Recently, Tom Simpson of Oxford University and myself co-authored a report, Academic Freedom in the UK for the British Think Tank Policy Exchange. The report garnered considerable attention in the British media, largely positive (see here, here, here, here for instance), with a couple critical — if revealing — comments as well (e.g. here, here).
The report is intended as a first step to move the UK debate on the threat to academic freedom forward – from high-profile anecdotes to a more systematic, representative, data-driven approach.
We also seek to get beyond diagnosis to concrete policy solutions. For, we believe that while it is vital to win the cultural argument, this is insufficient if institutions are captured or under the sway of norms set by an influential progressive minority. Thus, we argue, it is only by fine-tuning the institutional and legal machinery of government that the challenge to academic freedom, and the related downward spiral of viewpoint diversity, can be addressed.
UK Student Opinion: A House Divided
The report is divided into two main sections, one focused on student opinion, another on policy. For the first, we gathered a convenience sample of 505 UK undergraduate students aged 18-25 on the Prolific Academic crowdsourcing platform (a UK equivalent to Amazon’s MTurk). Such platforms skew young, liberal and tech-savvy, but offer a fairly good picture of young, educated opinion (Huff & Tingley 2015). We are also able to control for demographics, notably gender. The aim was to try to move beyond abstract support for free speech to explore student views on concrete high-profile cases and to examine what happens when the aims of free speech collide with an activist version of social justice. We also wanted to see how readily student opinion could be shifted by reading pro-free speech or pro-emotional safety passages.
In keeping with a previous survey of over 1,000 UK undergraduates (Hillman 2016), we find that there are three groups of students: one group consistently supporting emotional safety over freedom, another backing free speech over safety, and a significant undecided group.
Hillman’s study for the Higher Education Policy Institute found that 60 percent of students agreed universities ‘should never limit free speech’ and only 11 percent disagreed. By a 45-23 margin, students agreed that universities shouldn’t be comfortable, but rather places for debate and challenging ideas. Yet when the survey introduced a tradeoff between free speech and sensitivity to minorities, sensitivity often took primacy. For instance, 30 percent of students backed the idea that university publications, if ‘offensive to certain groups of students’ should be censored, with only 34 percent opposed. By a 45-17 margin, they agreed that ‘ensuring the dignity of minorities can be more important than freedom of expression’.
As important as the tilt of opinion is the considerable group of undecided students. Hillman’s work also noted an important gender split, with women 5-20 points more supportive of sensitivity-over-free speech positions than male students.
The results of our survey are summarized in figure 1. The first question we asked concerned a prominent but respectable ‘no-deal Brexit’ Conservative politician who has faced opposition on campus: ‘Should Conservative MP Jacob Rees-Mogg, who supports a ‘clean Brexit’ be prevented by the university from speaking on your campus?’ While 52 percent opposed no-platforming Rees-Mogg, 26 percent were in favour, with 22 percent saying they didn’t know.
We then asked students whether they agreed that: ‘Universities should regulate which kinds of attire people can wear to parties in order to protect people from cultural appropriation (such as wearing Mexican sombreros or Japanese kimonos), and set appropriate punishments for those who break the rules.’ 57 percent of students were opposed, with 21 percent in support and 22 percent not sure.
In order to probe several high-profile, concrete UK cases of threats to academic freedom, we asked students about the Jordan Peterson and Germaine Greer incidents. By a 41-31 margin, students backed Cambridge University’s decision to rescind Jordan Peterson’s fellowship, with 24 percent undecided and 4 percent saying they didn’t know. When asked whether Cardiff University should have overruled protesters to permit Greer to speak, just 37 percent agreed with the free speech position, while 42 percent backed Cardiff’s no-platform decision. 21 percent didn’t know or were unsure.
The aims of academic freedom and emotional comfort are often in tension, and we sought to establish which of these students prioritized. Asked which policy their university should back, ‘Prioritise free speech, even if this makes people upset’ or ‘Prioritise emotional safety, even if this limits free speech’, students, by a 52-31 margin, opted for free speech, with 16 percent undecided.
Given the significant undecided component in our and previous studies, we took the step of testing, through a survey experiment, whether passages emphasizing either Britain’s tradition of defending the speech rights of unpopular speakers or the need to protect minorities, women and trans people from emotional harm.
A third of students to read a pro-free speech paragraph, a third the safety paragraph and one third (the control group) read no passage. Results are shown in figure 2. Those who read no paragraph came out 49-34 for speech over safety, with 17 percent undecided. But when exposed to a short paragraph making the case for free speech, the share in favour of free speech over safety increased 14 points, to 63-30. By way of comparison, this means the sample became as pro-free speech as the small share of students in the sample who support Britain’s exit from the European Union (64-30), and more so than the men in the sample (56-29).
However, when respondents read a paragraph emphasizing the importance of protecting vulnerable minority groups, the proportion backing safety jumped 18 points. Much of these increases came at the expense of the ‘don’t know’ group, but some came from former free speech proponents. In other words, there is a considerable pool of malleable opinion, and even one paragraph of argument has a large effect for how students reason in any given moment. This indicates a sustained campaign, especially one backed by a university, could have a major impact on support for free speech.
All told, these results suggest a malleable middle ground with relatively impervious redoubts of pro-free speech sentiment (a core of 41 percent) and pro-safety feeling (30 percent). Students at the more elite Russell Group institutions tended to prioritize safety more than students in the post-1992 newer universities.
The Role of Gender
Gender, as expected, was important. Women leaned considerably more toward the protective/safety position. Men were 20 points more likely than women to oppose the banning of Jacob Rees-Mogg from campus, 23 points more opposed to dress restrictions at costume parties, 17 points more supportive of Jordan Peterson and 9 points toward Germaine Greer and over 20 points more likely to prioritise free speech over emotional safety.
Women were considerably more swayed by appeals to vulnerable minorities’ need for emotional safety than men. Figure 3 shows that women who read the safety paragraph became 20 points more supportive of safety over free speech, rising from 38 to 58 percent pro-safety. The free speech paragraph, by contrast, shifted women just 11 points toward free speech over safety, from 44-38 to 55-37. A core of 37 percent of the female sample remained wedded to the safety position even after reading the free speech paragraph. By contrast, men shifted just one point toward safety after reading the vulnerability passage, from 56-29 in favour of free speech to 55-34. The gender gap in the effect of reading this paragraph is a yawning 24 points.
On the other hand, while the free speech statement affected both men and women, it shifted men 18 points and women only 11 points. Thus male students who read the free speech paragraph came out 74-21 in favour of free speech whereas women who read the same paragraph recorded a 55-37 pro-free speech predilection. The fact women comprise 57 percent of UK undergraduates and men 43 percent is one factor – though not the main one – contributing toward more of an emphasis on emotional safety.
The gender gap is not principally about ideology – for instance there is no gap in the UK on anti-immigration sentiment or the Brexit vote. And women are more likely to endorse surveillance of Islamist students on campus in the name of safety than men, as Hillman (2016) shows. Yet the safety message may chime better with women because they are socialised into more caring roles in society, or because they are more likely to adhere to social norms – an effect we also see in the gender gap in populist right voting (Spierings & Zaslove 2017). Nigel Farage’s UK Independence Party, for instance, did almost twice as well among men than women in 2015 (Green & Prosser 2018). In the US, women were 11 points less likely to vote for Trump in the 2018 midterm elections (Tyson 2018).
Are Conservative Students Comfortable in Class?
High-profile cases of no-platforming are rare in Britain, but this may just be the tip of the iceberg. A more routinized ‘chill’ effect on conservative opinion could be operating. Indeed, our results show that, whether part of the 80 percent of students who support the Remain side, or the 13 percent who back Leave, all concur that it is much more comfortable to express Remain sentiment in class — only 45 percent said that a student in their class would be comfortable expressing support for Leave. As figure 4 reveals, under 4 in 10 Leave-voting students said a Leaver would feel comfortable expressing their opinion in class.
Are Academics Brainwashing Students?
One area where there is limited research – whether American or British – is on the source of student beliefs about free speech and emotional safety. Thus we wanted to find out where students got their views on the Peterson and Greer cases. The results are in figure 5. Lo and behold, 68 percent said social media. This was by far the most important influence on student opinion on these issues, with parents a distant second on 14 percent. New partisan online news sites like Vox, Buzzfeed, Breitbart, the Mail or the Guardian scored 8 percent. University professors and schoolteachers barely managed 1 percent each. This indicates that the content of what students are learning is not the key to understanding this phenomenon.
A further data point in favour of this interpretation is that older students (those over 20) were 19 points more likely than 18 and 19 year olds to back free speech over emotional safety. Nevertheless, it must be emphasized that we did not find this effect on our other questions, so more research is needed to substantiate this finding.
From Diagnosis to Policy
Our report makes a number of policy suggestions to address the interrelated challenges to academic freedom and viewpoint diversity.
First, as in the Canadian province of Ontario, we recommend that the British government, via the Higher Education regulator, the Office for Students (OfS), mandate that each university adopt an academic freedom commitment, such as the Chicago Principles, which clearly states that ‘debate or deliberation may not be suppressed because the ideas put forth are thought by some or even most members of the University community to be offensive, unwise, immoral, or wrong-headed’.
Second, that the OfS should ensure that universities each appoint an officer empowered to investigate allegations of academic freedom violations and/or instances of political discrimination. These individuals would also lead on collecting survey data on academic freedom within their institutions. Faculty and student whistleblowers would receive full protection.
In contrast to the plaintiff-slanted, hyper-vigilant ‘bias response team’ approach, in which discrimination is subjectively and expansively defined, all complaints would be required to meet a high and tightly-specified discrimination threshold based on the ‘rational person’ rather than ‘most sensitive person’ standard before investigation would proceed. The accused would be presumed innocent until proven guilty, and penalties would be calibrated to the seriousness of the offense, with a warning for a first infraction.
Third, the government should establish statutory duties against political discrimination in higher education and extend freedom of speech protections to encompass student unions, which are a major source of challenges to free speech in British universities, especially at elite institutions.
Finally, we recommend that a civil society organization be established along the lines of those established to advance equality and diversity. This would rank universities, as is the case with equality bodies, and their free speech ranking could form part of broader university rankings, which influence student and parent university choices.
Read the full report: Tom Simpson & Eric Kaufmann (2019). “Academic Freedom in the UK.” Policy Exchange.
Your generosity supports our non-partisan efforts to advance the principles of open inquiry, viewpoint diversity, and constructive disagreement to improve higher education and academic research.