President Trump displays a signed executive order to improve transparency and promote free speech in higher education in the East Room of the White House on March 21. (Andrew Harrer/Bloomberg)

Like a general always fighting the last war, Donald Trump’s executive order on campus free speech is designed to solve a problem that is already in retreat. Speaker de-platformings, disruptive protests, and restrictive campus speech codes — all of which drove the “campus free speech crisis” narrative in 2016 and 2017 — dropped dramatically in 2018. That’s why the savvier purveyors of the narrative have pivoted to a new and (according to them) more dire threat: student self-censorship.

So the crisis is back on, right? Not exactly.

The evidence for self-censorship

Critics of the state of free speech on campuses point to surveys showing widespread self-censorship among college students. For example, a 2017 survey sponsored by the Foundation for Individual Rights in Education (FIRE) found that although the vast majority of students (about 87 percent) report feeling comfortable voicing their opinions and ideas, many do admit to self-censoring in the past, whether in the classroom (54 percent report this) or elsewhere on campus (33 percent). And there is an ideological gap, as well: Conservative students are more likely to say that they have self-censored than liberal and moderate students. Other surveys tell a similar tale.

But there is more to the story.

Self-censorship on campuses isn’t distinctively prevalent

For one, reports of self-censorship on campus are not much more prevalent than reports of self-censorship among the general public. About 58 percent of Americans admit they stop themselves from saying what they believe to avoid giving offense, according to a 2017 Cato Institute survey. And considering the imposing restrictions that private employers place on what workers can say or do, actual levels of self-censorship could be much higher.

Not all self-censorship is bad

But so what? Even if students are no more likely to self-censor than the general public, this doesn’t let universities off the hook. After all, given the purpose of the university, isn’t any level of student self-censorship unacceptable?

Not necessarily. Using FIRE’s 2017 survey, it turns out that most self-censorship happens for fairly benign (if still perhaps undesirable) reasons. In the classroom, that includes reasons such as not knowing the answer or wanting to fit in. Smaller percentages of students are worried about causing offense or disagreeing with their professor.


Source: FIRE 2017 Speaking Freely Survey

Outside the classroom, the picture is similar. When students self-censor, they usually do so to avoid conflict, look smart or get along with friends.


Source: FIRE 2017 Speaking Freely Survey (FIRE 2017/FIRE 2017)

For those convinced that “political correctness” is stifling campus free speech, these numbers may look like vindication. But there are reasons to interpret them more cautiously.

Colleges and universities aren’t just places where knowledge is produced or transmitted. They are also social spaces in which students make friends, interact with strangers, and try out new identities. In other words, they are spaces where some degree of socialization, peer pressure and (yes) self-censorship is not only expected; it is healthy.

While we should want students to be free to speak their minds, we also want them to develop the skills necessary to navigate complex and diverse environments. This, in part, is what those skills look like.

Do students really need to self-censor to avoid a bad grade?

But hold on. Not all self-censorship may be bad, but some clearly is. For example, fully 21 percent of students who say they’ve self-censored in the classroom report doing so because they fear receiving a lower grade from their professor. And, again, conservatives are much more likely than liberals to report this fear. This is backed up by research from Heterodox Academy, as well as by numerous anecdotes.


Source: FIRE 2017 Speaking Freely Survey (FIRE 2017/FIRE 2017)

This fear is clearly real. It does not, however, have any basis in reality. According to all the available evidence, faculty do not give conservatives lower grades than liberals for equivalent work.

Researchers have tackled this issue from a couple of angles. In one experiment, students were asked to compose two essays, one on the Democratic Party (its values, goals, etc.) and the other on the Republican Party. These were then given to a mix of Democratic and Republican teaching assistants for grading. The students were told that the essays were voluntary and their identities would be kept secret, giving them no reason to self-censor. The result? Neither the partisan affiliations of the students nor of the teaching assistants made any difference in how the essays were graded.

In another study, researchers compared student performance in “liberal” departments (such as anthropology, women’s studies and education) to ones typically viewed as “conservative” (such as economics and business). Drawing on records for thousands of students and after accounting for their ability via SAT scores, researchers found that the more conservative the student, the more likely he or she would receive a good grade in a conservative department.

However, political ideology played no role in how students are graded in liberal departments. In other words, the more left-leaning the department, the fairer the grading. This conclusion is backed up by students themselves, who say on average that Democratic professors are more likely than Republican professors to grade assignments “fairly and consistently.”

A professor’s politics can affect how he or she grades in other ways as well. For example, there is some evidence that Democratic faculty tend to be more “egalitarian” in their grading, while Republicans tend to give more marks at the high and low ends of the grading scale. This study also found that Republican faculty tend to assign lower grades to black students than to white students of equivalent ability. Democrats exhibit this bias, as well, though to a lesser extent.

All of this complicates the traditional narrative of partisanship and grading, and there is still much we don’t know. For instance, it is possible that anti-conservative bias might only manifest itself when faculty are asked to grade assignments with more normative content (e.g. “Should abortion be legal?”). However, at present, there is no evidence supporting the claim that professors penalize conservative students for their beliefs.

Taking stock of self-censorship

So do universities have a student self-censorship problem? Possibly. On the one hand, most students are comfortable expressing themselves on campus, regardless of political party or ideology. And those who do self-censor tend to be trying to fit in, look smart or avoid conflict — hardly the stuff of campus critics’ nightmares.

On the other hand, universities clearly could be doing better at cultivating a culture of free speech, especially in the classroom. In particular, students need to know that their grades do not depend on their politics.

Yet it’s far from clear that Trump’s executive order would solve these problems. Early reactions suggest that the order may only increase polarization on campus and deepen conservative mistrust of faculty. That’s why if positive change is going to happen, it’s less likely to come from outside activists or angry politicians and more likely to come from within the campus community itself.

Jeffrey Adam Sachs is a lecturer in the department of politics at Acadia University.