How free is student speech at the University of North Carolina at Chapel Hill? Recently, a trio of scholars released a new report shedding considerable light on this question. Based on a survey of more than 1000 currently enrolled UNC students from across the ideological spectrum, it explores issues related to pedagogy, viewpoint diversity, self-censorship, and political tolerance. It is an impressive document and a valuable contribution to the debate.

Of course, we are not lacking for insights into free speech at UNC. In addition to the survey, we have a pair of reports on campus free speech by the UNC Board of Governors (which under a 2017 state law must be compiled annually), faculty research involving UNC students, and copious reports from the James G. Martin Center for Academic Renewal, a conservative institute based in North Carolina. As a result, we probably know more about the status of free speech in the UNC system than we do about any other state university in the country.

The picture that emerges is, overall, a positive one. Conservative fears of liberal indoctrination notwithstanding, most UNC students say that their professors rarely bring their personal political views into the classroom. Students also express a broad degree of support for free speech and viewpoint diversity, reflecting what one of the report’s authors calls a “hidden cross-ideological consensus.” And while right-wing voices are in the minority at UNC, the vast majority of students – including over a third of liberals – express an interest in bringing more conservative speakers to campus.

Still, not all of the new report’s findings are so positive. A sizeable minority of liberals say that it would be appropriate to deplatform a speaker they consider offensive, a response that The Atlantic’s Conor Friedersdorf says reflects an “authoritarian view.” A large number of students also report self-censoring in the classroom, particularly those who are conservative. And peer pressure exerts a powerful influence over what students are willing to say or do.

These are serious issues. How should we interpret them? And what can faculty do to help?

The Puzzle of Deplatforming

First, an important clarification. On the issue of deplatforming a speaker, Friedersdorf states that 25.5% of students consider such an action appropriate. However, this figure is incorrect. Friedersdorf seems to have simply added the percentage of students in each political category who approve of deplatforming (liberals=19.2%, moderates=3.3%, conservatives=3.0%) without weighting them according to their proportion of the student body. When corrected, the actual percentage of students who approve of obstructing a speaker is 13.2%. In other words, Friedersdorf makes the number of UNC students who hold this “authoritarian view” appear nearly twice as large as it actually is.

But so what. Thirteen percent is better than 25%, but it is still plenty bad, right? Well, that’s a complicated question. According to FIRE’s Disinvitation Database, there has not been an attempted deplatforming at UNC-Chapel Hill since 2009, when a talk by former congressman Virgil Goode was briefly disrupted by hecklers. Since that event, not a single recorded deplatforming attempt has taken place at the school. And this has not been for lack of targets, at least from a liberal point of view. In the last five years alone, Sebastian Gorka, David Horowitz, Corey Lewandowski, Ben Shapiro, Greg Gutfeld, and Christina Hoff Sommers have all spoken on campus. Yet none were deplatformed. None were even heckled. At the Shapiro and Lewandowski talks, students staged a silent walk-out, but that is about as disruptive as anyone got.

So what is going on here? If a substantial percentage of students say obstructing a speaker is appropriate, why do they so rarely act on it?

One possibility is simply that the students do not mean it. Affirming support for disrupting an offensive speaker may be a way of signaling one’s disgust with the speaker’s message. This seems especially plausible given that the UNC survey presented students with vivid, real-world examples of offensive speech. And studies do show that social desirability bias can seriously distort how university students answer survey questions, even when they are anonymous. If liberals attach greater value to signalling their disgust with offensive speech, this could explain why they are more likely than moderates or conservatives to “lie” about their support for deplatforming.

Another possibility is that the price of obstruction is too high. There is probably something to this. Under UNC’s campus free speech policy, students who attempt to deplatform a speaker face punishments ranging up to and including expulsion. And indeed, it was precisely because of this policy that the school’s Black Student Movement chose to stage a silent walk-out at the Lewandowski event, as opposed to something more extreme. However, UNC only adopted this policy in December 2017, so it cannot explain why there were no disruptions in the previous years. And while speakers at some of the events above were protected by security, this can only explain the lack of physical attacks or disruption, and not the absence of non-violent protests or heckling.

Finally, it is possible that students are genuine in their support for obstructing a speaker and do not perceive the costs as too high, but decline to act for strategic reasons. This could include the belief that a disruption will only draw further attention to the speaker’s message, cast the speaker as a victim, or discredit the students’ cause – in other words, create blowback. If true, this is precisely the kind of strategic learning I have argued elsewhere is taking place. Deplatformings, while initially a popular tactic for student social movements, may be growing less attractive due to speakers’ counter-tactics (e.g. invite-only ticketing) and growing opposition to disruptive student activism.

All of these theories deserve greater consideration. Past research shows that a significant minority of students claim to support deplatforming offensive speakers, yet the act of deplatforming itself remains extremely rare. Either offensive speakers are not coming to campus (which seems unlikely) or something else is going on. It is important to figure out what that might be.

Silence and Self-Censorship

The UNC survey also finds that while the vast majority of students report feeling comfortable in class and supported by their instructor, self-censorship does occur. Especially worrying are the number of students who report self-censoring due to being “moderately” or “extremely” concerned that the instructor would give them a lower grade (liberals=1.1%; moderates=8.0%; conservatives=16.9%), that a fellow student would file a complaint (liberals=0.6%; moderates=4.4%; conservatives=16.8%), or that they would lose the respect of their peers (liberals=5.6%; moderates=15.9%; conservatives=46.0%). What should we make of these fears?

We should begin by noting that the first, at least, is misguided. There is no evidence that faculty assign lower grades to students with whom they disagree politically. Furthermore, while some universities have Bias Response Teams to which students can complain about their classmates’ speech, UNC is not one of them. Yet students – and conservative ones in particular – nevertheless perceive these threats to be real. This could be because faculty and students are giving the impression, even if unconsciously, that certain viewpoints will be punished. It could also be the effect of media, some of which tells conservatives that they are a persecuted minority on campus.

So what can faculty do?

In terms of reassuring students that they can speak without losing the respect of their peers, I am not sure there is anything faculty can do. Student attitudes on moral or political issues are generally quite entrenched and tend to resist faculty pressure. Change, if it is to come, will more likely be student-led. And indeed, several promising experiments along those lines are currently underway.

But what about grading? The UNC report suggests that instructors affirm at the start of each semester that they will not assign grades based on politics. It also recommends adopting a standardized grading rubric, thereby minimizing the chance of bias. However, those most convinced of partisan grading may not find these steps reassuring.

For this reason, some scholars are recommending a different solution. Mark Carl Rom and Paul Musgrave, who have studied the issue of biased grading at length, suggest that faculty adopt a “paired assignments” approach. Essentially, students would be required to write two essays on a given proposition, one taking the “pro” position and the other taking the “anti”. Each essay would then be graded and the student’s final mark would be the average of the two. Thus, while a biased professor might still give a student a bad grade on the “liberal” essay, this would be offset by the better grade given to the “conservative” one.

As Rom and Musgrave explain (the precise methods and logical proofs can be found in their article), this strategy does not necessarily eliminate biased grading, but it would reduce it significantly. It would also create a proper ordinal ranking of the students. And perhaps most importantly, it would provide a body of data that faculty could use to examine themselves for bias.

Of course, essays will not be appropriate for every course. And faculty may well balk at the prospect of grading twice the number of papers per semester (though obviously their length can be adjusted). Nevertheless, paired assignments offer a smart solution to a vexing problem: how to persuade even the most skeptical of students that they are really being graded on their merits, and not their politics.

Taking Stock of the UNC

The UNC report provides much food for thought. Its authors are clearly correct when they note that conservative students face special challenges in expressing themselves. Other surveys suggest that this is a national problem, one that many members of Heterodox Academy have no doubt observed first hand. The UNC report is also right to point to the role of social sanction in limiting the scope of campus discourse. With few exceptions, the biggest obstacle to the free exchange of ideas is peer pressure. Until this problem is addressed, some students will continue to feel unfree.

Still, there is much in the report worth celebrating. The vast majority of students want a more open, more diverse conversation on campus. They oppose the most violent or obstructive forms of protest, and would welcome the opportunity to learn new ideas. They are also overwhelmingly pleased with their instructors, including their handling of hot-button political issues. To those looking for proof that academia is broken, you will not find it here.