for Jewish and Muslim targets were similar to those for Black and transgender individuals. This pattern indicates that the average student views both Jewish and Muslim people as minority groups deserving of similar levels of protection.

The second trial was described as a professor experiment. Participants read a scenario in which a faculty member made controversial remarks. The researchers randomized the identity of the target group, the context of the speech, and the severity of the statements. Severity ranged from the professor claiming a group plays the victim to get special treatment, to stating the country would be a better place without them. The context varied from a private text message to a public classroom lecture.

Students supported suspending or firing the professor when the speech was perceived to be more harmful. Saying the nation would be better without a specific group increased the probability of disciplinary action by 24 percentage points compared to the lesser offense. Students were also more willing to punish the faculty member when the harmful speech targeted a minority group rather than white individuals. While statements made during a class lecture were perceived as more harmful than private text messages, the context of the speech did not alter the students’ support for punishment.

The third trial was a student experiment. It mirrored the professor scenario but involved a hypothetical student repeatedly making offensive remarks after being warned by administrators. For this test, the target groups were limited to either Jewish or Muslim people. The severity of the statements ranged from questioning a group’s loyalty to calling them the root of all evil.

Again, the severity of the language drove support for punishment. About 17 percent of students supported suspending or expelling a peer for calling a group not loyal. Support for expulsion jumped by 45 percentage points when the hypothetical student called a given group the root of all evil. Participants also classified the statements differently based on severity. Only 11 percent of students labeled the loyalty statement as hate speech, whereas 90 percent categorized the root of all evil remark as hate speech. The researchers found no statistical difference in support for punishment whether the target was Jewish or Muslim.

Following these trials, the team asked participants directly about their views on speech rules. They asked whether rules should consider the target’s identity, such as race or gender. About two-thirds of the students supported the particularist principle, favoring identity-based protections. The remaining third held universalist views, fearing that special protections could lead to the selective suppression of unpopular viewpoints. Both particularist and universalist students placed equal importance on free speech, with roughly 75 percent of each group calling it very important.

When the researchers divided the participants based on these two philosophies, they noticed consistent behavioral patterns. Students who adopted the universalist approach did not support suspending or firing the professor in the second experiment when the target was Muslim or transgender. Universalist students also favored group-specific speech restrictions for white targets at higher rates than particularist students did.

In contrast, the particularist students showed a high willingness to punish faculty for speech targeting all minority groups. These students were highly likely to state that severe statements cause physical or psychological harm when directed at minority populations.

However, the researchers found that students sometimes abandoned their stated principles when ideological pressures surfaced. To test this, the team divided the sample by political ideology and by a continuous sympathy index regarding the Israel-Palestine conflict. They wanted to see if principled commitments to speech regulation held up against political alliances.

When strong social identities came into play, the students’ commitment to their principles weakened. Right-wing universalists expressed less support for punishing a professor who targeted Muslim and transgender people compared to white individuals. Pro-Israel universalists were less supportive of firing a faculty member who targeted a Muslim or transgender student. Pro-Palestine universalists were more likely to support firing professors who expressed hate toward Black or Muslim students.

Students in the ideological center were the exception. Those without strong affiliations to either side of the foreign conflict tended to uphold their free speech principles across all three experiments.

The researchers note some limitations and avenues for future inquiry. The student study specifically targets the university environment, which may not universally reflect the wider public domain. To test for generalization, the researchers conducted a subsequent follow-up study with a sample of United States adults.

The key patterns replicated in the adult sample. Perceived harm and target identity predicted support for restrictions. The adult sample was far more evenly divided between particularist and universalist principles, but ideological commitments still weakened these views exactly as they did in the student group.

Future investigations could look at how disagreements over campus speech evolve in response to changing social and political contexts. By tracking these attitudes over time, scientists can better map how broader societal debates shape the limits of expression.

The study, “Expression at the edge: Free speech boundaries amidst the Gaza crisis,” was authored by Ran Abramitzky, Guy Grossman, Yphtach Lelkes, Hani Mansour, and Tamar Mitts.

Headline options

• College students favor speech rules that protect minorities, survey finds
• Ideology overrides free speech principles on college campuses
• How college students draw the line between protected speech and hate
• Navigating free speech during the Gaza crisis on American campuses
• Are campus free speech values universal or particular to marginalized groups?
• Students want harsher punishments for offensive speech targeting minorities
• War in Gaza tests the limits of free speech on college campuses
• Survey reveals how college students view hate speech and campus protests
• When does offensive speech cross the line for college students?
• Free expression versus harm prevention on polarized university campuses
• Identity and severity shape student support for campus speech restrictions
• How universalist and particularist values clash in campus free speech debates