Academic Freedom in Crisis: Punishment, Political Discrimination, and Self-Censorship

Home , Academic bias, Cancel culture, Dog whistle (politics), Eric Kaufmann, Free Speech Union, Peer pressure

Eric Kaufmann*

Center for the Study of Partisanship and Ideology

Report No. 2

March 1, 2021

* Professor of Politics at Birkbeck College, University of London; Research Fellow, Center for the Study of Partisanship and Ideology.

Table of Contents Table of Contents ...... 2 Table of Figures...... 4 The Surveys ...... 7 Part I: Hard Authoritarianism ...... 9 No-Platforming and Firing ...... 9 The Iceberg Model of Discrimination ...... 12 Disciplinary Action and Bullying ...... 15 Hard Authoritarianism Testimonials ...... 18 Left-Wing Authoritarianism ...... 20 Testing Support for “Cancel Culture” Among Academics ...... 20 Who Backs Dismissal? ...... 24 Far Left Activists more likely to Back Dismissal ...... 25 Ideology and Intolerance ...... 26 Are Young Scholars Less Tolerant? ...... 26 Conformity or True Belief? ...... 34 Cross-Pressured Between Value Commitments ...... 35 A Lopsided Trade-Off: Free Speech and Political Correctness in Academia ...... 37 Decolonizing the Curriculum? ...... 43 Diversity Statement Experiment ...... 47 Academic Freedom Statement Experiment ...... 50 Age, Gender and Support for Academic Freedom ...... 53 Attitudes toward Nonconformity ...... 56 Academic Freedom or Social Justice? ...... 58 Not Against Academic Freedom, But Not For It, Either ...... 60 Part II: Soft Authoritarianism ...... 62 Prior Work ...... 62 From Hard to Soft Authoritarianism ...... 64 Political Discrimination ...... 65 Section IIa: The Ideological Evolution of Academia ...... 66 The Political Views of Academics ...... 66 Ideological Age Profile of Academics...... 79 Age and Selection Effects Inside the Academy...... 87 Left-Modernist Hegemony? ...... 89 Section IIb: Chilling Effects and Self-Censorship ...... 97 A Hostile Climate for Conservatives in Academia? ...... 97 Chilling Effects ...... 107 Expressing Political Views to Colleagues ...... 115 British Testimonials...... 124 North American Testimonials ...... 126 Career-Related Concerns ...... 128 Transgender Issues ...... 132 Leftist Complaints Against Conservatives ...... 132 Section IIc: Political Discrimination ...... 136 Political Discrimination: Quantitative Evidence ...... 136 Do Left and Right Discriminate Equally? ...... 142 Political Discrimination in Britain ...... 143

Is there Bias in Refereeing Grant and Promotion Applications and Journal Articles? ...... 150 Discrimination by PhD Students ...... 153 Who Discriminates? ...... 155 Soft Authoritarianism: Chilling Effects of Discrimination ...... 156 Summary of Discrimination Effects ...... 157 Social Aversion to Conservatives and Gender-Critical Scholars ...... 158 Discrimination: Summary ...... 164 Toward Soft Authoritarianism ...... 164 Political Discrimination and Illiberalism in Academia ...... 165 Part III Off-Campus ...... 169 Is Academia Different? ...... 169 Comparison with Non-Academics ...... 169 Non-Academic Comparator Study ...... 170 Non-Academic Workplaces Are Less Hostile to Expressive Freedom ...... 171 Willingness to Discriminate ...... 175 Authoritarianism ...... 178 Empirical Summary ...... 180 Policy Discussion...... 184 What to Do?: Policy Directions to Safeguard Academic Freedom ...... 184 Libertarian Approaches ...... 184 Interventionist Approaches ...... 185 Appendix ...... 187 References ...... 190

Table of Figures Table 1. Various surveys used in report ...... 9 Figure 1. Disinvitation Incidents, US Universities, 1998-2019 ...... 10 Figure 2. UK University-Related Free Speech Incidents in the Media, By Year ...... 11 Figure 3. “Iceberg” of Threats to Academic Freedom (Victim’s View)...... 13 Figure 4. “Iceberg” of Threats to Academic Freedom (Perpetrator’s View) ...... 14 Figure 5. Share Subject to Disciplinary Action, or Threat Thereof ...... 16 Figure 6. Experience of Disciplinary Action and Bullying, NAS, EU and UCU Respondents...... 17 Figure 7. Support or Oppose Campaign to Oust a Dissenting Academic, US Social Science/Humanities ...... 22 Figure 8. Support or Oppose Campaign to Oust a Dissenting Academic, UK Social Science/Humanities ...... 23 Figure 9. Support Campaign to Oust Dissenting Academic (SSH staff/PhDs only) ...... 24 Figure 10. Far-Left and Activist Share of Current SSH Staff ...... 25 Figure 11. Support Campaign to Oust Dissenting Academic (UK Academics) ...... 26 Figure 12. Predictors of Supporting a Dismissal Campaign (UK)...... 27 Figure 13. Predictors of Supporting Any Dismissal Campaign (US/Canada) ...... 28 Figure 14. Illiberalism on any Question, by Age and Far Left (UK) ...... 29 Figure 15. Support at least one Dismissal Campaign (Age 23-33) / Support Dismissal in Diversity-Performance Case (23-33) ...... 30 Figure 16. Predictors of Support for Dismissal Campaign ...... 31 Figure 17. Support for Any Dismissal Campaign by Position and Age Group (US, Canada, Britain) ...... 32 Figure 18. Would you support or oppose your school ALLOWING a speaker on campus who promotes the following idea: ...... 33 Figure 19. Not Silent, Not a Majority (Open and Secret Opposition to Dismissal) ...... 35 Figure 20. Probability of Supporting Counter-Letter, Center and Right Respondents Only (UK) ...... 36 Figure 21. PC: The Way People Talk Must Change (vs. People Too Offended), USA ...... 38 Figure 22. Thinking about political correctness, are you generally in favor of it (it protects against discrimination), or against it (it stifles freedom of speech)? (Britain) ...... 39 Figure 23. Predictors of support for political correctness among academics (Britain) ...... 40 Figure 24. Views of Political Correctness, UK PhD Holders, by Age Bracket ...... 41 Figure 25. Support for Political Correctness, by Age and Occupation, PhD holders (UK)...... 42 Figure 26. Support Diversity Quotas for Reading Lists (%) ...... 44 Figure 27. Net Willingness to Voice Views Publicly among Opponents and Supporters of Reading List Quotas ...... 45 Figure 28. Open and Secret Opposition to Decolonization...... 46 Figure 29. Favor Decolonizing the Curriculum by Mandatory Quotas ...... 47 Figure 30. Favor Decolonizing Reading List (UK) ...... 48 Figure 31. Political Correctness, Read Statement and Support for Reading List Quotas (UK) ...... 49 Figure 32. Support for Diversity Quotas on Reading List ...... 50 Figure 33. Effect of Treatments on Support for Decolonization (Britain) ...... 51 Figure 34. Support for Free Speech or Emotional Safety, British Undergraduate Students, by Type of Paragraph Read...... 53 Figure 35. Support Decolonization Initiative ...... 54 Figure 36. Support Quotas over Foundational Texts (Young/Older) ...... 55 Figure 37. Support Quotas over Foundational Texts (Male/Female) ...... 56 Table 2. Preferred Punishment for Quota Non-Compliance among Pro-Quota North American Academics ...... 57 Figure 38. Prioritize Academic Freedom or Social Justice ...... 59 Figure 39. Academic Freedom or Social Justice (North American Academics) ...... 60 Figure 40. Faculty Ideology in American Universities, 1989-2014...... 67 Figure 41. Number of Democratic Faculty Members For Every Republican in 25 Academic Fields ...... 68 Figure 42. The Political Composition of UK Academia, 1964-2015 ...... 69 Figure 43. The Political Composition of UK Academia, Selected Surveys, 2015-2019 ...... 70 Figure 44. SSH Academic Voting in 2016-2019 Elections, US, UK, Canada ...... 71 Figure 45. Ideological Self-Placement, EU, UK, North America ...... 72 Figure 46. Academics and Public Attitudes, Britain ...... 73 Figure 47. Academics and Public Attitudes (USA) ...... 74 Figure 48. Academics and Public Attitudes (Canada) ...... 74 Figure 49. Ideological Self-Placement, by Profession in America, 2016 ...... 75 Figure 50. Brexit Vote, by Profession, Education Industry (Britain) ...... 76 Figure 51. Brexit Vote by Sector, Degree-Holders Only (UK) ...... 77

Figure 52. Political Donations by Profession in America, 2016 ...... 78 Figure 53. Ideology by Age, American SSH Academics...... 79 Figure 54. Ideology by Age, UK SSH Academics (YouGov) ...... 80 Figure 55. Share of Far Left Academics, by Age, among SSH Faculty ...... 81 Figure 56. Share of Right-Wing Academics, by Age, among SSH Faculty ...... 82 Figure 57. Share in Favor of Reducing Immigration, by Age (SSH Only) ...... 82 Figure 58. Share in Favor of Increasing Immigration, by Age (SSH Only) ...... 83 Figure 59. Predictors of Left Ideology ...... 84 Figure 60. Partisanship among Americans with Advanced Degrees, 1964-2016...... 85 Figure 61. Ideology of Americans with Advanced Degrees, 1972-2016 ...... 86 Figure 62. Republican Vote, Undergraduate vs. Advanced Degree-Holders ...... 87 Figure 63. Left Identification, by Active/Retired and Age (UK) ...... 88 Figure 64. Left:Right Ratio by Student Status (Prolific) ...... 89 Figure 65. Left:Right Ratio by Education Status (UK Yougov)...... 90 Figure 66. My Political Views Wouldn’t Fit Academia, Graduate Students, by Ideology (STEM and SSH) ...... 92 Figure 67. Correlation between Left-Right Ideology and Academic Career Considerations ...... 93 Figure 68. Predictors of Interest in Academic Career ...... 94 Figure 69. Interest in Academic Career, by Ideology and Political Compatibility, SSH Graduate Students ...... 95 Figure 70. Interest in Academic Career, by Ideology and Political Compatibility (Masters only) ...... 96 Figure 71. Share of Graduate Students Interested in Academic Career...... 97 Figure 72. Ideological Self-Placement of Academics, by Field (Britain) ...... 99 Figure 73. Ideological Self Placement by Field (USA) ...... 100 Figure 74. Ideology, Gender and Field (Britain)...... 101 Figure 75. Ideology, Gender and Field (USA) ...... 102 Figure 76. % Share by Actual Ideology and Perceived Mean Ideology in Own Department (Britain) – STEM Subjects ...... 103 Figure 77. % Share by Actual Ideology and Perceived Mean Ideology in Own Department (Britain) – SSH Subjects ...... 104 Figure 78. Perceived Department Ideology (USA) ...... 105 Figure 79. % Share by Actual Respondents Ideology and Perceived Mean Ideology in Own Department (USA) ...... 106 Figure 80. Perceived Department Ideology (USA) ...... 107 Figure 81. A Hostile Climate for Your Beliefs in Dept? (Britain)...... 108 Figure 82. A Hostile Climate for Your Political Beliefs? (USA) ...... 108 Figure 83. A Hostile Climate for your Beliefs? (US-UK-Canada) ...... 109 Figure 84. Supportive or Hostile Climate in Dept (Britain)...... 110 Figure 85. Probability of Perceiving Hostile Dept Climate (Britain) ...... 111 Figure 86. Identify with Departmental Culture (British academics)...... 112 Figure 87. Identification with Culture of Department (US Academics) ...... 113 Figure 88. Identification with Culture of Departments (British Academics)...... 114 Figure 89. Identification with Culture of the Department (US and Canadian Academics) ...... 115 Figure 90. Would someone be comfortable expressing their Brexit view? (Britain) ...... 116 Figure 91. Would Trump/Biden supporter be comfortable expressing their view? (by own 2016 vote) ...... 117 Figure 92. Would Trump or Biden supporter be comfortable expressing their view to a colleague? (Leave/Trump voters only) ...... 118 Figure 93. Would a Leave Supporter be Comfortable Expressing View? (Current Staff Only), UK ...... 119 Figure 94. Have you refrained from airing views in teaching and research (Britain) ...... 120 Figure 95. Probability of Censoring One’s Views, Current Staff Only, UK ...... 121 Figure 96. Have you refrained from airing views in teaching and research (US SSH) ...... 122 Figure 97. Probability of Censoring Views (US and Canadian Academics) ...... 123 Figure 98. Self-censorship of Views in Teaching and Research, SSH only, by Sample ...... 124 Table 3. List Experiment Design ...... 137 Figure 99. Discrimination Against Trump/Leave Supporter in a Hiring ...... 139 Figure 100. Share Who Would Discriminate against a Leaver’s Job Application (Britain)...... 140 Figure 101. Share Who Would Discriminate against a Trump Supporter’s Job Application (US) ...... 141 Figure 102. Job Discrimination, by Own Political Orientation ...... 144 Figure 103. Discrimination Against Left Candidate, by Own Ideology/Discrimination Against Right Candidate, by Own Ideology ...... 145 Figure 104. Job Discrimination, by Own Political Orientation (US) ...... 147 Figure 105. Job Discrimination, by Own Political Orientation (Canada) ...... 148

Figure 106. Soft and Hard Discrimination in Hiring (North America vs. Britain) ...... 149 Figure 107. Would Rate Paper, Grant or Promotion Lower, by Rater’s Ideology (US) ...... 151 Figure 108. Would Rate Paper, Grant or Promotion Lower, by Rater’s Ideology (Britain) ...... 153 Figure 109. Openly Admitted Discrimination against Right-Leaning Papers, Grants and Promotions ...... 154 Figure 110. Predictors of Discrimination against Conservatives ...... 156 Figure 111. Comfortable Sitting Next to Someone of Opposing View at Lunch? (USA)...... 159 Figure 112. Comfortable Sitting Next to Someone of Opposing View at Lunch? (Britain) ...... 160 Figure 113. Would feel comfortable having lunch with Trump supporter, by Ideology (US and Canada) ...... 161 Figure 114. Would feel comfortable having lunch with a Leaver, by Ideology (UK) ...... 162 Figure 115. Would Discriminate Against Conservative or Leaver, by Own Ideology (Britain) ...... 163 Figure 116. Would Discriminate against Conservative or Trump Supporter, by Ideology and Comfort ...... 164 Figure 117. Predictors of Supporting a Dismissal Campaign (US and Canada) ...... 166 Figure 118. Predictors of Supporting a Dismissal Campaign (Britain) ...... 167 Figure 119. Illiberalism, by Far Left and Field (US and Canada) ...... 168 Figure 120. Illiberalism by Far Left and Willingness to Discriminate (UK) ...... 169 Figure 121. Would Leaver/Remainer Be Comfortable Expressing View? (by Sector, Britain) ...... 172 Figure 122. Probability of Reporting Hostile Environment, by Brexit and Workplace ...... 173 Figure 123. Revealed Share That Would Discriminate against a Leaver for a Job (Non-Academics, Britain) ..175 Figure 124. Revealed Willingness to Discriminate against a Leaver (Britain)...... 176 Figure 125. Share Who Know the Brexit Vote of over Half their Colleagues (%), by Sector ...... 177 Figure 126. Predictors of Expelling a Conservative Dissenter (Non-Academic Sample) ...... 179 Figure 127. Support for Affirmative Action, by Ideology (Non-Academics, UK)...... 180 Table 4. Previous and Current Studies of Academic Bias and Discrimination ...... 181 Table 5. Summary of Survey Results ...... 182

This report represents the most in a modular way by clicking on a heading comprehensive survey-based investigation in the table of contents. The reader can to date of academic and graduate student return to the table of contents at any time opinion on political discrimination, the by clicking on any heading in the text. The punishment of academics for speech, and same is true for the table of contents of experiences of hostility and self- tables and figures. censorship for political beliefs. It examines evidence from the perspective of both The Surveys victims, largely concentrated among the minority of 5-10% conservative or gender- This study relies largely on survey critical academics; and perpetrators, data of academics and PhD students at involving a substantial minority or universities in the United States, Britain, sometimes even a majority of scholars. and Canada. It also summarizes prior work The report is divided into four parts. on staff and students in the United States Part I concentrates on what I term and Europe and contextualizes findings hard authoritarianism, notably the with surveys of the general population and experience of being disciplined or the wider advanced degree-holding public. threatened for speech – or backing the The core of the study concentrates on firing or disciplining of controversial academic attitudes, where there has been academics; as well as support for policies less research than on student attitudes. such as mandatory reading list quotas that In order to mitigate the effects of abridge academic freedom. Part II social desirability bias – people giving examines soft authoritarianism, socially acceptable responses – I use a list encompassing views on political experiment.1 This is an innovation in work discrimination, as well as experiences of on political discrimination, and gives a hostility and self-censorship in more accurate indication of how prevalent scholarship, teaching, and other aspects of the willingness to discriminate on political academic life. This part replicates the grounds is. In the analysis, I focus more on findings of many existing studies but also the social sciences and humanities because uses new methods to go beyond them. Part political considerations are a larger aspect II is subdivided into three sections, the of these fields’ conceptual foundation, and first on the political leanings of academics, are thus assumed to exert greater influence the second on the extent of chilling effects over the culture and practices of these and self-censorship, and the third on disciplines. support for political discrimination. The UK data comprise the highest Part III goes off campus to ask quality sector-wide academic survey to whether the patterns of hard and soft date, based on a 61-76% response rate authoritarianism we find among academic from the approximately 1100 current and staff and graduate students are also present retired academics that happen to be on among college graduates working in YouGov’s 500,000 strong national British organizations off campus. We’ll see that panel. Since these are respondents who are there are differences, but also many in the YouGov system answering other similarities, between the two professional surveys, there is no danger that they are realms. Part III also discusses potential self-selecting into the survey due to their policy reforms to address the problems interest in answering questions on highlighted, comparing and contrasting academic issues. US and Canadian surveys interventionist approaches against those represent a mix of online surveys using that are more hands-off. convenience samples with low response The report is meant to be read as an rates from a large target pool, and integrated whole but can also be accessed platform-based surveys with very high

response rates from smaller target In addition to replicating previous populations. I also fielded a UK online studies of soft authoritarianism (political mailout survey to academics in the top 100 discrimination and chilling effects), I (out of 143) universities to compare with break new ground by examining support the relatively randomly-selected YouGov for hard authoritarianism – in this case UK sample and the online North American dismissal campaigns for controversial surveys. academics, and penalties for failing to Though response rates to the North implement mandatory curriculum diversity American and British mailout surveys are quotas. I also use the surveys to create only around 2-4%, the data they provide quantitative models that, in a departure tell a very similar story to the non- from existing studies, examine the selective YouGov and Prolific surveys relationships between the ideological skew (with 61-87% response rates), and to of staff, political discrimination, self- previous questionnaires fielded by other censorship, and support for cancelling scholars. The result is a triangulated body controversial scholars or employees. of replicated knowledge on political Table 1 shows the methods and discrimination and chilling effects that has population sampled in each survey not been contradicted and thus should no discussed in this report. More details of the longer be seriously disputed. surveys may be found in the Appendix.

Method Population Purpose Dates Sampled US and Canada E-mail Academics from Academic attitudes August 2020 top 100 US and experiences universities and top 40 Canadian universities, mostly social sciences and humanities (SSH), some STEM UK YouGov YouGov platform Academics from Academic attitudes March 2020 YouGov national and experiences British panel UK Mailout E-mail British SSH Academic attitudes September 2020 academics and experiences US and Canada Prolific platform PhD students Graduate student May 1-June 4, PhD Survey attitudes and 2020 experiences, future trends UK PhD Survey Prolific platform PhD students Graduate student May 1-June 4, attitudes and 2020 experiences, future trends US NAS E-Mail Members of the Perspective from a May 6-June 12, National political minority 2020 Association of Scholars, a majority- conservative academic organization Graduate Student Prolific platform Master’s and PhD Graduate student December 23, Academic Careers students in the US, attitudes and 2020-January 5, Survey UK, and Canada experiences, future 2021 trends, understanding how discrimination affects career goals UK Non-Academic Prolific platform Members of UK To compare with August 21-22, Survey public with college academics. 2020 degrees Table 1. Various surveys used in report.

Part I: Hard Authoritarianism person that a dissident scholar has been forced to leave. As we shall see, there is a small minority of academics that supports No-Platforming and Firing ousting those with controversial views. Administrators are often willing to use the This forms part of a growing university’s disciplinary apparatus to climate of political intolerance. In the enforce this sentiment, and it is not clear United States, the Foundation for that academics are sufficiently motivated Individual Rights in Education (FIRE) to resist this. While it is rare for an maintains a database of disinvitation academic to be fired, especially on incidents. These are presented below in ideological grounds, a number of worrying Figure 1. Incidents rose steadily in the cases have recently come to light in which 2000s, driven mainly by left activists, and life has been made so uncomfortable for a

spiked in 2015 when Nicholas Christakis Coddling of the American Mind.” Since was mobbed by a group of Yale students then, incidents have remained at an and Jonathan Haidt and Greg Lukianoff elevated level compared to the pre-2010 published their seminal piece on the period. phenomenon in the Atlantic, “The

Disinvitation Incidents, US Universities, 1998-2019

45 40 35 30 25 35 33 20 Unclear 15 by the Right

10 20 22 by the Left 5 17 20 2 3 5 6 9 11 13 13 13 0 0 4 12 14

4 12

1998

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017 2018 2019 Figure 1. Source: FIRE disinvitations database, accessed Nov. 12, 2020.

The National Association of summer of 2020, John McWhorter of Scholars (NAS) maintains a database of Columbia University received 150 (mainly) American academics who have messages from academics sharing their experienced campaigns calling for their anxieties about their work climate, and he dismissal. The database records 4 incidents continues to receive at least one per week.3 apiece in 2015 and 2016, 9 in 2017, 13 in In Britain we see a similar, if 2018, 12 in 2019, and a striking 65 in slightly delayed, pattern for deplatforming 2020. The database relies partly on reports and dismissals. The trend in combined free from users, thus it is unclear whether the speech incidents, as catalogued in two surge in 2020 reflects a disconcerting new separate lists, with a minority of rise in intolerance or the prevalence of overlapping cases, is presented in Figure 2 real-time over historic reporting since the and shows a noticeable rise from 2013, site was established in June 2020. and especially from 2018. Note that 2020 Regardless, 65 cancellations in one year data are incomplete, but already show a suggests the problem cannot be dismissed higher level of incidents compared to as purely anecdotal.2 Meanwhile, under 2017. the radar, in just one 3-week period in the

UK University-Related Free Speech Incidents in the Media, by Year 35 30 25 All Issues List 20 15 Gender-Critical List 10 5 0

Figure 2. Source: Wanstall, Mark, “The Banned List,” Academics for Academic Freedom; Biggs, Michael, “Academics and Others at British Universities Targeted for Questioning Transgender Orthodoxy,” http://users.ox.ac.uk/~sfos0060/GCtargets.shtml (limited to cases reported in the media).

A report by Civitas in December the data compiled by Wanstall or Biggs. A 2020 discovered that over half (53%) of all call for evidence on Twitter in mid-2019 137 British universities experienced by Professor Kathleen Stock of the demands for censoring speech around University of Sussex turned up 28 alleged “transphobic” episodes during testimonies of restrictions, threats, and 2017-2020. Many UK universities were harassment from British gender-critical also targeted by activists seeking to academics in just 7 days.5 suppress other forms of free speech. Over While successful dismissals are half (55%) of universities had at least one rare in academia, there have been a open-letter campaign, 37% reported social number of notable cases, including those media drives, and nearly a quarter bore the of Bo Winegard of Marietta State College brunt of a campus pressure group drive and Noah Carl and Jordan Peterson at seeking to curtail academic freedom or Cambridge University. In Britain, cases free speech. These forms of political include Chris Hill, formerly at the intolerance thereby encompassed a University of Central Lancashire in July majority of British universities.4 2018, and Andrew Dunn, formerly a Many other cases occur beneath Social Policy lecturer at the University of the radar: in a one-month period from Lincoln.6 Peterson’s case involved guilt by June 13 to July 14, 2020, the UK’s Free association with an Islamophobic fan Speech Union received 10 calls for whom Peterson did not realize to be anti- assistance from academics or contractual Muslim. The Hill case involves remarks staff asking for help to defend against some found offensive, a view upheld by online campaigns or university the university. Hill claims that accusers disciplinary charges brought against them were misrepresenting his views out of for public statements. None are listed in hostility to his conservatism. Likewise, the

Dunn case involved little more than anti- title IX trials and summarily had their lives conservative discrimination leading to the ruined by their respective schools. Kansas lecturer being forced out. State University being the worst offender.” There are also an unspecified –Politically right-wing sociologist, US number of disciplinary actions that have not been reported by the media, many of The Iceberg Model of Discrimination which are subject to privacy rules, but some of which are alluded to in comments We can think of the threats to I collected in my NAS survey, such as: academic freedom using the metaphor of an iceberg, with items that make the news “One professor at East Georgia State – such as deplatformings and dismissals – College was fired for ‘sexual harassment’ as the visible symptoms of a much deeper for disagreeing with the school’s sexual problem. Figure 3 and Figure 4 present harassment policy. I have heard many some of the main findings of our survey, stories from credible sources of from the perspectives of the victims and [sociology] professors, students and the perpetrators of discrimination. graduate students being subjected to secret

"Iceberg" of Threats to Academic Freedom (Victim's View)

No-platforming or Dismissal Incidents 0.03%

Threatened by Disciplinary Action for Speech, UK/EU (Mainly 6% left sample)

Threatened by Disciplinary Action for Speech (US NAS 23% academics)

Share of US NAS Academics Reporting Bullying for Beliefs 36%

Share of US NAS Academics Reporting Psychological Pressure for 50% Beliefs

Share of UK Right or Leave SSH Academics Reporting Hostile 50% Climate for Beliefs

Share of UK Academics Saying Leavers Uncomfortable 57% Expressing Beliefs, or Unsure

Share of UK SSH Academics Saying Leavers Uncomfortable 64% Expressing Beliefs to Colleagues

Share of Right-Leaning US Academics Reporting Hostile Climate 70% for Beliefs

Share of Right-Leaning US Soc Science/Humanities (SSH) 70% Academics Self-Censoring

Share of UK Leave SSH Academics Saying Leavers Uncomfortable 82% Expressing Beliefs, or Unsure

Share of US Academics Saying Trump Supporter Would not 86% Express Views, or Unsure

Share of US pro-Trump Academics Saying Trump Supporter 91% Would not Express Views, or Unsure

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100%

Figure 3.

As can be seen, deplatformings or are experienced by a majority of political dismissals affect only a tiny minority of minorities on the faculty. By and large, academics. Yet when one polls academics, most right-leaning academics feel a hostile particularly centrists or conservatives, on environment or self-censor in some form their experiences, we find that there are or another. some forms of political discrimination that

"Iceberg" of Threats to Academic Freedom (Perpetrator's View)

No-platforming or Dismissal Incidents 0.03%

Share of Academics Willing to Support a Given Dismissal 9% Campaign

Share of SSH Academics Willing to Support at least One of 25% Four Dismissal Campaigns

Share of UK Academics Who Would Discriminate Against 32% Leave Supporter for a Job

Share of American Academics Who Would Discriminate 40% against a Trump Supporter for a Job

Share of American PhD Students Who Support Dismissing an 43% Academic for Finding Against Diversity

Share of UK Academics Who Would Discriminate against a 44% Right-Leaning Grant

Share of American Academics Uncomfortable/Unsure about 57% Sitting Next to Trump Supporter

Share of UK Academics Uncomfortable/Unsure about Sitting 63% Next to Gender-Critical Academic

Share of American PhD Students Favoring Sanders Supporter 65% over Trump Supporter for a Job, Rather than Being Neutral

Share of Leftist American Academics Who Would 67% Discriminate against the Right on at least One Dimension

Share of American Academics Uncomfortable/Unsure about 71% Sitting Next to Gender-Critical Academic

Share of American PhD Students Who Would Discriminate 82% against the Right on at least One Dimension

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 4.

Surveys of the rest of academia desirability bias) a substantial minority or show that conservative scholars are not an overwhelming majority engages in imagining these things. Only a small discrimination. Among PhD students, the minority of academics is willing to results are even more extreme than they endorse dismissal campaigns. Nonetheless, are among other academics, with over depending on the question asked and the 80% of American PhDs willing to way that it is asked (i.e., whether a “list” discriminate against right-leaning scholars method is used to get around social on at least one dimension.

Disciplinary Action and Bullying mailout survey using the same UCU question wordings. These suggest that right-leaning academics are more likely to In July 2020, a high-profile mention that they have been the target of statement from 150 leading liberal writers disciplinary action for their public and academics, including J.K. Rowling, statements, research, or teaching. Noam Chomsky, and Salman Rushdie, My US online survey results show that warned of the rise of a “cancel culture” in right-leaning academics experienced 16- academia, the arts, and the media. “It is 100% more academic discipline than now all too common to hear calls for swift colleagues on the left on 5 of 6 measures. I and severe retribution,” they write, “in did not put these questions to my UK response to perceived transgressions of academic samples, but according to my speech and thought.” Noting the “hasty” Prolific PhD samples (with 72% of PhDs decisions of administrators to fire eligible taking the survey), right-leaning transgressors, they continue that “We are US and Canadian PhD students report 3 to already paying the price in greater risk 6 times the level of disciplinary action as aversion among writers, artists, and PhD students with centrist or leftist views. journalists who fear for their livelihoods if Thirty-eight percent of conservatives were they depart from the consensus, or even 7 disciplined, or threatened with discipline, lack sufficient zeal in agreement.” for their views in at least one of the However worrying the trend in no- following five spheres: teaching, research, platformings and dismissals, this is merely speech in private or public forums, or the most visible manifestation of a deeper speech elsewhere. and more pervasive problem in academia The Prolific North American PhD that rarely makes the headlines. The most survey and mailed out North American authoritarian aspect of academia that flies academic survey contain only small under the radar is disciplinary action and samples of right-leaning respondents (N= ideological bullying, much of which stem 40 PhDs and 43 academics). To rectify from fellow academics acting in the role of this, I ran a survey of the membership of departmental manager or peer. A study for the US-based National Association of the Universities and Colleges Union Scholars (NAS). Two-thirds report a right- (UCU), the main union for academics in wing identity, and 13-28% said they Britain, surveyed several thousand British experienced disciplinary action on any and European academics about their given dimension.9 Results show that experiences of being disciplined for 8 among this group, there is a substantially speech. The authors asked, “Have you elevated level of victimization, with 43% ever been subjected to informal or formal reporting being disciplined, or threatened disciplinary action, or the threat of with discipline, for their views in one of disciplinary action (up to, and including, the five spheres listed above. dismissal) because of academic views These results suggest that the expressed in [the following]?” Questions disciplinary threat level is higher for right- were asked about “academic views leaning academics and especially doctoral expressed” in teaching, research, within students. It helps explain why I find that institutions, in public fora, and elsewhere. right-leaning UK academics are 3-4 times Results appear in Figure 5. as likely as those of other political While I am unable to obtain the ideologies to report a “hostile climate” for UCU study’s crosstabulations by scholars’ their beliefs and over twice as likely to say ideology, I can compare their UK and EU that they self-censor. I estimate that right- data to American and Canadian survey leaning UK academics and PhD students results I have collected on my online are 50-100% more likely to experience

disciplinary threats than leftist and centrist academics.

Share Subject to Disciplinary Action, or Threat Thereof

40 38 35

30 28 25 23 25 23 23 23 20

% 20 18

15 12 13 9 10 8 6 6 7 6 6 6 7 7 5 5 5 4 4 5 4 4 5 5 3 3 3 3 2

0 EU UCU North North US left US right- NAS academics academics American American academics wing Academics (N=4141) (UK, right-wing center/left (N=760) academics (US, N=224) N=2299) PhDs (N=40) PhDs (N=43) (N=298) Teaching Research Non-Public/Meeting Public Elsewhere Research

Figure 5. Source: Karran and Mallinson. 2017. “Academic Freedom in the UK;” Membership survey of NAS academics, 18-25 May, 2020; Prolific Academic survey of US and Canadian PhD students, 11-22 July, 2020; Online survey of US academics 2020.

Across a broader range of More generally, around a quarter were disciplinary measures in Figure 6, we see disciplined or threatened for views the right-leaning membership of the NAS expressed in teaching and public or generally reporting 2-3 times as much university forums. authoritarianism as the overwhelmingly Finally, colleagues were the source left-leaning respondents in the Karran and of most of the problems reported by Mallinson EU/UCU survey. respondents in all datasets. Half of NAS Approximately 6-9% of NAS respondents reported psychological members experienced the most severe pressure from colleagues for their views forms of treatment for their views, and 36% said they experienced bullying including being moved to another for expressing them. The connection department or center (6%), demoted (8%), between colleagues, departmental being physically attacked (8%), or having authority, and higher-level administration research or facilities withdrawn (9%). A is important. Departmental colleagues quarter of NAS respondents reported being wield considerable power, both through falsely charged or threatened with charges informal peer pressure and via their formal for their views, and nearly a quarter were departmental managerial roles as well as given more, fewer, or different research, through faculty and college-level administrative, or teaching duties in committees that administer the university. response to things they said or wrote.

Experience of Disciplinary Action and Bullying, NAS, EU and UCU Respondents

4 Moved to Other Dept 4 6 4 Demotion 4 8 1 Physical Harm 1 8 7 Removal of Funding/Equipment 7 9 Threat/Punish for Views Expressed in 3 5 Publication 13 Threat/Punish for Views in Public Forum 4 6 (ie newspaper) 18 Given More/Fewer Teaching or 13 9 Research Duties 23 6 Threat/Punish for Views in Teaching 5 23 Threat/Punish for Views in Non-public 8 8 forum (ie Senate) 23 12 Given More/Fewer Admin Tasks 8 25 11 Charges Brought Against You 6 25 Threat/Punish for Views Expressed 7 5 Elsewhere 28 10 Other Sanction 5 29 23 Bullying by Colleagues 14 36 27 Psychological Pressure 16 50

0 10 20 30 40 50 60 UCU (2300) EU (4079) NAS (227)

Figure 6. Source: NAS survey May 6, 2020; Karran and Mallinson. 2017. “Academic Freedom in the UK.”

Hard Authoritarianism Testimonials and his lab scrutinized. He was placed under specific administrative guidelines for how he was to deal with this student.” Some reported being targeted by – Right, Communications, US the university’s disciplinary apparatus. The following are testimonials showing “Professor Dennis Gouws of the cases of hard authoritarianism. Humanities Department has been subjected to years of harassment, denial of “The university is right now considering sabbatical, placement on official disciplinary proceedings because of probation, threats of dismissal, and finally innocuous items I posted on my private administrative decree commanding him to blog. I think I’ll be alright, but stay silent or be fired (which he was forced conservative academics are somewhat to obey). The reason was his written and persecuted for their views.” – Right, spoken criticism of gender feminism.” – Geography, US Right, History, US

“Only recently has the attitude of the “A friend of mine at another university got administration turned to punitive measures a negative tenure vote despite a strong for people who are outspoken; thus most record for having unpopular opinions (and of us are being very careful about what we research findings). Fortunately the provost say and make sure we are covered by our overruled. One of his collaborators at a collective agreement to fend off the third school was subject to intense legal administration.” – Centrist, Classics, harassment (FOIA requests etc) though Canada this was *mostly* not internal to his university.” – Right, Sociology, US “Yes, every year I am called into my boss’s office because of some controversy “I had heard from a peer that one of their surrounding course curricula. At times, I close friends had been in the program the have been misunderstood. Other times, I year prior to ours, and halfway through his have to justify my work with students.” – department head sat him down and told Right, Education, US him he ‘wasn’t cut out to be a teacher,’ and that he was being cut from the “I have been dismissed as undergraduate program. My peer believed it was because programs chair in my department because he expressed strong conservative views in I am a gender critical feminist. Students class, perhaps not very diplomatically or currently have a petition underway respectfully.” – Right, Education, US supporting this dismissal because my faculty association is supporting me in a Several Leave-voting and grievance about it. I am also facing an conservative academics in my UK interestingly timed investigation into my YouGov survey report instances of research ethics, supposedly ordinary authoritarianism from administrators or procedure any time a faculty member is colleagues, impact on their freedom to featured in the news but the first time I’ve teach and research, and even threats to been investigated.” – Very Left, their jobs. Anthropology, Canada “Yes, indeed I have lost two senior jobs “A professor who declined to refer to a because I voted leave.” – Tory Leaver student by a false pronoun (a male wanted (Supporter of Brexit) to be called ‘she’) was visited by the diversity people and had all of his work

“I have been called in for a meeting with despite very positive feedback.” – Centrist University marketing, my Head of Remainer Department, and an HR officer after I published an article in a peer-reviewed “Yes – I avoid making political statements academic journal [redacted]. They asked – have also had head of department voice why I had not explicitly condemned strong disapproval for the sort of research conservatism as immoral within this I do and to use the ethics approval system article. I explained that I did not believe it to prevent certain research topics being was appropriate for me to use my position studied.” – Left Labour Leaver as a researcher to subjectively pass judgement on modern political ideology. I These comments go beyond the was told that there are some subjects I chilling effect of social opprobrium from shouldn’t remain neutral on, and that I colleagues and identify the potential for have a moral duty to condemn those on the abuse of power by those in the hierarchy political right. I was told that, if I insisted within the university. At all levels, the on remaining impartial within my institution can deploy sticks such as research, I was not to further research this assigning unpopular administrative tasks subject and warned I may face disciplinary or courses, limiting access to grant funding hearings if I did.” – Tory Leaver or promotion, or even forcing a staff member to leave. “Given the derogatory views regularly The latter is especially damaging, expressed by my colleagues about Leave not just for those who lose their jobs, but supporters, including the VC sending a for the signal it sends across the sector. University-wide email referring to us as Academic jobs are incredibly competitive ‘Little Englanders’, I have no doubt that if and specialized. A lecturer or professor on my views were known then it would a permanent contract who is dismissed or negatively affect the attitude of my forced to leave on a package is unlikely to colleagues towards me significantly. It find a permanent academic job in the same probably wouldn’t be career-ending, but it geographic area, especially if they are would reduce my influence, make it harder older, in a social science or humanities for me to deliver my teaching and field, and have acquired a reputation as leadership responsibilities, and quite likely conservative. The skill sets and networks force me into a position where I would of academics may also make it difficult to have to move institutions.” – Tory Leaver make a lateral move to another profession mid-career. All of this instills a powerful “I have to be careful about where I place sense of caution in many full-time my research because of two different areas academics, with a desire not to jeopardize of my research. Colleagues have attempted a fulfilling and reasonably (though not to stop me teaching.” – Tory Remainer highly) paid career. This leads them to self-censor, further constricting viewpoint “I had a professor who was very left wing diversity. This impairs the exchange of and who reduced the promotional chances ideas that is vital for both the academic of anyone that was centrist or slightly right enterprise, and for beginning to of centre.” – Tory Remainer intelligently negotiate between advocates on both sides of society’s major political “A previous line manager had a large divide. photo of Jeremy Corbyn on her desk. When I failed to approve (I said nothing) she had me removed from the programme

Left-Wing Authoritarianism empirically measure the extent to which scholars may disagree with the substance Hard authoritarianism on campus is of findings but agree that those often spearheaded by a left-wing undertaking unpopular research should not authoritarian minority. Recent work in be driven from their posts. The questions I social psychology identifies three distinct selected are as follows: clusters within the umbrella concept of left-wing authoritarianism. The first 1. If a staff member in your revolves around economic orientations, institution did research showing with high agreement with phrases such as that greater ethnic diversity leads “The rich should be stripped of their to increased societal tension and belongings and status.” The second poorer social outcomes, would you focuses on culture, notably “Deep down, support or oppose efforts by just about all conservatives are racist, students/the administration to let sexist, and homophobic.” A third is the staff member know that they focused on authority, i.e., “I must line up should find work elsewhere? behind strong leaders who have the will to [Support, oppose, neither support stamp out prejudice and intolerance.” The nor oppose, don’t know] last two are especially germane to my analysis. The underlying orientation is a 2. If a staff member in your preference for using power or violence to institution did research showing upend the current moral order and replace that the British empire did more it with a new regime that compels people good than harm, would you support to conform to new principles. This or oppose efforts by students/the orientation is characterized by moral administration to let the staff absolutism, intolerance of dissent, and a member know that they should find preference for censorship, exhibiting work elsewhere? [Support, oppose, considerable overlap on many neither support nor oppose, don’t psychological measures with right-wing know] authoritarianism.10

Testing Support for “Cancel Culture” 3. If a staff member in your Among Academics institution did research showing that children do better when How does left-wing authoritarianism brought up by two biological manifest itself within the professoriate? parents than by single or adoptive Having surveyed the hardest forms of parents, would you support or internal discipline in academia, I next oppose efforts by students/the move to examine the level of backing for administration to let the staff “cancel”-style dismissal measures. To do member know that they should find so, I probe responses to controversial work elsewhere? [Support, oppose, research findings adapted from actual neither support nor oppose, don’t cases, but worded in an abstract enough know] way that individuals are unlikely to make a connection to a particular high-profile 4. Please imagine a member of your incident. While most people do not organization has done work endorse these research statements, and showing that having a higher share expect that most academics feel likewise, I of women and ethnic minorities in sought to use them as a litmus test, to organizations correlates with

reduced organizational willing to back campaigns to fire performance. Several thousand academics who dissent from norms on hot- professionals, some from your button issues. organization, have signed an open As a comparison, I include a question letter calling for the staff member on whether academics would support or to be fired in order to protect oppose firing a member of staff who wants disadvantaged groups from a immigration to be reduced. Eight percent hostile learning environment. A of US faculty would favor the dismissal of small group have started a counter- a restrictionist academic, but 78% would petition defending the staff oppose this, with 14% uncertain. Most of member on grounds of academic those surveyed would undoubtedly take a freedom. Would you: [a) Sign the progressive position on these five open letter, which called for the questions, but most also recoil from staff member to be fired, b) campaigns to remove academics who Support the views expressed in the adopt a dissenting view on them in their open letter, but choose not to sign research. it, c) Not support nor sign either The flip side of this largely positive letter, d) Support the counter- portrait, of course, is that there is a 7-18% petition, but choose not to sign it, minority of American faculty who would e) Sign the counter-petition, f) support dismissal campaigns that directly Don’t know.] violate academic freedom. In addition, the share of academics who oppose a given Results are displayed in Figure 7. dismissal campaign is not higher than 52% These show American faculty’s for any issue other than the immigration willingness to support a “cancel” question, and falls to just 31% in the campaign to dismiss an academic on the example of a study finding that a higher one hand, or, on the other hand, to oppose share of women and minorities lowers it publicly or privately. Findings reveal an organizational performance. Many are important reservoir of support for noncommittal, including about half of academic freedom among staff at US those on the question just cited – unwilling universities, with just 7-18% of lecturers to cancel but also unwilling to oppose and professors, depending on the issue, those who would seek to do so.

Support or Oppose Campaign to Oust a Dissenting Academic, US Social Science/Humanities 100%

80%

60% 76% 40% 50% 49% 52% 31% 20% 18% 0% 7% 8% 8% 8% Diversity as Empire as Traditional Women and Restrict Negative Positive Parenthood Minorities Lower Immigration Better Performance Support Oppose Unsure/Neutral

Figure 7. Source: US online mailed survey. August 2020. British YouGov data, in Figure 8, only 54% who would oppose these show very similar findings. Just 6-13% actions. support dismissal across the four As in America, the softest campaigns, but no more than 51% would opposition to dismissal was in the “women oppose a dismissal campaign. Eighty-three and minorities lower performance” case, percent of British academics would oppose with just 27% willing to oppose a firing a member of staff who favored dismissal campaign and 6 in 10 unsure. reducing immigration, slightly higher than The British mailout online survey showed in the US data. For the British mailed results nearly identical to those in the survey, the numbers are similar: 4-18% YouGov survey. favoring dismissal, but with a maximum of

Support or Oppose Campaign to Oust a Dissenting Academic, UK Social Science/Humanities

100% 90% 80% 70% 60% 50% 83% 40% 53% 50% 54% 30% 27% 20% 10% 9% 12% 10% 13% 0% 8% Diversity as Empire as Traditional Women and Restrict Negative Positive Parenthood Minorities Lower Immigration Better Performance Support Oppose Unsure/Neutral

Figure 8. Source: UK YouGov survey. March 2020.

These findings replicate with PhD dismissal, and for the diversity- students: in the UK, just 25% would organizational case, fully 41% back oppose dismissal in the “diversity lowers cancellation. These figures are worrying performance” case, but 65% would in the inasmuch as they concern the generation immigration restrictionist case and 31-34% entering academia. It seems that acquiring would for the diversity, empire, and a post generates a new appreciation for job parenthood questions. For North American security, countering these impulses PhDs the analogous numbers are similar: a somewhat, but, as we shall see, younger mere 19% opposing cancellation in the academics are more politically intolerant diversity-organizational performance case, than their elders, portending a rising 50% opposing dismissal for an academic illiberalism problem. who backs immigration restriction, and 29- The general pattern, therefore, is 50% against firing on the diversity, that the majority of academics is empire, and parenthood hypotheticals. foursquare against cancellation only in the PhD students on both sides of the case of an immigration restrictionist Atlantic are noticeably more likely to academic. For the hypothetical case of a endorse dismissal than academics: in the scholar finding that race and gender UK, 9-10% of PhD candidates back diversity reduces organizational dismissal for immigration restrictionists performance, only 19-31% of academics and those doing research on the or PhDs across all five surveys said they parenthood, diversity, and empire would oppose, publicly or privately, an questions, rising to 27% for the diversity- open-letter campaign to get the academic organizational question. In North America, fired even as few actively endorse a cancel fully 25% of PhD students would seek to campaign. force an immigration restrictionist academic from their job. On three other scenarios, between 11 and 17% back

Who Backs Dismissal? Comparing across the US, Britain, and Canada in Figure 9, including Compiling responses to the first academics and PhD students, shows that a four binary questions (all except “restrict majority of both staff and PhD students do immigration”) in Figures 7 and 8 above not support dismissal campaigns on hot- into an index allows us to ask about the button issues around race, gender, and characteristics of people who support sexuality. Indeed, only 2% of American illiberalism when it advances the cause of academics and PhD students support all perceived racial and gender justice, four dismissal campaigns. broadly construed. Combining the four However, fully half of North questions, each with answers as 0 for anti- American social sciences and humanities dismissal or 1 for pro-dismissal, into one (SSH) PhD students backed at least one measure (also scored as 0 or 1) allows me illiberal measure, as did 36% of UK PhDs to develop an approximate “likelihood to and 20-25% of academics. PhDs are more expel” score for each person in the authoritarian than academic staff. Is this surveys.11 because they don’t have a job to worry Using this measure, an average of about losing? Perhaps. But, as we shall 20% of current UK faculty, rising to 25% see, another (less powerful) reason is in the social sciences and humanities, back because they are younger than professors. at least one dismissal campaign. On the Age plays an important role among both other hand, just 5% of current British professors and PhD students in predicting academics in the social sciences and their willingness to cancel an academic for humanities back all four illiberal measures. transgressing progressive values.

Support Campaign to Oust Dissenting Academic (SSH staff/PhDs only) 60 49 50 43 40 36 30 30 24 25 23 18 18 20 20 16 12 12 11 12 % favoring dismissal %favoring 10 7 8 8 8 9 8 8 8 8 8 8 10 5 6 4 0 US (706) UK (235) Canada (260) US/Cdn PhDs (124) UK PhDs (77)

Diversity as Negative Empire as Positive Traditional Parenthood Better Women and Minorities Lower Performance Restrict Immigration Support One or More Dismissal Campaigns

Figure 9. Note: number of respondents given in parentheses for each category. Figure 9 above shows that 24% of rough guide, therefore, around 1 in 10 American SSH academics, 20% of SSH academics back any single dismissal Canadian SSH academics and 25% of campaign and close to a quarter support at current British SSH staff endorse at least least one of the four hypothetical cases one of four dismissal actions – even offered. Against this, between a quarter though only 1-2% back all of them. As a

and a half oppose a given dismissal Far Left Activists more likely to Back campaign. Dismissal Yet in almost all cases there is a large undecided group. Around half are Far-left and activist scholars are a unsure in 3 of 5 cases – they neither substantial minority of SSH academics in support dismissal nor oppose it. By North America and Britain. Figure 10 contrast, only 1 in 5 are unsure on the shows that far-left scholars who agree, to a immigration restrictionist case while more greater or lesser extent, with the statement than half are unsure on the diversity and “I would consider myself as an activist” organizational performance case. The make up between 8 and 18% of SSH staff unsure group is often the silent majority across the US, Britain, and Canada, with among staff and PhDs when it comes to far leftists at 16-28% and self-described the more controversial cases. I would activists at 26-38%. The British YouGov argue that this lack of certainty reflects a sample, which is arguably the most cross-pressuring between a defense of representative, shows a lower share of far- academic freedom and job tenure on the left activists than the UK mailout survey. one hand, and a desire to protect On the other hand, the PhD surveys, which disadvantaged identity groups on the other. captured 62-85% of PhD students on For instance, as we shall see, most British Prolific Academic, found that 35% of 124 academics diverge substantially from the North American SSH PhDs identified as wider UK public by overwhelmingly far left and 44% as activists. In Britain, the (76%) supporting the notion that the 77 SSH PhDs were less likely to self- virtues of political correctness in identify in these ways, with 22% far left protecting minorities outweigh its threat to and 19% activist. free speech.

Far-Left and Activist Share of Current SSH Staff 40 38 36 35 31 30 28 26 25 25 22 18

% 20 16 15 13 13

10 8

0 US SSH (706) UK SSH Yougov (235) Canada SSH (260) UK SSH mailout (216) Axis Title

Far Left Activist Far Left Activist

Figure 10. Note: number of respondents per survey in parentheses.

Ideology and Intolerance Canadian, and 41% of British academics back at least one dismissal campaign. Illiberalism is much higher among academics who identify as far left, Are Young Scholars Less Tolerant? especially if activist. Far-left activists are often in the forefront of “social justice” Even among far-left activists, it is activism within the university, and many noteworthy that a bare majority would not challenges to academic freedom stem from support forcing dissenting voices from the this group. academy in any given case. The UK data Thirty-seven percent of US in Figure 11 show how support for academics, 40% of Canadians, and 40% of dismissal rises as we move from all British faculty who identify as far left current and retired staff surveyed, to favor at least one firing campaign. Among current staff only, to current SSH staff, far-left activists, 40% of US, 51% of and then to far-left and far-left activist faculty.

. Support Campaign to Oust Dissenting Academic

35% (UK academics) 33% 31% 30%

24% 25% 22% 20% 20% 20% 18% 18% 16% 15% 15% 13% 12% 9% 10% 8% 10% 7% 8% 6% 6% 6% 5%

0% Current & Retired Current (484) Current Social Far Left (85) Far Left Activist (47) (820) Science/Humanities (235)

Diversity as Negative Empire as Positive

Traditional Parenthood Better Women and Minorities Lower Performance

Figure 11. Note: number of respondents given in parentheses for each category. Is age still important when we take outcome measure. The results for Britain a person’s ideology into account? Maybe are shown in Figure 12. young people are just more leftist, so the Ideology and age are the main two factors are confounded? To answer predictors of a scholar’s support for this question, I construct a model, with dismissal and have separate and important willingness to dismiss (0 for no, 1 for yes) effects. Left-right ideology is the most on any of the four questions as my important predictor of viewpoint intolerance on the four aforementioned

items in Britain, but only “very left” (when or moderate, tend to be more illiberal than contrasted to centrist) is statistically older staff. This is concerning, as it points significant in picking out the intolerant to the possibility that the potential for academics in the British YouGov sample, authoritarian activism is likely to rise in with those on the moderate left not more the years ahead. Much turns on whether likely to favor hard authoritarianism than this reflects the fact that younger those in the center.12 The average academics prioritize social justice over likelihood of an academic endorsing a academic freedom more than senior staff; dismissal campaign rises 50 points as we or whether this is a maturity issue, with move from a “very right” academic to younger staff likely to outgrow their someone on the far left. relative illiberalism as they age. But age also matters. Younger staff, regardless of whether they are far left

Predictors of Supporting a Dismissal Campaign (UK)

Very Left*** 0.220 Age (younger)** 0.151 Ethnic minority+ 0.063 Fairly Left 0.055 Political Attention 0.053 Activist 0.052 Current Staff 0.038 Soc Sci/Humanities 0.026 Female 0.008

Standardized Effect

Figure 12. Note: R2=.088. Reports standardized beta coefficients. Significance at +p<.1, *p<.05, **p<.01, ***p<.001. N=820.

Among the other characteristics on the race-related issues. Being female that one might assume are associated with was not a statistically significant predictor favoring speech restrictions in the name of of opinion on any campaign. SSH sensitivity are being female and nonwhite. academics were not more illiberal once I However, in the model in Figure 12 above, controlled for ideology. This is because neither reached conventional statistical SSH disciplines are more far left than significance. Ethnic minorities were STEM, and when this is screened out, the slightly more favorable to removing effect of SSH falls away. British academics that perform dissenting research academics who share their opinions on on racially-sensitive issue topics when I Twitter are also somewhat more likely to disaggregated the outcome to focus only

back expulsion, though this is only express sentiments (such as feeling significant in some models. censored) that are more similar to The situation in North America in conservatives whereas in Britain centrists Figure 13 shows a very similar pattern, report low self-censorship and seem closer albeit with some shifting of the predictors to moderate leftists in their appraisal of and a stronger overall model. Far leftism is questions of academic freedom. again a powerful predictor of illiberalism, In addition, women and minorities along with age. Younger staff are are more illiberal than men and whites in consistently more likely to favor dismissal North America, whereas these – indeed this is the most powerful characteristics did not predict higher levels predictor in the North American model. of support for dismissal in Britain. The US Interestingly, for North America, SSH findings for minorities are accounted for academics are again no more illiberal than by a somewhat higher effect among their STEM colleagues when you control African-American respondents (N=36), for the fact that SSH has more far-leftist although black academics were not more members of staff. illiberal than others in Canada or Britain, Activist and moderate-left granted with very small numbers being academics are also more supportive of polled. Finally, there was no significant dismissal in North America than centrist difference between American and staff, whereas in Britain this was not Canadian academics in their support for significant. These findings reflect the fact dismissal campaigns.13 that centrist academics in North America

Predictors of Supporting Any Dismissal Campaign (US/Canada)

Age(younger)*** 0.205

Very left*** 0.182

Activist*** 0.130

Minority** 0.093

Fairly left* 0.083

Female** 0.077

Soc Sci/Humanities 0.042

American 0.026

0.000 0.050 0.100 0.150 0.200 0.250 Standardized Effect

Figure 13. Note: R2=.133, N=1,093. Reports standardized beta coefficients. Significance at *p<.05, **p<.01, ***p<.001. Concentrating on the main drivers between age 30 and 70, the chance of of illiberalism in the British YouGov data backing the ouster of an academic for results in the chart in Figure 14. This politically incorrect research decreases by shows that being far left increases the around 15 points. Together, they exert a chance of favoring a dismissal campaign powerful effect. Thus a 30-year-old far-left by about 25 points. Age matters as well: British academic has a 1 in 2 chance of

endorsing one of the dismissal campaigns, while a 70-year-old who is not far left has barely a 1 in 10 chance of doing so.

Figure 14. Pseudo-R2=.128. Far Left and age are significant at the p<.01 level. Among both staff and PhD contain more women and minorities and students, there is a significant age are younger) show that PhD students are gradation, with younger staff more more intolerant than academics, even intolerant. Are PhD students more pro- controlling for age. However, the cancellation than academics? Figure 15 difference is concentrated among STEM compares American PhD students and PhDs. Among scholars in the social academics in the 23-33 age range, which sciences and humanities, there is no contains 8 in 10 PhD students in my detectable difference between PhDs and sample. There is also an important number academics (this is confirmed in a model of academics in this age bracket in my that excludes STEM participants). In academic survey, resulting in a combined addition, combining American, Canadian, dataset for this age group consisting of and British respondents weakens the 60% PhD students and 40% academics. difference between academics and PhD Results of a model that controls for students in Figure 15, leaving it gender, race, and age (PhD students statistically insignificant.

Figure 15. Note: N=361 between ages 23 and 33. 60% PhDs, 40% academics. Controls for age, gender and race. Dismissals model Pseudo-R2 = .073, and for diversity model = .069. PhD student is significant in dismissals model at p<. 05, and borderline in diversity case model.

To examine the raw effect of Age is the strongest predictor of support studying for a PhD as compared to holding for dismissal campaigns, followed closely an academic position, where a person may by left-right ideology. PhD status is also feel the need to protect against risk to job strongly significant, along with being loss, I run a model in Figure 16 looking at female rather than male. Being North all age groups using a combined sample of American rather than British, or in STEM 2 North American and 3 British surveys, rather than SSH, does not predict pro- combining PhD students and academics. cancel sentiment.

Predictors of Support for any Dismissal Campaigns -.173 Age***

Ideology (left-right)*** -.159

PhD not Academic*** .091

Female*** .085

White+ -.038

SSH -.027

US and Canada not Britain .016

-0.200 -0.150 -0.100 -0.050 0.000 0.050 0.100 0.150 standardized effect (beta)

Figure 16. N=2,633 (1,431 North America, 1,202 Britain; 2,129 academic, 504 PhD). Pseudo-R2=.107, OLS R2=.117. ***p<.001; +p<.1.

There is an important wrinkle, PhDs, and a 32% chance for British however, that makes the transatlantic academics. While there is only have dimension important. Namely that sufficient data points for PhDs 35 and American, and to a lesser degree under in North America, it is clear they are Canadian, PhD students are noticeably more intolerant than academics the same more pro-cancellation than either British age, especially in the US where Millennial PhD students or American and Canadian academics are 18 points (38% vs. 56%) academics. This is evident in the model in more tolerant than PhDs of the same Figure 17. This shows that American PhD cohort. students 35 and under (Millennials) have a Millennial academics are twice as 56% chance of backing at least one of the intolerant as those 55-64 in Britain, four dismissal campaigns and Canadian PhDs times as intolerant in Canada, and close to the same age have a 44% chance of doing three times as intolerant in the US. While so. younger academics are more pro- This compares to a 41% chance of cancellation, there also appears to be a supporting at least one dismissal campaign discontinuity, with the Millennial cohort among academics 35 and under in Canada, substantially more intolerant than those in a 38% chance among young American the next 36-45 cohort in three of four cases academics, a 36% likelihood for British where I have comparable data.

Support for Any Dismissal Campaign, by Position and Age Group (US, Canada, Britain) .60 .56 .50 .44

.41 .40 .38 .36 .32 .30 .27 .24 .23 .23 .16 .17 .20 .22 .20 .14 .16

probability of supporting supporting dismissal of probability .20 .10 .16 .10 .10

.00 35 and under 36-45 46-54 55-64 65 and over Canada Academic US Academic Britain Academic Canada PhD US PhD Britain PhD

Figure 17. N= 1,098 US, 323 Canada, 1,093 Britain. Controls for age, gender, SSH/STEM and race. US model Pseudo-R2 = .147, Canada model = .089 and British model = .037.

Age matters more than any other freedom-oriented than younger graduate factor for political toleration. In addition, students, which could be due to workplace being a Millennial is also important, with socialization, role change, or the desire to those 35 and under especially likely to be protected from the risk of dismissal. support dismissal campaigns even when Nevertheless, young academics remain accounting for their age (continuously only half as tolerant as older academics. measured in years). While PhD students Intervening to familiarize undergraduate, are significantly less tolerant than master’s, and PhD students with the academics, this mainly concerns STEM history and importance of academic PhD students in North America being freedom would seem an urgent task. Later significantly more pro-dismissal than we shall see, however, that by the time a STEM academics. student enters a PhD program, their views Support for academic freedom may on these questions may be difficult to shift. wane as older academics retire, and as a The emerging problem of less tolerant cohort of Millennial PhD Millennial illiberalism may stem from a students and academics replaces the more rising culture of intolerance among free speech-oriented Boomer and Xer younger cohorts rooted in a feelings-based generations. It is encouraging that ethic that some have termed “therapeutic Millennial academics appear more totalitarianism.”14 Among American

undergraduates, for instance, support for • Some racial groups are less free speech appears to be very limited, intelligent than others? 85% with a reflexive intolerance for most forms opposed. of controversial speech. The Foundation • The US should support Israeli for Individual Rights in Education military policy? 50% opposed. (FIRE)’s 2020 survey asked 20,000 • All white people are racist? 74% undergraduate students enrolled in four- opposed. year programs at 55 leading US • Transgender people have a universities the following: mental disorder? 72% opposed. • Christianity has a negative Would you support or oppose your influence on society? 56% school ALLOWING a speaker on opposed. campus who promotes the following idea: Only the small minority of conservative students consistently backed • Abortion should be completely free speech across all questions.15 Figure illegal? 72% opposed. 18 shows that just 7-18% of US • Black Lives Matter is a hate undergraduates show “strong support” for group? 75% opposed. free speech, with strong opposition at 40% • Censoring the news media is or higher across five questions. necessary? 64% opposed.

Figure 18. Source: FIRE 2020 College Free Speech Rankings, p. 19. These findings resemble those of acceptable to shout down a speaker, and other surveys. A 2018 Knight Foundation 44% agreed that “people who don’t respect survey found that students, by a 53-46% others don’t deserve the right of free margin, favored the aim of a “diverse and speech.”16 Academics are more tolerant inclusive society” over protecting free than PhD students, who might be more speech. Thirty-seven percent said it was tolerant than undergraduates. However, it

is unclear whether the rising cohort of new preference falsification: in almost all academics will import the intolerant ethos surveys, those who opposed dismissal of the current generation of students from were more likely to say they would elite universities into the professoriate. publicly sign a counter-petition than remain silently opposed. Conformity or True Belief? Of course, it may be that when the rubber hits the road those who say they According to Timur Kuran, many would sign the counter-petition against people may appear to support an dismissal get cold feet. Arif Ahmed, a authoritarian regime while privately Reader (Associate Professor) in opposing it. Faced with harsh sanctions for Philosophy at the University of Cambridge speaking out, many come to believe that reports that it was very difficult for him to others support the regime when in fact acquire the 25 signatures needed to get his most oppose it.17 Sunstein terms this free speech motion on the ballot at Cambridge, but it passed with 80% faculty “pluralistic ignorance,” arguing that this 19 characterized mass publics during the Nazi support. and Soviet regimes, and does so in Even so, the notion that there is a organizational life today.18 Is this the silent majority motivated to oppose nature of contemporary academia with cancellation but afraid to do so does not respect to support for cancelling receive backing from the data presented academics on “social justice” grounds? here. If this were the case, I would expect That is, do most professors oppose cancel to see a clear majority opposed to culture but remain silent? My fourth dismissal across all questions, but with illiberalism question helps shed light on most saying they would only do so the problem. Recall the earlier question on privately. how individuals would respond to a In fact, there is a large group that is finding that more women and minorities in normatively undecided. Even if there was organizations correlates with worse no fear of speaking out, there might not be performance. a majority opposing a dismissal campaign. Among social science and As Figure 19 shows, the pro- and anti-free humanities academics currently in post in speech positions have broadly similar Britain, just 7% would sign the open letter backing, but there is a large group of denouncing the professor. Another 11% undecided or neutral respondents, would support the views expressed in it. comprising 30-60% of the faculty, Sixty percent wouldn’t support either side depending on the question. I found a similar result with students in my previous or said they didn’t know. Nine percent 20 would support but not sign the counter- report. petition while 13% would sign a counter- Moreover, as Figure 19 – which is petition. These figures suggest that while limited to those on the pro-freedom side – only 18% of serving British SSH shows, most of those on the pro-freedom academics take an illiberal stance, just side are willing to nail their colors to the 13% would publicly stand for academic mast. Indeed, slightly more are likely to freedom against a “social justice” say they would sign publicly than are challenge. Another 9% fit Kuran’s supporters of dismissal. The only preference falsification pattern of privately exception was among North American backing the counter-petition while staying (88% American, 12% Canadian) PhD silent. students, where there is more evidence for Looking across the range of five a pattern of preference falsification in surveys in North America and Britain in which 14 of 22 (about 70%) of people Figure 19 shows limited evidence for

would support but not sign their name to an anti-dismissal counter-petition.

Not Silent, Not a Majority (Open and Secret Opposition to Dismissal) 20 19 19 18 18 17 16 16 14 14 13 12 11 10 9 9 8 8 7

6 %option selecting 4 2 0 UK US Canada UK Mailout UK PhDs N America PhDs

Sign counter-petition Support but not sign counter-petition

Figure 19.

Cross-Pressured Between Value against a campaign to fire them is Commitments generally not higher than 50%. Even among right-leaning Earlier, we saw that more than 6 in professors and lecturers, just half in 10 British academics and 5 in 10 Britain and 60% in North America express American academics are noncommittal on private or public support for the academic whether to defend the hypothetical freedom counter-signatories in the academic who found that diversity diversity-organizational example. This worsens organizational performance. indicates that many academics have not Many in the academy are torn between the made up their mind on questions where claims of self-esteem egalitarianism and academic freedom and “social justice” those of freedom and reason. The upshot collide. This reticence likely also informs of this is that the problem of academic their belief that opposing such measures freedom in academia is predominantly a may out them as a heartless conservative. battle for hearts and minds, rather than just It does not support the idea that a clear a collective action problem like Kuran’s “silent” majority favors academic freedom “private truths, public lies” that we see but is too scared to speak up: the among those who conceal their private “emperor’s new clothes” scenario in which views in authoritarian regimes. Among it takes just one person to prick the North American and British academics, balloon. there are few active authoritarians, but the The noncommittal stance of most share who care enough about academic staff may also have to do with the fact that freedom to even privately support (i.e., free speech activism on behalf of without saying anything) an academic politically incorrect researchers may be implicitly associated with the right. In

addition, many on the left may implicitly with significantly greater support for accept the group affirmation that actively promoting academic freedom extremists provide because, as Clark notes coming from those on the right compared in a review of the psychology literature on to those in the political center. political tribalism, “Extreme, rigid, Gender seems to match left-right dogmatic defenders of our political ideology in explanatory power. The model ingroups demonstrate tribal loyalty that in Figure 20 shows, controlling for can be appealing to those of us who care ideology, that there is a 54% chance of a about our political coalition’s success, 70-year-old non-leftist British male even if we are more moderate or have academic publicly or privately backing a more nuanced beliefs and policy counter-petition, but this falls to 17% preferences.”21 among a 30-year-old female non-leftist Focusing only on centrist and academic. Even among conservatives and conservative academics who do not centrists, women and the young appear to endorse dismissal – options c) through e) be more cross-pressured between above, and controlling for ideology, I find progressive and academic freedom that women and younger lecturers are commitments. especially likely to opt to remain neutral,

Figure 20. Pseudo-R2=.072. Female and age are significant at the p<.01 level. Combining British, American, and right-wing academics and PhD students Canadian respondents, we see a similar say they would publicly oppose dismissal pattern. What is vital for the report, compared to 21% of 1,719 respondents on however, is to appreciate the connection the left. Among those under 40, who between viewpoint diversity and resistance represent the future of academia, public to intolerance. 35% of 700 centrist and 214 opposition is even more concentrated

among conservative academics (34%) that were unjust, unhealthy, compared to centrists (20%) or leftists repressive, and oppressive. These (11%). Thus the problem of hard traditional societies were authoritarianism does not simply arise reprehensible because of their because the larger post-1990s pool of far- deep-rooted inequality, left activists in the professoriate produces exploitation, and irrational more agitation for disciplinary action and, traditionalism ...there is much work as a result, a chilling effect. It also to be done to dismantle the involves a weakening of resistance to powerful vestiges of inequality, authoritarianism due to the paucity of exploitation, and repression.23 centrists and, especially, conservatives. This limits the ranks of those who would On Smith’s account of the liberal actively resist authoritarian measures, progress narrative, a set of “social justice”- clearing the path for institutionalized based normative commitments are illiberalism. prioritized above even the search for truth. The question of political correctness is a A Lopsided Trade-Off: Free Speech and useful barometer for assessing how far Political Correctness in Academia outside the mainstream academia lies. More in Common’s Hidden Tribes report The cross-pressuring among segments US and British society into academics on the question of dismissal values “tribes” based on their survey does not mean that most faculty resemble responses to ideological questions. The the median voter. In reality, the typical “Progressive Activist” segment of society SSH academic trades free speech and makes up 8% of the American adult emotional safety off considerably further population and 13% in Britain. It is largely in the direction of the latter. white, urban, and professional, leaning Noting the way that processes of consistently left across all issues, and social closure operate, some argue that bulks larger in academia than elsewhere. academia is becoming a moral community, In Britain, this group is 6 times more with a set of norms that constrain the likely than average to post political content questions scholars may ask. What is on social media, and in America, 3 times termed “social closure” within particular more likely. When it comes to political occupations like the arts or academia correctness, Progressive Activists stand involves a boundary being drawn around a apart from all other groups in support. For community, so enabling a group-based instance, 80% of Americans say “political identity to be constructed.22 At the correctness has gone too far” but just 30% extreme, some consider these norms to be of Progressive Activists agree. In Britain, the comparable numbers are 72% and sacred values, with the campus a “safe 24 space” that must be kept pure, free of 28%. heretics who would profane values of Younger Americans are more pro- cultural equity and diversity. For instance, PC. Within the general population, the sociologist Christian Smith speaks of a 2016 American National Election Study teleological “liberal progress narrative” (ANES) pilot survey asked: that has come to dominate the field of sociology in America. He summarizes it as There’s been a lot of talk lately follows: about “political correctness.” Some people think that the way people Once upon a time, the vast talk needs to change with the times majority of human persons suffered to be more sensitive to people from in societies and social institutions different backgrounds. Others think

that this has already gone too far born in 1940. Though ideology and party and many people are just too easily identity are far stronger predictors, age is offended. Which is closer to your still associated with support for political opinion? correctness among American whites when other variables are held at their mean Among whites, and controlling for values. That is, even when screening out education level, Figure 21 shows that the ideology and partisanship, a 16-year-old youngest respondents (born in 2000) are, remains 15 points more politically correct on average, 25 points more likely to favor than a 76-year-old. the politically correct position than those

Figure 21. Source: ANES 2016 pilot study. N=874, Pseudo-R2 = .02. Full model with ideology, party identity, income and gender has a Pseudo-R2 of .153.

Most Britons and Americans Thinking about political oppose PC. However, most also oppose correctness, are you generally in hate speech, which is often used to justify favor of it (it protects against PC. When the question is phrased as a discrimination), or against it (it trade-off between speech restrictions and stifles freedom of speech)? discrimination, a clearer picture of people’s priorities emerges. I use UK data Figure 22 shows that when anti- due to the enormous sample size of discrimination and free speech are set YouGov’s UK panel, from which my against each other in this way, 47% of academic survey is drawn. Consider a 164,000 members of the British public question asked across YouGov’s British oppose political correctness and 37% panel: support it. Against this, the balance among the 603 academics in my YouGov sample

(includes both STEM and SSH, active and against. For comparison, even among the retired) who also answered the PC sample of 56,000 university graduates in question as part of separate polling that is the sample, the pro-PC tilt is just 48-41, included in the Profiles database, is 64% in much narrower than in academia. In a favor and 31% against. Among the 176 related vein, Samuel Abrams finds that in currently employed academics in the the US, 78% of Democratic, but only 39% social sciences and humanities, however, I of Republican, academics support safe find 76% backing PC and a mere 20% spaces.25

Thinking about political correctness, are you generally in favor of it (it protects against discrimination), or against it (it stifles freedom of speech)? (Britain) 80% 76% 70% 64% 60% 47% 48% 50% 41% 37% 40% 31% 30% 20% 20% 10% 0% Public (164k) Degree Public (56k) Academics (603) Current SSH Academics (176)

Favor Neutral/Don't Know Opposed

Figure 22. Source: UK YouGov Profiles data, accessed April 19, 2020; Own YouGov survey matched to Profiles data. Note that only 603 of 820 in my sample could be matched. Sample size in parentheses.

Within the sample of 603 British In academia, two other academics who answered the political characteristics are associated with correctness question in Figure 22, the main increased support for political correctness: predictors of their political correctness are being a professor rather than a lecturer, ideology, politics, and activism, as shown and being female. Given the influence of in Figure 23. This suggests that the professors in running universities, the political skew of academia renders it less former finding is especially concerning for friendly to freedom of speech than other those who fear for academic liberty. No occupations when the competing value of significant difference was found by age, or emotional harm-avoidance comes into by whether an academic is retired or play. active, however, once other factors had been taken into account.

Predictors of support for political correctness among academics (Britain)

Ideology (Right vs Left)*** Labour Voter** Female** Leave voter** Tory Voter* Use Twitter to Share Opinions** Not an Activist* Soc Sci/Humanities Academic* Lecturer, not Professor* Age (older) Minority

-0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 EFFECT SIZE (STANDARDISED BETA)

Figure 23. Note: statistical significance denoted by *p<.05, **p<.01, ***p<.001. For ideology, right is the higher value in the models. Notice that age does not 1,200 mainly non-academic PhD holders significantly predict attitudes to PC among drawn from the UK YouGov panel. Here British academics in Figure 23 above. This there is a big age gap: PhD holders under is in marked contrast to the picture among age 40 are over 30 points more inclined to the general public in the US data in Figure support PC than PhD holders over 60. 21. Figure 24 presents a sample of roughly

Views of Political Correctness, UK PhD Holders, by Age Bracket 80% 70% 67% 60% 55% 48% 50% 41%

% 40% 35% 30% 20% 20% 10% 0% 25-39 40-59 60+ Favour Oppose . Figure 24. Source: YouGov Profiles, Nov 22, 2020. N=1,217, with 340, 463 and 412 across the three age groups. But this age gradation does not academia will henceforth remain PC as exist among academics, especially SSH they age, altering the ideological academics. According to Figure 25, there composition of non-academic professions. is an important divergence between British On this reading, the high support for PC PhD holders who remain in academia and among older academics, especially in SSH those who leave for other fields. The fields, is because they are a left-modernist difference is especially noteworthy among ideological vanguard in society that is academics currently teaching in the social leading where others will follow. If the sciences and humanities, where being latter interpretation is correct, this means older does not appear to correlate with that non-academic professional spheres are diminishing enthusiasm for political likely to become more PC in the years to correctness. come. Therefore, while PhD holders However, the age profile of left- under 40 inside and outside academia are wing frequent social media users is flatter fairly similar in their high support for PC – than it is for PhD holders, indicating that a finding that we’ll see also holds (on ideological fervor counteracts the related measures) for current PhD students conservatism that is associated with being of all ages – there is a wide divergence older (whether for life cycle or between academic and non-academic generational reasons). Among British non- PhDs over age 60. This points to possible academics who are frequent social media academic socialization, especially within users, identify with the Labour Party, and SSH fields, as a stimulant to retaining are far leftist, there is no difference support for political correctness. between young and old. Indeed, the over- But there is another possibility, 60s in this group are as or more PC than with wider implications. Namely, that the their academic counterparts. This evidence change is generational, and PhDs outside suggests, in line with the model in Figure

23, that leftist consciousness, social media own appears to have only a modest impact, use, and gender are central, with age per se if any, on support for PC.26 a smaller factor. Education level on its

Support for Political Correctness, by Age and Occupation, PhD holders (UK) 90% 80% 70% 60%

% support % 50% 40% 30% under 39 40-59 60+ Non-Academic PhD holder Academic (current or retired, STEM and SSH) Current SSH Academic Non-Academic Left Party ID, Frequent Soc Media User Non-Academic Far Left, Left Party ID, Frequent Soc Media User

Figure 25. Source: YouGov Profiles, Nov 22, 2020. N=1,217 PhD holders, with 340, 463 and 412 across the three age groups. N=1127 for left party ID, far left social media users and 14,093 for left party ID social media users; YouGov 2020 UK academic survey, N=603.

Support for political correctness is and feelings for the plight of the poor grounded in empathy for historically helped to mobilize anti-bourgeois disadvantaged groups, especially those genocide in communist regimes. In today’s based on race, gender, and sexuality. Yet, context, empathy for minorities can lead to as Paul Bloom warns, empathy can be a hostility towards whites or conservatives. flawed guide to morality.27 An effect of Rather than see the whole and trust in empathy for a particular person or group empirical regularities to guide one’s anger can be that others who deserve sympathy and calibrate one’s response, powerful disappear from view. Worse, those who attachments and narratives take over, are viewed as harming the group with leading to flawed moral reasoning. which one empathizes may cause one to Moreover, when empathy for a particular become angrier than would be justified by subset of disadvantaged groups and a more wide-angle, evidence-based view. hostility to their purported oppressors This is illustrated by the fact that becomes a leading priority, values other empathy for some identified victims can than cultural egalitarianism – notably be used to rouse people to commit freedom and reason – which universities atrocities. As Bloom remarks, empathy have long proclaimed, may become towards white women was used to fan endangered. anti-black violence in the Jim Crow South,

Decolonizing the Curriculum? writers on the subject to satisfy a quota of 40% females per reading list mandated by Returning to my academic surveys, the University of Lund’s political science I sought to probe faculty views on a less department.28 dramatic test of academic freedom than Figure 26 summarizes the results support for firing heretics: across all surveys, focusing only on SSH subjects. It highlights that this question led Please imagine there was a new more people to jump off the fence, with initiative in the Humanities, Arts the share of neutral and don’t know and Social Sciences at your answers, 16-26%, much lower than in the institution, stipulating that on each dismissal scenario. This is 2 to 3 times less reading list, at least 30% of uncertainty than for the question of readings must come from women whether to dismiss an academic who and 20% from authors of color. argues that diversity impairs organizational performance. Response categories were the same The second major aspect to note as in the diversity-organizational with this question is that a near majority of performance dismissal scenario, permitting SSH academics support the decolonization us to gauge not only people’s views, but agenda across all jurisdictions: between 44 the extent to which academics censor their and 48% of academics in the first four private beliefs: surveys back the idea of racial and gender reading list quotas. In all surveys, this comfortably exceeds the share who oppose a) Publicly express opposition (e.g. at quotas though there is significant a department meeting) opposition: between 29 and 34% of academics, depending on the survey. b) Privately oppose the initiative, but Social science and humanities PhD not say anything publicly students in both Britain and North America are especially enthusiastic about c) Neither support or oppose the quotas, with 61% of British SSH PhDs and initiative 70% of their North American counterparts in support. This tallies with the views of d) Privately support the initiative, but the 1,216 PhD holders under 35 in the not say anything publicly YouGov non-academic sample, who leaned 67-20% in favor of political e) Publicly express support (e.g. at a correctness. While these figures might be department meeting) expected to dip somewhat when PhD students enter academia, have to make up f) Don’t know their syllabi, and find their intellectual wings clipped, it is a worrying sign of how This kind of reading list campaign far the balance between social justice and has occurred in a number of jurisdictions, academic freedom has tipped in the with one Swedish academic, Erik direction of the former among the Ringmar, complaining that he had to youngest cohorts entering the academy. cancel an entire course on fascism due to his inability to find sufficient women

Support Diversity Quotas for Reading Lists? (%) 80 70 70 61 60 51 48 50 44 44

% 40 33 34 30 29 30 19 20 14 10

0 UK YouGov UK Mailout US SSH Canada SSH UK SSH PhD N America SSH SSH PhD Opposed Unsure Support

Figure 26.

Respondents who backed or clear pattern of self-censorship among opposed the reading list quotas tended opponents, here there is more evidence of again to say they would do so publicly a chilling effect on those opposing the rather than keep their beliefs to decolonization agenda. Figure 27 indicates themselves. Naturally, a portion of those that supporters of reading list quotas are who claim they would stand publicly may far more likely than their academic fail to do so in the breach, but Figure 26 freedom opponents to publicly state their does not support a Kuran-Sunstein picture views rather than keep them private. of mass conformity to social justice claims Among academics, supporters are between concealing majority support for academic 4 (315%) and 8 (693%) times more likely freedom. to voice their opinion publicly than I also have data from my survey of privately. By contrast, opponents of National Association of Scholars decolonization are much less willing to go members, where 91% of the 227 public, being just 26-174% more likely to respondents opposed the decolonization speak up than stay silent. initiative and just 3% backed it. Among This tips over into a preponderance the opponents, 61% – two-thirds – said of silence when we shift from academics they would publicly sign a counter-petition to PhD students, where opponents are while 30% said they would keep their between 33 and 58% more likely to stay views private. silent than speak up – reflecting the heavy However, this doesn’t mean there dominance of pro-decolonization is no concealment. Compared to the sentiment among graduate students. dismissal scenario, where there is less of a

Net Willingness to Voice Views Publicly among Opponents and Supporters of Reading List Quotas 800% 693% 700% 591% 600%

500% 473% 411% 400% 315% 300%

200% 174% 130% 100% 47% 26% 39% 0% UK YouGov SSH UK Mailout US SSH Canada SSH -33% -58% -100% UK SSH PhD N America SSH PhD

Opponent Supporter

Figure 27.

In particular, it is noticeable that themselves rather than voice opposition among the 14-20% minority of PhD publicly. This could be related to the students who oppose decolonization greater support for hard authoritarianism quotas, there is more reticence to come out among PhDs compared to academics. This publicly than is the case among professors comports with a 2020 FIRE survey that who oppose quotas. Among PhDs, a finds that conservative students tend to majority of opponents in Figure 28 say self-censor more at relatively liberal they would choose to keep their beliefs to universities.29

Open and Secret Opposition to Decolonization 30

25 24 20 20 20 17 14

% 15 13 12 9 9 10 10 8

5 4

0 UK YouGov UK Mailout US SSH Canada SSH UK SSH PhD N America SSH SSH PhD Oppose Publicly Oppose Privately

Figure 28.

The final point to note is that in b) that course content should this question, across almost all surveys, feature the most intellectually opinion among academics under 40 foundational books and articles generally contained 5-10 points fewer in the field. neutral responses than among the over- 40s, possibly indicating that there are more c) Don’t know. settled views of this question among younger staff, in the direction of both Results are plotted in Figure 29. diversity quotas and, to a lesser degree, These show that academics back content academic freedom. This age-pattern was over quotas in both the UK and North not apparent in the previous question on America. North American faculty and, organizational performance, however, so it especially, graduate students are relatively may not hold more generally. more favorable to quotas over To probe the malleability of this foundational content compared to British group of undecided scholars, I asked faculty, who clearly privilege content over respondents: quotas. Both surveys of UK social science and humanities academics show that they If you had to choose, which of prioritize intellectual merit over quotas by these do you think is more 18-25 points. British PhD students break important: 50-50 between the two options. On the other hand, North American PhD students a) that course content should be incline 15 points more toward quotas over inclusive, representing the foundational texts. Note the difference between these racial and gender makeup of numbers and those in the similar question the students. in Figure 26 above. This time, I introduce

a more explicit trade-off between quotas In Britain, mention of intellectual and intellectual merit (“foundational rigor and the key value of education seems texts”) than was presented in the previous to have swayed academics even more than question asking people to merely support in North America, lowering support from or oppose an initiative “stipulating 44% (YouGov) and 51% (mailout) in the that…at least 30% of readings must come case of support for the diversity initiative from women and 20% from authors of to 32% (YouGov) and 34% (mailout) on color.” Among North American the question of quotas versus foundational academics, support for quotas has slipped texts. So too among PhD students: UK 6 points in the second version (45% PhDs reduce support for the quota option supported quotas in the first question, from 61% in Figure 26 to 46% in Figure falling to 39% support in the second). 29 while North American PhD students lower their support from 70% to 55%.

Favor Decolonizing the Curriculum by Mandatory Quotas 70%

57% 60% 55% 52% 50% 47% 46% 47% 40% 39% 40% 34%

32% % 30%

20%

10%

0% N America PhDs N America academic UK PhDs UK academic mailout UK academic YouGov

Quotas Don't Know Foundational Texts

Figure 29.

Diversity Statement Experiment achievement by supporting diversity and equal opportunity in In order to examine the firmness of its education, services, and these beliefs, I conducted a survey administration, as well as research experiment in each of the five surveys. and creative activity. This Before answering the question, half the organization particularly sample read an equality and diversity acknowledges the acute need to statement modeled on that of the remove barriers to the recruitment, University of California30: retention, and advancement of talented students, faculty, and staff Please imagine an organization is from historically excluded committed to the full realization of populations who are currently its historic promise to recognise underrepresented. Many personally and nurture merit, talent, and identify with underrepresented groups in the curricula (that is, the

stories, histories, findings and but not among Brexit supporters, as Figure research of women and BME 30 shows. Clearly many British academics members, among others) and the have values that render them open to a organization is determined to serve Social Justice appeal. them through an education aligned The typical academic who voted to with social justice and inclusion. remain in the European Union in 2016 (the vast majority of academics) shifted 16 In the British YouGov sample, points after reading the diversity among those who didn’t read a statement, statement, from 27% to 43% support. By just 25% supported the view that course contrast, the minority who voted to Leave content should represent the racial and the EU barely budged, from 13% among gender makeup of students instead of the those reading nothing to just 15% among intellectually foundational texts in a field. those reading the diversity statement, This jumped to 39% among those who showing no statistically significant read the UC diversity statement. The effect difference. seemed to register across most academics,

Figure 30. Pseudo-R2=.119 with controls, .041 without. Treatment statement is significant at the p<.001 level. Leave interaction is not significant at p<.05 level, though Leave sample for treatment is small (N=134).

Another important factor affecting of PC. In addition, as Figure 31 shows, the a person’s view on course content is greater an academic’s support for PC, the support for political correctness. PC more susceptible they are to being supporters were, as one would expect, far convinced of quotas after reading the more supportive of quotas than opponents diversity statement. Here it seems that

institutional initiatives around equity and current social sciences and humanities diversity, like diversity statements, strike academics favor the concept to help fertile soil among the pro-PC majority of combat discrimination, even if it stifles academics: recall that three-quarters of free speech.

Figure 31. Pseudo-R2=.091 with no other controls. Treatment statement is significant at the p<.001 level. All interactions significant at p<.05 level.

Looking across the wider array of Second, groups that already have high results in Figure 32 unearths two patterns. support for quotas – young, women, the First, culturally conservative groups, far left – shift less. It seems that a broad notably Brexiteers (known as “Leavers”) receptive group in the middle – center-left, and those favoring immigration restriction, Remain, men, pro-PC – move most toward have low support for quotas and are not backing quotas. affected by reading the diversity statement.

Support for Diversity Quotas on Reading List 58 58 60 51 48 49 50 44 43 41 40 38 38 39 38 40 31 28 28 30 30 25 23 20 1415 15 % Agreeing with with quotas % Agreeing 1314 10

Read No Statement Read Diversity Statement

Figure 32. Pseudo-R2=.091 with no other controls. Treatment statement is significant at the p<.001 level. All interactions significant at p<.05 level. Source: UK YouGov academic survey. Academic Freedom Statement Experiment The first point to note is that 40- 45% of younger academics (under 40) The diversity statement had a oppose racial and gender quotas for profound effect on British academic reading lists. While younger academics are opinion, suggesting that opinion in the more supportive of the “decolonization” academic center is malleable on this agenda, a near majority are not, and their question. Might it be shifted in the other lack of movement after reading the direction, away from quotas? The answer, diversity statement suggests there is an it appears, is no. important group of young faculty who are In order to test whether academic aware of diversity claims, but reject them freedom arguments might work, I when they conflict with academic introduced a new statement group, in freedom. addition to the diversity statement group That said, the survey experiment and the control group who read nothing: shows that there is a flexible component of academic opinion in Britain, just as there Throughout the ages, those who is with the student opinion I analyzed in a have expressed unorthodox previous co-authored report, which I opinions – on religion, on the reprise in Figure 34. Yet these results only government, on homosexuality, show that academic opinion can be swayed socialism or capitalism – have been by a social justice/diversity appeal, which subject to persecution or is less institutionalized in Britain than censorship. Britain [America] has North America. Might an appeal to free an especially important history of speech shift this group away from the pro- resisting threats to freedom of decolonization position? expression from the authorities. Sometimes these threats take the

form of being killed or jailed, with is contrary to the spirit of free while at other times they involve speech. being turned into a social pariah, losing one’s livelihood or being Figure 33 summarizes the impact subject to reputational shaming. of the two treatment statements on my Many have died for our right to mailout sample of UK SSH academics. freely speak our minds and The results reveal that opposition to exchange ideas, our precious decolonization is similar among both those inheritance. Liberalism is not easy who read nothing (57%) and those who because it is about tolerating ideas read the free speech passage above (60%). we don’t like, which is why it is This shows that reading the free speech rare in history and in much of the vignette has no impact on academics’ world. The role of the university is value priorities when academic freedom to permit a wide variety of and diversity are in tension. However, opinions to be expressed and those who read the diversity statement debated, even if they challenge were nearly 20 points more likely to favor social convention or offend quotas – replicating findings in Figure 32 people’s sensibilities. The quest to from the YouGov data. The “don’t know” teach and research the highest truth responses are similar, indicating that many must prevail over other priorities. changed their view from the “foundational Tolerating only views we agree texts” position to the quotas position after reading the diversity statement.

Effect of Treatments on Support for Decolonization (Britain) 70% 60% 60% 57% 50% 47%* 40%* 40%

% % 29% 30% 27% 20% 10% 0% Diversity Passage Read Nothing Free Speech Passage Oppose Don't Know Support

Figure 33. Source: UK Mailout academic survey. N=222. *p<.05 in chi-squared test.

For the North American academic already well-established among North and PhD surveys, and for British PhDs, American scholars of all ages and younger neither the diversity nor free speech cohorts in British academia, with views experimental passages made any fixed in both directions. Thus new significant difference to their responses to messaging on academic freedom or free the quotas-versus-merit question. This speech to PhD students and faculty may suggests that diversity-based quota logic is have little impact on their trade-off

function between social justice and free just 41% backed free speech among those speech. who read a paragraph about the need for The lack of responsiveness to the emotional safety for disadvantaged groups. free speech message across all academic Thus the free speech treatment moved and PhD surveys stands in contrast to that opinion 14 points (49 to 63%). The of British undergraduates, who are more emotional safety passage had a similar biddable to the free speech side. In impact the other way, shifting views in the previous work, I found that the free speech direction of emotional safety by 14 points passage above had a significant impact on (34 to 48%). Female undergraduates undergraduate student opinion on the shifted more than male undergraduates in question: response to the emotional safety passage while both genders were equally moved by When in doubt, which policy the free speech appeal. should your university support? It appears that British academics and PhD students have relatively fixed a) prioritize free speech, even if views. British academics are amenable this makes people upset; only to a diversity-based appeal for restrictions on academic freedom, not to b) prioritize emotional safety, even free speech appeals. North American if this limits free speech; academics and PhDs have settled views on the matter so are resistant to either c) don’t know. message. Advocating for free speech is therefore more likely to bear fruit if Figure 34 reproduces that work, directed at undergraduates. By the time showing that 63% of undergraduate students have been socialized into the students who read the free speech passage diversity and emotional safety-oriented backed the free speech option but only academic culture at the PhD level, it may 49% of those who read nothing did so, and be too late.

Support for Free Speech or Emotional Safety, British Undergraduate Students, by Type of Paragraph Read

70% 63% 60% 49% 48% 50% 41% 40% 34% 30% 30% 20% 17% 11% 10% 7% 0% Free speech Emotional safety Don't know No paragraph Free speech paragraph Safety paragraph

Figure 34. Source: Simpson, T. and E. Kaufmann. 2019. “Academic Freedom in the UK.”

Age, Gender and Support for Academic Freedom

We have seen that views on the In general, younger faculty lean decolonization versus academic freedom more toward quotas than older faculty. question have crystallized in North Figure 35 shows how academics under 40 America and among British PhDs, and are are, depending on the survey sample, 4-23 relatively impervious to competing points more likely to back the reading list perspectives. This suggests stronger quota initiative than those over 40, while divisions on this question could prevail PhDs under 30 are 10-29 points more going forward. The question this then begs likely to do so than PhDs over 30. The age is whether opinion is shifting away from differences are somewhat larger among merit toward quotas due to generational North American than British academics turnover. In order to explore what may be and doctoral students. coming, I parse the views of faculty and doctoral candidates by age.

Support Decolonization Initiative 65 65 66 60 56 58

47 46 43 44 37

% 29

UK YouGov UK Mailout US SSH Canada SSH UK SSH PhD N America SSH SSH PhD Young (under 40, or PhDs under 30) Older (over 40, or PhDs over 30)

Figure 35.

When the question explicitly to the greater penetration of curriculum juxtaposes “intellectually foundational quota activism in North America, which texts” against reading list quotas (see has roots in the 1960s and with the Figure 29), support for quotas falls across multicultural education movement of the the board, but as Figure 36 demonstrates, late 80s and early 90s. Only recently has the age gap remains – though only in the there been a major push on this front in North American samples. This may point Britain.31

Support Quotas over Foundational Texts 80 70 67 58 60 48 50 43 42 39 37 38 39 % 40 33 35 33 30 20 10 0 UK YouGov UK Mailout US SSH Canada SSH UK SSH PhD N America SSH SSH PhD Young (under 40, or PhDs under 30) Older (over 40, or PhDs over 30)

Figure 36.

Gender is also a consistent speech versus emotional safety predictor of attitudes to diversity/safety questions.32 Having said this, an important versus freedom questions. Figure 37 charts group of between 24% and 34% of female a wide gender divide that ranges from 12 academics and graduate students favor to 38 points but is greater than 23 points in intellectually-foundational texts as the all but one survey. This echoes findings in basis for course reading lists rather than student surveys, which also reveal diversity quotas. consistent gender disparities on free

Support Quotas over Foundational Texts 66 70 62 64 60 56 57 50 44 40 32 34 34

% 29 30 30 24 20 10 0 UK YouGov UK Mailout US Canada UK PhD N America PhD Female Male

Figure 37.

As new generations of PhD Attitudes toward Nonconformity students enter the academy and, if the share of women rises, we should expect the balance of internal opinion to move in If a new initiative mandates that the direction of emotional safety over people meet diversity quotas for reading academic freedom. In addition to limiting lists, what is the punishment for those who the freedom of academics to set the course refuse to yield their academic freedom to texts they believe to be most pertinent for these rules? While one may see diversity teaching, further moves in the direction of as a positive goal, are those in favor emotional safety are likely to mean that willing to own the authoritarian research on controversial topics around implications of such a policy for those race, gender, and sexuality may become who refuse to comply? I put the question increasingly off limits. This circumscribes to my sample of over a thousand American and distorts knowledge produced by the and Canadian academics. Table 2 lists, university, reducing its value.33 among those in favor of quotas, their preferred sanctions, from least to most punitive, for academics who refuse to comply with the quota requirement:

Answer %

Don’t know 15

No action of any kind 12

No formal or informal pressure, but must take extra implicit bias awareness training 28

No formal disincentives, just social pressure 27 Give them less favorable teaching and administrative roles, or access to research funding 7

Cancel the course, make them teach a course that meets the quotas 9

Terminate employment due to breach of contract 3 Table 2. Preferred Punishment for Quota Non-Compliance among Pro-Quota North American Academics. N=474 pro-quota academics in US and Canada (out of total 1,093 sample). The appetite to punish severely is only 2% of academics over 50 selected not as strong as the desire for quotas. Only this option. Those who identified as 3% recommended firing dissenters, though activists and “very left” were similar in 19% supported formal disincentives from their inclination to punish, and being dismissal to being allocated less favorable female had an effect only slightly weaker teaching, research, and administrative than age, activism, and ideology. positions. Another 27% favored social Yet much of this is an artifact of pressure while a further 28% backed these groups’ stronger support for quotas mandatory implicit bias awareness over intellectual merit. When limiting the training. These punishments are the mirror analysis only to those in favor of quotas, image of the complaints outlined in Figure the relative authoritarianism of the under- 6, which found that 38% of right-wing 30s, and lack of it among the over-65s, is PhD students in North America and 43% what jumps out of the models. Ideology of the mainly right-leaning National and being female no longer predict Association of Scholars (NAS) members authoritarianism, though self-described experienced formal or informal sanctions activists are somewhat more inclined to for expressing their views in research, punish than others (this effect was just public fora, and teaching. These showed outside the boundary of conventional that the most prominent form of statistical significance). punishment experienced by What this question does not ask is nonconformists was social bullying, whether people would be willing to followed by adverse teaching or publicly object if the university decided to administrative roles or reduced research dismiss those who resist, or to punish them funding. severely. We saw that for controversial The strongest proponents of severe research, about half of academics opposed punishment were academics under 40, dismissal campaigns, and for less with 11% of them advocating forcing controversial views – such as favoring dissenters to cancel their course and teach reduced immigration – around 8 in 10 did. a quota-compliant module. By contrast A little over half the opponents said they

would be willing to publicly express their d) I haven’t thought about it opposition. Assuming that a recalcitrant much, and don't have a strong academic who does not wish to comply view either way with mandatory reading list quotas attracts responses intermediate between these e) I prioritise academic freedom, positions, this leads me to estimate that but don’t have strong beliefs in close to half of academics would oppose a this area dissenter being terminated, but that little more than a quarter would say so publicly. f) I prioritise academic freedom, This would possibly clear a path for and have strong beliefs in this administrators to adopt a hard area authoritarian position against those who refused to comply with a reading list quota When phrased in these more directive. intellectualized and abstract terms, the balance of answers shifts away from the Academic Freedom or Social Justice? social justice position towards academic freedom. Whereas American and Canadian A related but distinct question is academics favored foundational texts over whether the tendency to favor quotas over quotas 47-39, Figure 38 reveals that they intellectually foundational texts, or to back prioritize academic freedom over social new diversity quota initiatives, reflects a justice by a whopping 56-27 margin (58- general preference for social justice over 26 in the US). Even among North the freedom to seek the truth, wherever it American PhDs, a 15-point lead for the lies. Here the question is: quota position has eroded to just a 6-point lead for the social justice view in this When it comes to tension between formulation. UK PhDs move less: from an the freedom for academics to even position between foundational texts publish research on their interests and quotas to +4 for academic freedom. and concerns over social justice for The difference can likely be attributed to disadvantaged groups, which both the intellectual appeal of the term comes closest to your view? “academic freedom” and the more abstract nature of the question. Elsewhere I find that more concrete trade-offs involving a) I prioritise social justice, and actual people, policies, or events tend to be have strong beliefs in this area resolved more in favor of the social justice position.34 b) I prioritise social justice, but It is also noteworthy that North don’t have strong beliefs in this American PhD students favor social area justice over academic freedom by 6 points (9 points among SSH PhDs), and that c) I have thought about it, but social justice is nearly on par with don’t have a strong view either academic freedom among British PhDs. way This indicates that support for academic freedom may be less robust among the emerging generation of faculty.

Prioritize Academic Freedom or Social Justice? 70

60 58 53 50 40 40 37

34 34 33 % 30 26

0 US Academic Canada Academic N. America PhD UK PhD

Academic Freedom Social Justice

Figure 38.

The correlates of these positions the spectrum, but with more limited are similar to those in the quota versus gender disparities on the right. Those of merit models, with younger scholars, those “other” gender (29 in dataset) also tend to on the far left, activists, and women more favor the social justice option over inclined toward the social justice position. academic freedom. Crucially, men – even Being a racial minority is not statistically far-left men – tend to favor academic significant, bearing out a pattern whereby freedom over social justice. race is a less important correlate of views In terms of statistical effect sizes, on academic freedom questions than age, identifying as an activist had the strongest gender, or ideology. effect on prioritizing social justice, Figure 39 shows how gender and followed closely by left-right ideology, ideology interact to condition attitudes to gender, and age. American academics academic freedom or social justice, with were marginally more likely to favor women increasingly diverging from men academic freedom than Canadian scholars. toward social justice as we move left on

Figure 39. Note: scale runs from 1 to 3 only. Pseudo-R2 in ordered logit is .140; R2 in OLS=.234. N=1,093. Controls for activist, US/Canada and minority. Not Against Academic Freedom, But Not For It, Either

The upshot of our analysis of the even when using my expansive measure of academic freedom vs. social justice support for firing (backing any one of four questions is that most academics are cross- hypothetical campaigns). In any given pressured between their commitments to campaign, the majority of far-left cultural progressivism and their academics will oppose dismissal. attachment to academic liberty and reason. Fear plays a part for an important On a positive note, just 20-25% of SSH minority of pro-free speech academics, professors and lecturers, rising to 36-45% inclining them to remain silent rather than of SSH PhD students in North America advocate publicly for freedom, but sincere and Britain, would support at least one of belief and not fear is the main obstacle to four authoritarian measures against a tackling today’s illiberal campus climate. dissenting conservative academic. This Activists and administrators are the still leaves three-quarters of SSH instigators of authoritarian policies, but in academics as consistent opponents of the absence of the climate of opinion drives to remove professors from their prevailing within the faculty, where half posts for politically incorrect research. Far are hesitant about protecting academic leftists are considerably more likely to freedom when it collides with progressive back illiberalism, but even here, the far- aims, their efforts would meet more left faculty divides fairly evenly between resistance. supporters and opponents of dismissal

The fact that many academics do strongly promoting this agenda. This gives not publicly nail their colors to the internal diversity proponents within the academic freedom mast due to their beliefs professoriate a metrics-based rationale to is not a good sign. This is especially so if move in the direction of mandatory targets younger academics do not become more or even quotas. In North America, such inclined to defend academic freedom as initiatives are even more entrenched as a they ascend the academic ladder. The combination of staff and student activists rising share of women in academia may and better-funded (compared to Britain) also affect active support for academic equity and diversity administrators enforce freedom due to their somewhat higher compliance. These have truly become predilection for diversity arguments – even what Jonathan Haidt and John Ellis term though women are not more likely to “social justice” rather than “truth” endorse authoritarian punishments than universities.37 men who share the same belief system. While Cambridge University’s Most academics and PhD students governing body famously rejected, by an in the United States, Canada, and Britain 80-20 margin, a proposed policy shift that don’t actively oppose academic freedom, would have mandated “respect” for but they don’t actively support it either. identities over toleration of difference, the Most are unlikely to speak up for vote involved just 1 in 8 academics at the colleagues who dissent from social justice- university, and included retired academics inspired restrictions on their academic and some non-academic staff. While my freedom. Thus progressive survey results intimate that the vote would authoritarianism succeeds through a kind still have prevailed had all staff been of sin of omission. This indicates that polled, they also suggest that the final vote intervention would need to occur at the would have been closer.38 institutional level, to ensure activist-driven These findings also indicate that proposals cannot become university rebalancing internal messaging from policy. At present, college authorities are equity-diversity toward academic freedom generally silent on academic freedom or would have a significant impact on subordinate it to the countervailing attitudes among undergraduate students, imperatives of equity, protection from and – to the extent that messaging around harassment, diversity, and college emotional safety for protected groups is reputation. reduced – also among UK professors and Currently, the momentum inside lecturers. Where views have largely the university lies largely with those who crystallized, as among North American would abridge academic freedom and free academics and PhDs, and among UK speech in response to equity and diversity PhDs, messaging is less likely to persuade. claims on behalf of disadvantaged groups. Government policy can play an This is well-established in the American important role in conveying the will of the and Canadian cases.35 But there is also a democratic majority and the letter of the rising volume of diversity initiatives at law (at least among institutions in receipt British universities, spurred on by the of public funds), while issuing guidance to Race Equality Charter (REC) of Advance ensure that university administrators HE and ATHENA SWAN, influential (though not academic staff) act in a charities that get universities to sign up for politically-neutral manner. This their monitoring and compliance pledges, intervention could also help set the tone at building these into their research universities based on what society and the assessment scores.36 Both Universities UK law – as well as most academics – expect. (UUK), the industry body for universities, As Cass Sunstein notes, law often signals and the major sector union, the UCU, are the public mood and can dispel

misperceptions about the approved values scholarly community, it would damage of an organization.39 their careers. This is especially pointed in view of the scarcity of academic positions, with hundreds of applications for Part II: Soft Authoritarianism permanent posts. Hyper-specialization results in limited mobility – especially if a Most academics reject hard researcher wants to live in a given authoritarianism. But the verdict is less geographic area or work at a high-status clear for soft authoritarianism. As noted in university. Meanwhile, reputations travel Part I, faculty matter because they play a quickly through the “invisible colleges” cardinal role in setting the culture and that affect hiring choices, making it very environment of departments on a day-to- difficult for academics to move jobs. This day basis. Whether permanent, or on time- produces even greater risk. As a result, limited contracts as researchers and/or dissenters prudently adapt to this teachers, they do so by shaping how manifestation of John Stuart Mill’s research and teaching are conducted, and “despotism of custom” through self- by being responsible for the variety of censorship, limiting their academic decisions in which, for instance, papers are freedom, and constraining the truth- accepted for publication; grants are seeking mission of the university. awarded or not; job offers and promotions are received; and students are admitted. Prior Work More generally, faculty set the tone. Taken individually, very few In recent years, a growing body of actions by individual faculty members are work has begun to examine the possibility of decisive significance. What is of of discrimination on ideological or interest and significance are larger political lines. “Political” or “ideological” patterns, which help shape the culture of discrimination involves one’s professional the university and ultimately influence the judgment about another person or their public contribution it makes. To what work being affected by the extent to which extent are academics free to pursue they agree with the political or ideological research or public engagement that may orientation of the person. have a political interpretation, without This becomes even more urgent in paying a social or professional penalty? light of the growing political polarization This part of the report examines of the 2000s, initially in the US, but political discrimination against increasingly in Canada and the UK as conservative and gender-critical academics well.40 Indeed, political polarization is and its corollary, the chilling effects connected with political discrimination. produced within these target groups. In For instance, in 2014, Shanto Iyengar and line with the “iceberg” models presented Sean Westwood found that affective earlier, it pursues the question through a polarization – hostile feelings towards focus on both the perpetrators and victims opposing political partisans – was far more of discrimination. The former involves pronounced than negative feelings based half or more of staff and students at the on race. Further, affective polarization university while the latter is narrowly impacts discriminatory behavior to a concentrated on conservative and gender- significantly greater degree than racial critical academics and students. prejudice.41 Targeted scholars either experience While political discrimination direct victimization or feel chilling effects could occur in any sector of society, it is of by anticipating that if their views were to particular interest within academia, where become known to colleagues or the wider political viewpoint may directly shape or

affect the content of someone’s work. their grant applications.46 Honeycutt and Accordingly, some studies focus on Freberg replicated this result nearly political discrimination within academia exactly, but showed that the willingness to specifically. Key contributions to this discriminate was symmetrical on both literature include studies by George political sides. In their sample, 33% of Yancey,42 Yoel Inbar and Joris liberals were willing to discriminate Lammers,43 Nathan Honeycutt and Laura against conservatives in hiring decisions Freberg,44 and Uwe Peters and co- (answers varying from “somewhat” to authors.45 “very much”), which was equally matched Yancey found that 30% of his by conservatives’ willingness to sample would discriminate against a discriminate against liberals, at 32%.47 Republican while 50-60% of This study also found that the more anthropologists and English professors ideologically committed a respondent was, said they would be less likely to hire an the more likely they were to be willing to evangelical Christian. The latter three discriminate.48 From the Peters study, the contributions adopted a similar question highest levels of willingness to wording so can be readily compared. Each discriminate were again reported for hiring used an online mailout survey to gauge the decisions. For a right-leaning hire, over attitudes and opinions of a given sample. 55% of left-leaning philosophers reported Inbar and Lammers sought to evaluate the being willing to discriminate attitudes of psychologists in the US, and “occasionally” to “all the time.” For a left- focused on the attitudes of and towards leaning hire, the equivalent proportion of political conservatives (n = 508). right-leaning philosophers willing to Honeycutt and Freberg extended that discriminate was over 45%.49 study, using a sample of academics in A larger sample was collected in California across disciplines, repeating 2017 by Terrence Karran and Lucy these questions for political liberals in Mallinson.50 These researchers surveyed addition to conservatives (n = 618). Peters 2340 academic members of the University et al. used an international sample of and College Union (UCU), the main union philosophers, and asked about a more for academic staff in Britain and an even variegated set of political orientations and larger complement of European scholars. viewpoints, rather than a simple left-right The UCU survey focused on threats to division (n = 794). These studies academic freedom, germane to my study confirmed, for their respective samples, here, and principally on internal threats to the same central finding as that of Sam academic freedom arising from managerial Abrams (see Figure 40 below), namely favoritism, unspecified forms of that the American professoriate has an discrimination such as methodological overwhelmingly left-liberal orientation tribalism, and imperatives flowing from and a paucity of political conservatives research and teaching assessment relative to the general population. For exercises. instance, in the Peters study, 75% of Qualitative comments gave insight philosophers leaned left, 11% were into a series of “invisible” threats to moderate, and 14% leaned right. academic freedom that are rarely On the willingness to discriminate identified. The most widespread was politically, the findings were dramatic. In psychological bullying from colleagues, Inbar and Lammers’ study, over 1 in 3 of reported by 27% of respondents. There those who identified as liberal would were also reports of disciplinary actions discriminate against conservatives in from department heads or managers. Such hiring decisions, while 1 in 4 would individuals, often senior staff in rotating discriminate against them in reviewing appointed positions at department or

School/Faculty level, hold considerable penalties? Reported chilling effects are power over individual careers. Survey not, in themselves, direct evidence of a respondents mentioned cases of threatened problem. Someone’s fear of penalties may or actual dismissal from employment, be irrational, and should thus be handled increasing an academic’s teaching or through personal resilience rather than administrative load, removing them from a external protection. cherished teaching or administrative To address this, I consider a suite position, berating them during of attitudes that affect the social and performance reviews, and withdrawing professional context in which academics grant funding. Other instances I am aware work. A key aspect of this is the degree to of include departmental or higher-level which the perceived political import of decisions compelling conservative their work may affect their career. To speakers to be balanced by progressives (a assess such political discrimination, I requirement not imposed on left-only focus on attitudes towards grant panels), or repeatedly delaying an event applications, papers submitted for without cancelling it. publication, and applications for These studies have been vital for promotion. Another is the softer but highlighting dynamics within universities arguably no less consequential issue of in terms of both hard and soft workplace social climate. Are academics authoritarianism. Of course, each of these comfortable interacting in an informal studies depended on participants opting in context with colleagues who hold to take the survey. This creates a potential politically-opposing or controversial risk of self-selection bias. Among the views? surveys presented in this report, that risk also applies to the North American From Hard to Soft Authoritarianism academic survey, the NAS academic survey, and the UK mailout academic Hard authoritarianism, being fired survey, but not to the UK YouGov or threatened for one’s views, is a more academic survey or the Prolific surveys of serious violation of academic freedom American, Canadian, and British PhD than soft authoritarianism, not being hired, students. By comparing these samples, we promoted, awarded a grant, or published in are able to show that the online mailout a journal. But both matter for academic samples do not differ systematically from freedom. Active social bullying is more the accidental, non-selective surveys. For punishing than social ostracism, which is more information on survey data and in turn worse than socially avoiding methods, see the Appendix. someone or not including them in one’s In addition, previous surveys rely social circle. Even so, there is a sliding on participants’ willingness to report their scale from hard forms of authoritarianism attitudes in a way that is susceptible to to softer “everyday” exclusions that can social desirability bias. This report is affect a large number of staff. All forms of based on a series of studies that address political sanction matter, however, and this through the use of a concealed list ideally would be kept to a minimum given technique that can reveal the true the importance of collegiality to the willingness of academics to engage in academic enterprise. political discrimination. Part I examined both perpetrators’ Do academics feel comfortable willingness to endorse hard expressing their views on politically- authoritarianism and victims’ experiences salient or controversial issues, and have at the sharp end of such measures. Part II they decided not to express their views in considers soft authoritarianism: political research or teaching because of the risk of discrimination and its impact, in the form

of chilling effects, on nonconforming chilling effect are based in reality, and academics. Here Mill’s “despotism of self-censorship becomes a reasonable custom” rather than activist bullying or response to the threat of genuine antipathy university disciplinary power is what and unequal treatment. ensures compliance. The degree of faculty support for This doesn’t mean people should dismissing academics who take an not be allowed to discriminate in their unpopular position on a sensitive issue is social interactions. There is an important one barometer of whether those who distinction to be drawn between the perceive a chilling effect are correct to do negative duty not to exclude, i.e., “I won’t so. This connects hard authoritarianism to associate with conservatives,” which the soft authoritarianism of public opinion. should be unacceptable, and the positive The perpetrator test here is whether desire to socialize more often with those academics back the idea that someone who share common beliefs or identity, i.e., should lose their job or post, or be subject “I like to associate with neo-Marxists.” to an online campaign, for research that is Limiting discrimination is not coterminous deemed to have contravened a moral with positively mandating patterns of norm. socializing, which is contrary to the spirit of free association that is central to Political Discrimination liberalism. Something similar holds for inviting speakers to campus, where staff or Soft authoritarianism arises from student interests may organically result in political discrimination. That is, the invitations skewed to the left. This is quite willingness of perpetrators to penalize different from actively rejecting someone for their perceived politics. Short conservative speakers by stretching the of dismissal, this may involve definition of racism to encompass their discrimination in hiring, promotion, and views and thereby justify excluding them refereeing; or social ostracism in collegial from campus. spaces. The surveys below test for these In addressing the question of forms of discrimination. Though short of academic freedom, I begin from the dismissal, these less dramatic forms of standpoint of the victims who experience discrimination may nevertheless prove discrimination or anticipate that they will extremely important in shaping the be penalized for their views. Academics’ perceptions of those who might otherwise decisions to restrict what they teach or have chosen to pursue their genuine research, on the basis of concern about research interests. In this manner, speech colleagues’ potential discriminatory is subject to soft authoritarianism. reactions, is what is often termed a Second, I consider the problem “chilling effect.” This is at the heart of soft from the other end: the chilling effects authoritarianism, as opposed to hard reported by the victim side. Here I authoritarianism, which relies on threats of measure how academics perceive their disciplinary action or dismissal. scholarly environment. This sets out what Is there a rational basis for scholars the perceptions are, and whether they – notably conservatives or gender-critical correspond to actual risks. feminists – to feel a chilling effect? To This furnishes two foci for the answer this question, there is a need to work on soft authoritarianism that follows. focus on perpetrator attitudes, namely First, the willingness to engage in political those of the leftist and (to a lesser extent) discrimination; and second, the centrist academic majority toward the experiences of those who report being conservative minority. If these views are subject to (or fear being subject to) indeed penalizing, then perceptions of a viewpoint-based penalties. Both are

connected, as we shall see, to the low and professors in the United States, ideological diversity of the professoriate. Britain, and Canada would not discriminate on political grounds in any Section IIa: The Ideological Evolution way. This includes a slight majority of of Academia those on the left. In each country surveyed, for instance, over half of left-wing academics would not discriminate against The ideological composition of the a Trump or Brexit voter in a job professoriate is germane to this application. conversation. For, as we shall see, when two sides discriminate against each other, The Political Views of Academics there may be parity in discriminatory intent, but a structural imbalance in discriminatory effects because there are There have been several studies of more perpetrators on one side than the the political leanings of academics in the other. This is very much the dynamic in United States and Britain since the 1960s. academia, where there is a pronounced In 1969, a Carnegie Commission survey ideological imbalance among staff. found that 28% of American university Consistent with prior research, I find that a faculty were conservative, 27% centrist, significant majority of academics incline and 45% liberal or left. Thus the left:right left in terms of political outlook. ratio was below 2:1. While Sociology Academics’ attitudes on political subjects leaned left by a 4:1 margin, the ratio in the such as Trump, Brexit, abortion, or Humanities and Law was 2:1 and in the 51 immigration are very distant, in aggregate, Sciences 1:1. In 1984, the Carnegie data from those of the average voter. This is not showed that just 39% of faculty were left simply because the “ivory tower” is or liberal, and the conservative share had somehow an outlier. Those with advanced risen to 34%, a balance approaching 1:1. degrees in all sectors hold views The left share among social science and intermediate between professors and the humanities academics was high but general public. Interestingly, as we shall appeared to have declined from 66% in see in the non-academic part of this study, 1969 to 56% by 1984. This finding led academics as individuals seem to Hamilton and Hargens to claim that “the discriminate on political grounds at similar incidence of leftism [among faculty] has 52 rates to non-academic professionals with been considerably exaggerated.” In 1989 analogous levels of education. and 1997, the Carnegie surveys showed a Yet academia is different for two leftward shift to a 2:1 ratio compared to reasons. First, the scale of the left-liberal 1984. This was confirmed by 1989 and 53 majority is more exaggerated than in most 2001 HERI studies. By the mid-2000s, other professional settings. Second, however, this skepticism was difficult to colleagues’ political views are more sustain. A study by Gross and Simmons visible in their work – at least among (2007) showed that Democratic voters social science and humanities scholars – in outnumbered Republicans in the universities. The structural effect of having professoriate 3.5 to 1, with an 8:1 ratio in so few conservatives and a progressive the social sciences and humanities, the majority, in combination with the same tilt uncovered by two other studies at 54 difficulty in hiding one’s views, results in the time. a high level of net discrimination against Likewise, among American social non-progressives. psychologists in 1960, the ratio of liberals It is vital to emphasize that, on my to conservatives was about 2 to 1. But this evidence, a narrow majority of lecturers began to change in the 1980s, and by 2014, the ratio of liberals to conservatives

in social psychology had reached 15 to 1.55 of tens of thousands of professors from the Thus the initial decline in the number of Higher Education Research Institute conservatives seems to have taken place (HERI) at UCLA, Sam Abrams found that between the early 1980s and early 2000s, a the share identifying as left-wing increased period when the share of conservatives to from around 40% in the mid-1990s to 60% liberals in the population remained by 2010-11.56 This occurred during a relatively unchanged. period when the political leanings of the More recent trends, between the US public remained roughly steady at mid-90s and early 2010s, point to a around 25-30% left, thus wider societal continuing decline in the share of shifts cannot account for the change conservatives among college faculty among the professoriate. Figure 40 shows combined with a new decline of the center the trend. and rise of the left. Using triennial surveys

Figure 40. Source: Faculty Ideology in American Universities, 1989-2014. Abrams, Sam, “Professors Moved Left since 1990s,” Heterodox Academy, Jan. 9, 2016.

More recently, groundbreaking and 13:1 in a sample of 66 liberal arts work by Mitchell Langbert used a colleges. In terms of academic disciplines, comprehensive representative sample of Langbert uncovered a staggering ratio in a tenure-track professors from leading number of social science and humanities research universities (over 7,000) and fields (see Figure 41). In liberal arts liberal arts colleges (over 9,000). Tracking colleges, in Anthropology, the political registrations of professors, he Communications, and the “Studies” fields found that, excluding military and (race, gender, sexuality) he found a perfect religious colleges, the ratio of Democrats monoculture, with no registered to Republicans was 11.5:1 in the SSH Republicans at all.57 Looking across over departments of major research universities 7,000 permanent faculty in Economics,

Journalism, History, Law, and Psychology Moreover, other studies find that while at major research universities, Langbert academic Democrats are solidly left- and colleagues found just 4.3% to be leaning across all policy issues, Republican, compared to 50% Democrats, Republican professors tend to be more 16% unaffiliated, and 29% unregistered. ideologically diverse.58

. Figure 41. Source: Langbert. 2018. “Homogeneous.” Figure 1.

Developments in Britain broadly impact on left-wing Labour and the parallel those in the United States. In Greens, whose combined share was 44%. 1964, 35% of UK academics voted for the By 2015, however, Chris Hanretty, Conservatives and 47% for the left-wing using randomly-sampled Understanding Labour Party, a modest difference. Society data, found that closeness to left- Support for the moderate Liberal wing parties (Labour and Green) had Democrats or SDP rose substantially, reached 68%, with closeness to the reaching 35% by 1989, during Thatcher’s Conservatives down to 15%. Figure 42 tenure in office, and squeezing support for seems to show – even as I cannot the Tories among the professoriate to just definitely prove this given shifts in 18%.59 The Liberal Democrat rise had less question wording and methodology – that

the same shift from centrism to the left that conservative decline followed by a second occurred in America between the early phase in which the left expanded at the 1990s and 2000s also took place in the expense of the center. UK.60 As in the US, we see a first phase of

The Political Composition of UK Academia, 1964-2015 60 55 50 47 40 40 37 35 29 35 % 30 27 20 18 17 15 14 10 7 13 0 0 0 1964 1976 1989 2015 Conservative/UKIP Labour Lib Dem/Liberal/SDP Green SNP/Other

Figure 42. Source: Carl, Noah. 2018. “The Political Attitudes of British Academics.” Open Quantitative Sociology and Political Science; Hanretty, Chris. 2016. “Is the Left Over- Represented within Academia?” Medium.

To triangulate more recent data, I My results are broadly in line with provide a compilation of recent surveys of Understanding Society’s smaller-sized British academic opinion in Figure 43. random sample, as well as the Times This includes self-selected Times Higher Higher’s opt-in surveys, increasing (THES) data, where readers offer their confidence in the results. These show that views, randomly-sampled data from about half of academics vote Labour and Understanding Society, and my UK between 10 and 20% favor the YouGov survey, where a passive sample is Conservatives or UKIP/Brexit Party. approached by a survey firm, ensuring less Restricting to current academics and using bias toward those who like to venture their survey weights, the figures show that 16% opinion and may therefore not represent of my sample voted Tory or Brexit Party the median academic. Unweighted results in 2019. from the YouGov UK 2020 survey show academics’ 2019 and 2017 (recalled) vote.

The Political Composition of UK Academia, Selected Surveys, 2015-19 60 55 51 50 50 46 46 40

30 28 20 20 22 19 12 17 15 18 10 10 8 7 9 6 5 6 3 0 2015ukhls 2015thes 2017yougov20* 2019thes 2019yougov20*

Conservative/UKIP/Brexit Labour Lib Dem Green SNP/Other

Figure 43. Source: Understanding Society 2015, Times Higher Education Supplement 2015 and 2019, YouGov 2020. Note that my survey results are shown with an asterisk.

Using my data, and comparing scoring higher on the psychological traits across the US, UK, and Canada, focusing of agreeableness and trust, who tend to only on currently active social science and lean left, may also be more likely to humanities faculty, produces the picture in complete voluntary surveys.61 Against this, Figure 44. The most representative data those who are interested in filling out a come from Britain, where there is a 6:1 survey concerned with academic freedom ratio of support between left or liberal and may lean more centrist or conservative. conservative parties among SSH faculty. All told, the 12:1 Democrat-to-Republican In the US, the ratio in my survey is 12:1 (D:R) ratio in my data matches Langbert et and in Canada 10:1. In a UK online al’s (2016) representative voter mailout sample of over 200 UK SSH registration-based findings for the top 40 academics, using the same method as the US Economics, Law, Journalism, History North American survey, and achieving a and Psychology departments. It also similar response rate, I found a 12:1 ratio matches his 12:1 D:R ratio for social rather than the 6:1 from the YouGov science academics in 51 leading liberal survey. arts colleges using the same method, The PhD surveys, which are not though it is well below his 32:1 ratio in the convenience samples, show a 9:1 ratio of humanities for these institutions. It also left to right in both Britain and North seems in line with previous such studies American academic figures may overstate and thus a good representation of the the degree of left-liberalism across all US ideology of the faculty of leading North colleges, mainly because they do not American institutions. include low-ranking institutions. Those

SSH Academic Voting in 2016-19 Elections, US, UK, Canada 100 90 83 82 80 72 73 70 60

% 50 40 30 20 14 10 6 6 7 0 US (N=687) UK YouGov UK Mailout Canada (N=246) (N=235) (N=221) Left, Liberal or Green Conservative, Libertarian or Populist Nonvoter

Figure 44.

Finally, the US sample is limited to especially on immigration and attitudes to the top 100 institutions and in Canada to the European Union.63 For Canada, the top 40. Thus the US data are most Nakhaie and Adam (2008) showed that 11- exclusive, followed at a very considerable 13% of Canadian SSH and STEM distance by Canada, with Britain offering academics voted Conservative or Reform an unselective cross-section across all in the 1993-2000 period, with the left ranks of university. Since academics at outnumbering the right 5:1.64 They elite institutions are, from what we know suggested Canadian academics were thus from previous research, more left-leaning, more centrist than their American this may mean that North American counterparts. Our data does not support academics as a whole are somewhat less this contention, though ours is a more left-leaning than my surveys suggest, and elite, SSH-leaning and considerably may not differ as much from their British smaller sample. counterparts as appears to be the case here. Incorporating my data and that Another way to examine ideology from a large-scale survey of European is to ask people to place themselves on a academics and members of the 100,000 5-point or 10-point left-right scale. In a strong UCU, which represents about 80% review of five recent studies, Shields and of Britain’s academics (Mallinson et al. Dunn find the share of conservatives in the 2017), the data in Figure 45 tell a fairly social sciences to be 5-15%, and in the consistent story. The left-right ratio is humanities, 4-8%.62 In the European lowest in continental Europe, at 5:1, with Union, data from the high-quality various Anglosphere surveys picking up European Social Survey, which is a ratios of between 6 and 15 to 1. Right- random stratified rather than convenience leaning academics form around 5% of the sample and obtains a nearly 80% response total in SSH-leaning samples, similar to rate, find professors to be substantially the figures recorded by Langbert et al. more left-wing than other professions, (2016) using voter registration data.

Regardless of differences in modestly greater than in SSH. The biggest sampling technique, the broad trends are gap between SSH and STEM is with similar, pointing to a left-right ratio of respect to the share of far leftists, which is between 6 and 12 to 1 among SSH much lower among STEM faculty. Note academics in North America and Britain. that the British YouGov data include 40% STEM academics are consistently 12-16 retired academics and is half STEM, while points less left-leaning across all datasets, the British mailout survey is limited to though the share of conservatives is only currently active SSH academics.

Ideological Self-Placement, EU, UK, North America

EU (N=3640) 15 74

UK UCU (N=2148) 6 80

UK Mailout (N=221) 5 66

UK YouGov (N=820) 9 53

Canada (N=290) 4 73

US (N=803) 5 73

0 10 20 30 40 50 60 70 80 90 %

Right Center Left

Figure 45. Note: includes both SSH and STEM academics. As with the van der Werfhorst dramatically from those of the wider study on Europe, my data, presented in population. For instance, 52% of the Figure 46, show that academics’ voting public voted to leave the European Union, and attitudes on leading cultural issues compared to only 17% of academics. such as Brexit and immigration differ

Academics and Public Attitudes, Britain

Voted Conservative or Brexit 46 Party in 2019 16 52 Voted to Leave the EU 17

9 Increase Immigration 29

54 Reduce Immigration 16

0 20 40 60 Public Academics

Figure 46. Sources: YouGov Academics survey; National election results; Ipsos Mori survey on immigration, 22 Nov, 2019. Data for the US, in Figure 47, and mailout sample mirrors the US and Canada, in Figure 48, show an even more Canadian online mailout results, showing skewed profile, with the share of just 7% favoring an immigration decrease, American academics favoring limits on eight times lower than the public. We may immigration nearly 9 times lower than that surmise that a representative sample of of the population, and nearly 7 times lower North American academics would likely in Canada. While this is a greater gap than lean more toward the three- or fourfold the threefold distance between British level of difference from the public seen in academics and the public, my UK online the British YouGov sample in Figure 46.

Academics and Public Attitudes (USA)

46 Voted for Trump in 2016 4

34 Increase Immigration 65

34 Reduce Immigration 4

0 20 40 60 80 100

Public Academics

Figure 47. Source: Academic mailout survey; ANES 2019 pilot survey; national election results.

Academics and Public Attitudes (Canada)

36 Voted Conservative or PPC in 2019 7

13 Increase Immigration 52

40 Reduce Immigration 6

0 20 40 60 80 100 Public Academics

Figure 48. Source: Academic mailout survey; EKOS immigration poll 2019; national election results.

Much of the difference on cultural- teachers, though, as Figure 49 shows, populist questions is accounted for by differences are on the order of just ten education level. Around 2 in 3 Britons points more than senior levels of the with a master’s or PhD voted to Remain in teaching profession. Part of the difference the EU, with little more than a quarter is accounted for by the fact that academics opting for Leave.65 A closer examination nearly all have advanced degrees whereas of YouGov’s UK Profiles data, with a many teachers do not. Among advanced sample size of several hundred thousand, degree-holding teachers, differences are shows that degree-holding professors and more modest between academics and lecturers in my sample (most of whose teachers, though still persist. The share of participants can be matched in Profiles) right-leaning academics stands out as are more left-wing than secondary school especially low.

Ideological Self-Placement, by Profession, Education Industry 100% 8% 10% 90% 18% 26% 29% 23% 80% 37% 32% 33% 70% 60% 50% 40% 30% 51% 55% 20% 47% 46% 44% 33% 32% 31% 10% 28% 0%

Left Center Right

Figure 49. Source: YouGov Profiles data, accessed April 19, 2020. unweighted sample size in parentheses. “Don’t knows” counted as centrist. *Denotes data from author’s YouGov survey. A similar difference holds for the shown) are more similar to university EU referendum vote, as Figure 50 shows. professors and lecturers in their Brexit Advanced degree-holding teachers (not vote, and in their views on immigration.

Brexit Vote, by Profession, Education Industry (Britain) 100% 90% 12% 17% 80% 30% 30% 31% 70% 14% 29% 37% 43% 60% 50%

40% 82% 77% 65% 30% 62% 60% 59% 54% 52% 50% 20% 10% 0%

Remain Leave Other

Figure 50. Source: YouGov Profiles data, accessed April 19, 2020. Unweighted sample size in parentheses. “Don’t know” included in ‘other’ category. *Denotes data from author’s YouGov survey. Finally, Figure 51 shows that other sectors all lean toward Remain. degree-holders in the education profession Brexit voting among academics (80% as a whole (universities and schools) lean Remain, 16% Leave) locates them as even modestly more toward Remain compared more Remain-leaning than graduates to other degree-holding professionals. working in media, marketing, PR, and Education is not the most Remain-leaning advertising. profession, however, and degree-holders in

Brexit Vote by Sector, Degree-Holders Only (UK) 100% Remain Leave 90% 80% Other 70% 68% 67% 65% 63% 63% 70% 62% 62% 59% 59% 58% 57% 56% 55% 60% 51% 50% 39% 40% 30% 17% 20% 10% 0%

Figure 51. Source: YouGov Profiles data, accessed 19 April 2020. Universities included under Education sector. Sample size per category is between 382 and 3,949 (with over 10,000 for ‘Other’ and unemployed categories). Only includes university graduates working in these sectors. Data are more difficult to come by Union. In the United States, political for North America or continental Europe, donation data from the FEC find but work by van der Werfhorst (2019) academics rank 7 of 62 professions in their showed that in the European Union, only propensity to donate to Democrats over artists were more left-wing than Republicans. As Figure 52 illustrates, 90% academics. Teachers, while left on the of academic donations went to the economic dimension, placed well to the Democrats, with near-total Democratic right of academics on cultural-nationalist unanimity among professors of English, issues like immigration or the European History, and Law.

Figure 52. Political Donations by Profession in America, 2016. Note: Blue denotes donations to Democrats, red to Republicans. Source: “Democratic vs. Republican Occupations,” Verdantlabs.com, accessed Nov. 15, 2020. Much of academia’s left-liberal humanities. Those with the most advanced tendency is connected to the fact that those degrees are then disproportionately who score higher on the big-5 personality employed in academia. Moreover, the trait of openness tend to self-select into academic sector is, like the school system, higher education.66 Once there, they come largely publicly funded, and thus leans to be influenced by like-minded people, economically left. Finally, occupational especially in the social sciences and typing, akin to the stereotype of bus

drivers as men and nurses as women, may selecting the top 2 public and 2 private add a feedback loop to the process, locking institutions across 30 states, including a in a process that discourages political mix of SSH and STEM departments, they diversity.67 I consider further evidence for show that assistant professors, typically this later in the report. the youngest faculty, have the highest Democrat:Republican ratio (10.5:1) Ideological Age Profile of Academics compared to associate (8.7:1) and full 69 professors, the oldest group (8.2:1). The ideological changes that have My data for the United States is been charted in academia since the 1960s much less of a complete sample than should have left an imprint in today’s Langbert and colleagues’ findings, though academic age structures. In the US, it has the offsetting benefit of including Langbert et al. (2016) looked at five SSH unregistered and non-identifiable staff. fields in the top 40 US universities using Like Langbert et al., I find that younger voter registration data. They reported that SSH academics in Figure 53 are clearly among those under 35, Democrats more left leaning (41% far left) than their outnumbered Republicans 23:1, falling to elders. This is not true for the share of 16:1 among academics aged 36-45, 12:1 right-wing academics, however, which is for those 46-55, 10.5:1 for those 56-65 and higher in the youngest group (8%), 10:1 for those over 65.68 In a second study perhaps indicating greater polarization at younger ages.

Ideology by Age, American SSH Academics 70%

60%

50% 41% 40% 30% 30% 27% 21% 21% 20%

8% 10% 6% 3% 4% 2% 0% under 40 40-49 50-59 60-69 70+ very left fairly left centrist right

Figure 53. N=803. Source: Mailout survey.

In Britain, Figure 54 shows there is of SSH academics, but no clear linear a slightly higher share of far leftists in the pattern, with those over 70 nearly as far youngest cohort (22%) than older cohorts left (18%) as those under 40.

Ideology by Age, UK SSH Academics (YouGov) 60%

50%

40%

30% 22% 20% 16% 18% 14% 10% 10% 9% 10% 7% 6% 3% 0% under 40 40-49 50-59 60-69 70+ very left fairly left centrist right

Figure 54. N=235. Current SSH academics only. Source: YouGov survey. Figure 55 shows the broad age decades, with each new generation more pattern for four main ideological left-wing than the previous one. But this is categories across four academic surveys much less clear in the UK and Canada, as for SSH academics. The American trends well as among PhD students in the US, fit the classic story of younger SSH staff UK, and Canada, where there is no being more leftist than their elders. This consistent pattern. It is therefore difficult looks like a pattern of cohort change that to be sure that there is a general pattern tallies with surveys over the past few across the three countries.

Share of Far Left Academics, by Age, among SSH Faculty 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% under 40 40-49 50-59 60-69 70+ USA Canada UK Yougov UK Mailout

Figure 55. Looking at the other end of the note that the right-wing share consists of scale in Figure 56, there is no evidence very small numbers of academics: below that right-wing academics are going 10% of SSH faculty. All of which makes it extinct. If anything, there is a modest trend difficult to screen noise from signal – of younger SSH staff being slightly more indeed, there is no statistical significance right-leaning than their elders – though to these age trends in my models. this is not a consistent pattern. However,

Share of Right-Wing Academics, by Age, among SSH Faculty 16% 14% 12% 10% 8% 6% 4% 2% 0% under 40 40-49 50-59 60-69 70+ USA Canada UK Yougov UK Mailout . Figure 56.

It could be that the interpretation of is also no clear age trend across studies, as right and left differs between generations. Figures 57 and 58 show. Roughly 50-70% Yet, on the acid test issue of modern across surveys favor an increase and 5- populist conservatism, immigration, there 15% seek a reduction.

Share in Favor of Reducing Immigration, by Age (SSH only) 50%

40%

30%

20%

10%

0% under 40 40-49 50-59 60-69 70+ UK Yougov US Canada UK Mailout N America PhD UK PhD

Figure 57. SSH active academics only.

Share in Favor of Increasing Immigration, by Age (SSH only) 100%

80%

60%

40%

20%

0% under 40 40-49 50-59 60-69 70+ UK Yougov US Canada UK Mailout N America PhD UK PhD

Figure 58. SSH active academics only. Taking a more systematic approach they will be mostly, or wholly, retired by which controls for confounding variables, around 2015-20. I thus cannot comment on Figure 59 demonstrates that age is not a pre-Baby Boom shifts, which appear to significant predictor of left-right ideology have been decisive.70 We are therefore on a 5-point scale except, marginally, in comparing Boomers with Generation X the case of North American PhD students. and Millennials, and here there appears to The effect of being older is pointing in the be no significant ideological difference direction of being less left-wing and more between the generations. right-wing, but does not reach While it is hazardous to read cohort conventional (p < .05) statistical change directly off age effects because significance in any survey. On people may change their views as they age immigration, age is only related to more across the life cycle, the existing literature restrictionist attitudes among North suggests the direction of change would be American and British PhD students, but expected to run in a conservative direction even here in opposing directions, over the life course.71 The absence of confounding clear conclusions. statistically significant age effects in Recall from our earlier discussion Figure 59 work against an explanation of evidence from previous studies that the based on more leftist cohorts replacing leftward shift of the professoriate occurred conservative cohorts within the same in two waves, one between the 1960s and demographic (i.e., younger left-wing men the late 1980s, when the share of replacing older conservative men). The conservatives to liberals declined, and the clear age gradation we see in the US second between the late 80s and the survey, with the young more left-wing, present, when the share of far leftists fails to turn up in the form of significant increased while centrists declined along age effects in all models. with a continued modest decline of What may be more consequential conservatives. To the extent that these is gender, with women significantly more shifts concern the passing of the Silent left-wing among both North American Generation (born before 1949), there will academics and PhDs, and among British be no record of this in my data because academics in the YouGov survey. Even in

the two surveys where women were not only the plot of all respondents showing a more left-wing, the effect was in the sloping pattern of younger people being predicted direction. Nonwhite academics more left than the old (due to the were only (statistically) significantly more compositional effect of more females at left-wing in one of the four surveys where younger ages). It could be that women race was measured, so increasing racial leave the profession in larger numbers than diversity does not seem as likely to have men, offsetting what might be an even altered the ideological composition of the larger shift to the left. SSH faculty as the greater proportion of Finally, being in an SSH discipline women. as opposed to a STEM field is associated Even so, the data don’t tell a clear with being more left-wing in three of the story of growing leftism arising as a four surveys (no STEM were sampled in consequence of a larger share of female the UK mailout). This aligns well with recruits at younger ages. Plots of ideology existing research.72 This said, STEM by age within male and female subsamples respondents were only significantly more do not show, as we might expect from the right-wing than SSH respondents in 1 of 4 gender-driven leftism thesis, flat age lines surveys. By contrast, STEM scholars were (with the young no more left than the old) less likely than STEM scholars to identify for both male and female subsamples; with as far left in 3 of 4 surveys.

Predictors of Left Ideology 0.300 0.277*** 0.250 0.200 0.148*** 0.139*** 0.141** 0.126*** 0.150 0.114 0.094 0.100 0.078* 0.053 0.050 0.000

standardized effect standardized -0.013 -0.050 -0.028 -0.048 -0.100 -0.063 -0.096+ -0.150 N America UK Yougov UK Mailout N American UK PhD PhD Age Female SSH Canada

Figure 59. Note: UK mailout survey contains no STEM respondents. *p<.05;**p<.01;***p<.001. The mystery is also not solved by between 1966 and 1992, some return to the examining changes in the wider political Republicans during the Clinton years, environment. The shift of those with followed by a shift back to the Democrats degrees, especially advanced degrees, to in the post-George W. Bush period. Note the Democrats is more recent. Figure 60 that even in 2016, about 35% of advanced shows there was a modest shift among degree-holders identified as Republican, advanced degree-holders to the Democrats far above the approximately 5% share

among American SSH academics recorded in my survey.

Partisanship among Americans with Advanced Degrees, 1964-2016 70% 60%

50% 57% 40%

30% 36% 20% 10%

1966 1972 1978 1964 1968 1970 1974 1976 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2008 2012 2016 Democrat Republican

Figure 60. Source: ANES cumulative file, 1948-2016. N=3,915.

When it comes to ideology, Figure the mid-1990s, it may be connected to the 61 shows that a statistically significant continued fading of conservatism from trend toward liberalism among American academia after 2004. Even so, the observer advanced degree-holders is mainly a post- is struck by how little of the longitudinal 2012 development linked to Trump. There trend charted by Sam Abrams and others is also a steady decline of conservatism can be tracked in the wider national and modest rise in far leftism between opinion data. More educated Americans 2004 and 2012 that doesn’t reach have been becoming more liberal over the statistical significance. While this cannot last few decades, but the ideological explain the American academic macro transformation of academia has been much trend of declining conservatism prior to more extreme.

Ideology of Americans with Advanced Degrees, 1972-2016 60%

50% 50% 40%

30% 32%

20%

10% 6%

1978 1980 1994 1996 1972 1974 1976 1982 1984 1986 1988 1990 1992 1998 2000 2002 2004 2008 2012 2016 Extreme Liberal Liberal Conservative . Figure 61. Source: ANES cumulative file, 1948-2016. N=3,370.

Having said this, it is also and right due to political events or noteworthy that advanced degree-holders ideological sorting do not therefore have been consistently more Democratic- account for change in the partisan leaning than undergraduate degree-holders composition of the professoriate. The in most years, a statistically significant emergence of more populist Republican phenomenon (see Figure 62). There is an leaders such as George W. Bush and education effect, but it doesn’t seem to Donald Trump cannot explain why explain the chronological shifts to the left Republicans were largely absent from the picked up in the academic data since the SSH academy by 1999. late 80s. Changes in the meaning of left

Republican Vote, Undergraduate vs Advanced Degree- Holders

70% 60% 50% 45% 40% 30% 36% 20% 10%

1966 1974 1982 1990 1998 1964 1968 1970 1972 1976 1978 1980 1984 1986 1988 1992 1994 1996 2000 2002 2004 2008 2012 2016 Undergraduate Advanced

Figure 62. Source: ANES cumulative file, 1948-2016. N=8,097 undergraduate, 3,915 advanced degrees.

Age and Selection Effects Inside the Academy

Something distinctive is occurring 72:15 in 15 years. While their sample may in academia, but if this involves a have exaggerated the extent of the change, narrowing of the intake, we should see this would be in keeping with the thesis more of an age gradient, and we don’t, that the 1960s generation was remaking unlike older studies that did. Lipset and the academy.74 Ladd, for instance, using 1972 data, found My UK YouGov sample consists faculty under 30 to be 35 points more of 40% retired academics with a median liberal than those over 50 and nearly 50 age of 70. While a purely generational points more supportive of black activism.73 theory would presume that pre-Boomer By 2006, however, Gross and Simmons retired academics (born before 1949, thus reported that faculty under age 36 were over age 70) would be more right-wing less likely to be on the left than those aged than Boomer retirees (aged under 70), I 50-64, though the youngest cohort found no significant relationship. contained no more conservatives than the Does this mean that the leftward 1960s Baby Boom generation. Those over shifts charted by Abrams and others for 65 (representing the Silent Generation) the US and Halsey for Britain are an were more conservative than the Boomers artifact of sample bias? Not quite. One though the left still outnumbered the right reason why age data may not align with in this group by over 3:1. The authors cohort shifts over time is selective exit. reasoned that the Baby Boom generation Selection effects are important for had introduced the greatest ideological explaining academia’s ideological change. Indeed, Rothman and his makeup. As early as 1959-61, a study of colleagues, analyzing Carnegie 1984 and Berkeley students found that socialists 1999 survey data, reported that the were four times more likely than left:right ratio had shifted from 39:34 to conservatives to “realistically consider” an

academic career.75 The same may hold for it is younger retirees 65 and under who are leaving the profession. The self-selection more conservative than academics of the of non-leftists out of academia would same age who remain in the profession (in explain why older academics are not more my British YouGov sample of current and conservative than young scholars even as retired academics).76 Conservatives or academia has been shifting left over the centrists are more likely than leftists to generations. Indeed, Figure 63 shows that select out of academia before age 65.

Figure 63. N=820. Pseudo-R2=.02. Under/Over 65 variable x retired interaction is significant at p<.05 level, with controls for age and SSH/STEM. 467 active staff under 66 and 51 active staff over 65. 245 retired staff over 65 and 57 retired staff under 66.

The share of ex-academics may be (mainly working outside academia) are considerably larger because only those considerably more conservative than PhD who listed their profession as university holders working as academics. The professor or lecturer with YouGov were ideological climate may thus operate not counted as such. Those who switch only to discourage conservatives and professions and state a new one are not centrists from entering, but to captured by this data, yet the YouGov disproportionately nudge the conservatives Profiles data on PhDs in Britain visited who start their careers in academia out the earlier shows that older PhD holders door.77

Left-Modernist Hegemony?

It may be the case that the somewhat less so (between 8 and 9 to 1), predominantly left-wing ideological and non-students between 18-24 markedly profile of sociology professors recorded by less so (3.6:1). In Britain, there is a steady Lipset and Ladd in 1972 has come to education gradient, with each higher level spread across the entire academy and even of education more left-leaning than the down to an entire generation of students. previous level. The general pattern is one Figure 64 summarizes Prolific in which further education correlates with demographics for various student greater leftism. The Prolific data are categories, drawing on a 100% sample of heavily skewed to the left compared to the those taking surveys on the site. US data more representative ANES data we visited show that PhD students on the platform earlier, where 35% of American advanced are the most left-leaning (10.8:1), with degree-holders voted Republican in 2016, master’s and undergraduate students with no significant variation by age.78

Left:Right Ratio by Student Status (Prolific) 15.6 16.0 14.0 12.1 12.0 10.8 10.0 9.2 8.8 7.9 8.0 ratio 5.7 6.0 3.6 4.0 2.0 0.0 PhD student Masters student Undergraduate 18-24 Non- Student USA UK

Figure 64. Note: US Ns=202 PhD, 584 master’s, 2200 Undergraduate, 79 Non-student; UK Ns=212 PhD, 556 master’s, 2219 Undergraduate, 104 Non-student. Only those who had attained prerequisite degree qualification are included.

The best benchmark is the UK step-change between master’s and PhD YouGov Profiles data, which is more study. This in turn compares to a 1:1 ratio representative of the wider population. in the population at large. On immigration Figure 65 shows that in Britain, there is a attitudes, young non-students back 2:1 left-right ratio among 18-24-year-olds restriction over increasing immigration 2.6 who are not in education, and among to 1, just below the population average. undergraduate and master’s students alike. This slips to around 1.5:1 among This jumps to nearly 5:1 among PhD undergraduate and master’s students, and students, however, indicating an important 1:1 among PhD students. Here PhD

students stand out on the left:right similar gap shows up in the data on measure, though immigration attitudes academics. move along more of a steady gradient. We also know that there is an These data speak to powerful important political gap between advanced political selection effects as students degree-holders who work in academia and ascend the academic ladder. Do those with those who work outside it, which widens inquiring minds who pursue knowledge with age. This points to the content and score higher in the Big 5 personality trait ethos of SSH academia as an important of openness to new experience and thereby factor. How much of this is legitimate tend to embrace left-modernism? My (social problems naturally attract the left- Prolific PhD surveys show that STEM inclined) or illegitimate (universities and PhDs are 16 points less likely to be on the especially SSH departments are hostile left than SSH PhDs in North America and environments for conservatives, and thus 13 points less likely to be so in the UK. A fail to ask important social questions due to bias)? This is the question I turn to next

Left:Right Ratio by Education Status (UK YouGov) 6.0

5.0 4.8

4.0

3.0 2.6

ratio 2.4 2.2 2.1 2.0 1.6 1.4 1.0 1.0

0.0 18-24 Non UG student (2254) Msc Student (442) PhD student (549) Student (3693) Student Status

Left:Right Lower:Higher Immigration

Figure 65. Source: YouGov Profiles, accessed Dec. 15, 2020. Number of cases in parentheses. Research into why conservatives in the idea of pursuing graduate work select out of academia by Woessner and while conservative interest remains static. Kelly-Woessner (2009) finds that Conservatives also report substantially conservative undergraduates are only half more dissatisfaction with their SSH as likely to major in SSH subjects as left- elective classes than their STEM electives, wing students. Conservatives tend to in contrast to left-wing students. In prioritize practical over expressive goals political science classes, the authors note, more than leftists. Even so, between their “we found clear evidence that students freshman and senior years, only left- who believed they were at odds with their leaning students become more interested professor’s politics were generally more

critical of the professor, the course, and interest in an academic career. This was the subject matter.” Conservatives are also followed by, “How much do the following less likely to report being mentored by considerations affect your decision?” and a (mainly liberal) academic staff. This battery of six questions, each on a 7-point suggests that while self-selection is the scale from strongly agree to strongly strongest reason for academia’s skewed disagree: political profile, the academic environment • “Academic jobs are too hard to get, is probably a contributing factor.79 especially where you want them.” The authors mention the possibility that a hostile environment may be shaping • “I can earn more money in a preferences, though their data do not ask different job.” this directly. It is also worth inquiring into • “Academia is too isolating, I prefer the extent to which selection pressures more social interaction at work” operate further down the academic • “There is too much form-filling pipeline at the master’s level. To examine and hoop-jumping in academia” these questions further, in a December • “My political views wouldn’t fit, 2020 Prolific survey of 843 master’s and which could make my life PhD students, I asked: difficult.” • “I’m not the academic type.” How interested are you in pursuing a career as a university academic The first point to note in Figure 66 is (i.e. Lecturer, Professor)? how clearly ideology structures the answers to the “my political views Answers were provided on a 5- wouldn’t fit” question, with conservatives point scale from “not at all interested” to much more likely to believe their views “extremely interested.” For analysis, I could “make my life difficult” if they collapsed the “very” and “extremely” pursued an academic career: answers into one category to signify strong

My Political Views Wouldn't Fit Academia, Graduate Students, by Ideology (STEM & SSH)

70% 62% 60%

50%

40% 34% 30%

20% 18%

10% 8% 6%

0% Very Left Somewhat Left Centrist Somewhat Very Right Right

Figure 66. N= 843, includes master’s and PhD students, STEM and SSH, 434 UK, 368 USA, 41 Canada. 14% right-wing responses due to oversample of right-wing master’s students. The argument that conservatives position where they want one. The latter are repelled by academia’s low earning may indicate that conservatives are more potential, at least among those who made oriented toward geographic stability than it to graduate school, finds no support in liberals. This may also point to Woessner the data. Figure 67 presents ordered and Kelly-Woessner’s observation that logistic regression coefficients for models conservatives prioritize starting a family predicting responses to all six questions, more than liberals.80 In any event, within with controls for age, gender, income, the self-selected group of graduate SSH v. STEM, master’s or PhD, and students taking the survey, the prospect of country. The chart shows that left- and being able to earn more money elsewhere right-leaning graduate students only differ does not seem to deter conservatives more in their answers on two of the six than liberals, echoing findings in previous academic career considerations. Namely work.81 Moreover, those on the left are, if political views, which is by far the anything, somewhat more likely to say strongest effect, and in their view that it they are “not the academic type.” would be too difficult to find an academic

Correlation between Left-Right Ideology and Academic Career Considerations

0.511*** Political views wouldn't fit -0.224*** Academic jobs too hard to get where you want them

-0.068 Not the academic type

0.067 I can earn more elsewhere

-0.049 Academia too isolating

-0.029 Too many forms and requirements

-0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 unstandardized effect

Figure 67. ***p<.001. N= 843, includes master’s and PhD students, STEM and SSH, 434 UK, 368 USA, 41 Canada. 14% right-wing responses due to oversample of right-wing master’s students. Pseudo-R2s on models range between .008 and .023.

The foregoing doesn’t mean that not fitting in even seems mildly correlated the political atmosphere in academia is a with interest in an academic career. Being deal-breaker for all conservatives. “not the academic type” and saying that According to Figure 68, among graduate academia is too isolating are the stronger students, concerns about political fit rank predictors of who is interested in being a fourth, and then only for right-leaning professor. graduate students. For those on the left,

Predictors of Interest in Academic Career -0.34*** Not the academic type PhD student 0.18*** Isolating -0.15*** Political views not fit x Right-wing -0.11* Earn elsewhere -0.10** Political views not a fit 0.08* SSH 0.08* Right-wing 0.06 Too many forms -0.06 Income -0.02 Too hard to get 0.01

-0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 standardized effect (beta)

Figure 68. *p<.05; **p<.01;***p<.001. N= 843. R2=.278. Includes master’s and PhD students.

An interesting dynamic concerns academic career than are left-wing or the difference between master’s and PhD centrist counterparts. SSH students. Right-wing SSH master’s Sensitivity to a hostile political students who sense they won’t fit atmosphere thereby acts to deter politically are significantly less likely than conservatives from pursuing an academic leftist and centrist SSH master’s students career when they are studying at the who say they won’t fit politically to be master’s level. The data show that interested in an academic career. Figure 69 graduate students who say they are “very shows that just 9% of right-leaning SSH right” are about 30 points more likely to master’s students who say their views say their politics doesn’t fit academia than would make a poor fit with academia have those who are “fairly right.” We might a strong interest in pursuing an academic also surmise that the conservative master’s career compared to 29% of leftist or students who say their views would make centrist SSH master’s students who say a difficult fit with academia are more their politics would fit poorly. culturally conservative whereas those who On the other hand, 42% of right- say fit is not a problem may be more leaning SSH master’s students who say economically conservative. On the left, the that political compatibility is not a meaning of saying one’s politics don’t fit problem say they are interested in an may be different, perhaps reflecting a academic career. While there is just a 4- more general feeling of nonconformity to point difference between leftist/centrist institutional dictates. master’s students who think their politics Importantly, the few conservatives is, or is not, a fit with academia, the gap who commit to studying for the PhD are reaches 33 points on the right. And, as we less likely to be put off by academia’s have seen, right-wing graduate students political atmosphere. Conservatives are already much less likely to say their unwilling to mute their views or put up views would make a good fit with an with hostility from the left have arguably

selected out by this stage. Importantly, the Second, right-wing PhD students share of “very right” compared to “fairly who see their politics as a difficult fit with right” graduate students does not vary academia are more interested in pursuing from master’s to PhD level, or between an academic career than their politically- those with an interest in academia and incompatible master’s counterparts. those who have no interest. This holds out Whereas there is a 32-point gap among hope that a small remnant of conservatives right-wing master’s students between will remain in SSH academia into the those who view their politics as fitting foreseeable future. versus not fitting an academic career, What, then, accounts for self- right-wing SSH PhD students who see selection effects being stronger at master’s their politics as incompatible are actually than PhD level? To begin with, the share 10 points more likely to express interest in of right-wing SSH PhD students who say an academic career – though this is not their politics are incompatible with statistically significant as there are only 20 academia is 33%, lower than for SSH right-leaning SSH PhD students in the master’s students on the right, where 49% sample. say their politics are incompatible with academia.82

Interest in Academic Career, by Ideology and Political Compatibility, SSH Graduate Students 60% 50% 50% 44% 38%* 40% 40%

30% 27% 20% 18%

% interested % 20%

10% 6%*

0% Left/Centre Left/Centre Right Right Compatible Incompatible Compatible Incompatible SSH Masters SSH PhD

Figure 69. N=361, including 49 right-wing SSH graduate students (29 master’s, 20 PhD). *Chi-squared test on political compatibility and political interest crosstabulation is significant at p<.05 level among 29 right-wing SSH master’s students. No other relationships significant.

Running the master’s portion of the questions, results in Figure 70. This above as a statistical model, with strong illustrates that it is among right-wing interest in becoming an academic as the master’s students that perceived political outcome measure, and controlling for age, hostility makes an impact. Resilience, an gender, income, country, university rank, ability to conceal one’s views, or a greater and the five other academic career comfort navigating the rules of academia

probably explain why right-wing PhD center who see themselves as politically students are better adapted than right-wing incompatible, whether in STEM or SSH, master’s students for academia. The are actually more likely to express an effects are driven by the tendency of right- interest in academia, suggesting that the wing master’s students in SSH fields to meaning of political incompatibility may view themselves as politically vary by ideology. incompatible. Those on the left or in the

Figure 70. Includes SSH and STEM. N=370, Pseudo-R2=.131. Right x Political Fit interaction significant at p<.01 level. Controls for age, gender, income, country, university rank, SSH/STEM and the five other academic career questions.

Finally, it should be said that Notice as well that the transition from political considerations are not a master’s to PhD seems to coincide with significant predictor of interest in an reduced interest among STEM students academic career among STEM graduate and increased interest among SSH students at any level. Figure 71 shows that students, which likely has to do with more political compatibility is only consistently plentiful career opportunities for STEM related to interest in an academic career PhDs outside academia. Social isolation among SSH graduate students – and even rather than loss of earnings is the dominant here, it is significantly correlated only factor in deterring graduate students among right-wing SSH master’s students. overall from pursuing an academic career.

Share of Graduate Students Interested in Academic Career 50% 45% 44% 42% 38% 40%

30% 25% 23% 20% 20% 12% 10% 0% SSH Politically SSH Politically STEM Politically STEM Politically Incompatible Compatible Incompatible Compatible Master's PhD

Figure 71. N=843. These findings largely comport of what a leading academic should look with those of previous researchers. As like.83 Gross and Fosse summarize, “Woessner and Kelly-Woessner (2009) find that twice Section IIb: Chilling Effects and Self- as many liberal as conservative college students aspire to complete a doctorate.” Censorship Other work (Gross and Simmons 2006) shows that being conservative, A Hostile Climate for Conservatives in Republican, or an evangelical is associated Academia? with lower confidence in universities and viewing professor as a lower-prestige occupation. The authors themselves argue One reason for conservatives to that academia is “typed” as a liberal stay away from, or leave, academia is a occupation the way the military is typed as perceived hostile work environment. conservative, and that this occurred in part Academia may lean left, but is this due to a growing post-1960s identification, infringing on the liberty of academics to in the minds of many, of academia with teach and research? Not if the composition the left. of the academy is the result of self- Though the authors are liberals selection in an atmosphere supportive of who stress the importance of self- all points of view. If, however, it results selection, they also allow that from deliberately exclusive or hostile conservatives may face hostility for their actions from institutions or academics, views. Moreover, they acknowledge that then this does represent a constraint on entire disciplines and elite universities academic freedom and the mission of construct identities (“images of intellectual universities to pursue knowledge. personhood”) that include being left-wing. Before proceeding, it is important Thus, they continue, hiring committees at to address a popular theory on the right: elite universities may be biased against that universities are indoctrinating students conservative applicants because right-wing into left-wing beliefs. Repeated studies, candidates don’t fit their idealized image including my previous co-authored report Academic Freedom in the UK show little

to no effect of classroom exposure on 9% right. For currently employed students’ political views.84 Thus, rather academics in the social sciences and than impacting students’ opinions, the humanities, the figures were 61% left, more pressing effect of political 30% centrist, and 7% right. discrimination – if this is indeed taking What is especially noteworthy is place – is on the academic experience of the significantly higher presence of those staff and students, the quality of research, describing themselves as “far left” and as and the principles of expressive freedom “activists” within social science and and the pursuit of truth. humanities disciplines as compared to My first aim therefore is to take the STEM fields. Indeed, among currently temperature of the academic workplace, a serving academics, 16% of SSH academics climate predominantly established by are on the far left compared to 6% for colleagues, students, and university STEM. Twenty-six percent of current SSH administrators. How strong is the pressure academics consider themselves activists to conform? Are there fears that compared to 15% of STEM staff. These nonconformity will be punished? This findings mirror those from US scholars perception, if widespread, would restrict who similarly found a big difference in the expressive freedom of a large number self-identified activists, radicals, and of faculty, and present a far greater Marxists between SSH and STEM.85 The challenge to academic freedom than no- proportion of centrists in SSH fields, at platforming. 32%, is also markedly lower than the 43% I began by asking about subjects’ for STEM subjects.86 Gross and Simmons own political ideology on a 5-point left- also found that younger professors were right scale, then asked where they thought less Marxist, radical, and activist. In my the average member of their department data, academics under 40 are indeed less was on the same scale. UK YouGov likely to be activists (16%) than those over survey results are summarized in Figure 65 (27%) but the under-40s are 5 points 72. In terms of respondents themselves, more likely to be far left (15.6%) than 53% identified as left, 35% as centrist, and those over 65 (10.5%).

Ideological Self-Placement of Academics, by Field (Britain) 46 50 42 41 40 28 31 30

% 18 20 15 7 10 7 10 1 1 0 Consider Very left Fairly left Centrist Fairly Very myself right right an activist STEM SSH

Figure 72. Note: Includes both serving and retired academics in my YouGov sample. Figure 73 shows an even larger gap and 35% for Canadian SSH, academics. between STEM and SSH subjects in the The UK online mailout results also show United States than in Britain. The British 22% “very left” compared to 15% for the online mailout SSH sample shows 30% YouGov UK sample, 26% of US SSH activists, similar to the 28% for the sample, and 23% for the Canadian SSH YouGov sample, and well below the 38% sample. activist component recorded for US SSH,

Ideological Self-Placement by Field (USA) 48 50 44 38 40 33 30 28

% 20 20 18 12 9 10 4 1 1 0 Activist Very Fairly Centrist Fairly Very Left Left Right Right STEM SSH

Figure 73.

As Figure 74 shows, SSH men, though the gender gap is twice as academics in Britain are significantly more wide in STEM, whereas SSH fields tend to left-wing than STEM academics. UK be uniformly left-leaning across gender. women are also more left-leaning than

100

Figure 74. Note: UK YouGov data. Pseudo-R2=.021. Controls for age, academic rank and minority. Female and SSH are significant, but interaction of female and SSH is not significant at p<.05 level.

The data for the US (as for to be more left-wing, and women are again Canada), presented in Figure 75, also find more left-wing than men – as they are in social science and humanities academics the wider population.

101

Figure 75. Pseudo-R2=.030. Controls for age, academic rank and minority. Controls for age and minority. Female and SSH are significant, but interaction of female and SSH is not significant at p<.05 level.

British academics’ perceptions of in their department. Though the two their departments reflect the reality that questions are not directly comparable, the SSH fields are more strongly left leaning. answers suggest that most academics have In Figures 76 and 77, I compare self- a fairly good sense of the climate of identified ideology with academics’ opinion in their workplace. perceptions of the average political leaning

102

% Share by Actual Ideology and Perceived Mean Ideology in Own Department (Britain) - STEM Subjects 50 47 43 40 35 34 30

20 10 10 10 8 5 4 1 0 3 0 Very left Fairly left Centrist Fairly right Very right Don't know Actual Perceived

Figure 76. Note: chart compares the percentage of respondents in each ideological category of self-identity (“actual”) with the ideological variation in what respondents think is the median in their department (“perceived”).

This said, while British SSH contrasts with STEM staff. While this academics are more left-wing than their could be an accident of my sample, it may STEM counterparts, Figure 77 shows that also be due to the clearer cues about British SSH staff view their departments political values provided by the content of as 8 points more left-wing than teaching and research in SSH fields. themselves. The degree of this perception

103

% Share by Actual Ideology and Perceived Mean Ideology in Own Department (Britain) - SSH Subjects

70 60 60 50 45 40 30 30 21 20 16 8 10 6 4 6 1 0 3 0 Very left Fairly left Centrist Fairly right Very right Don't know

Actual Perceived

Figure 77. Note: chart compares the percentage of respondents in each ideological category of self-identity (“actual”) with the ideological variation in what respondents think is the median in their department (“perceived”).

In particular, right-leaning SSH and 4 SSH) identify as such. On the other academics tend to perceive a greater left hand, leftist and centrist SSH staff view skew in their departments compared to their departments as only slightly more both left-leaning SSH academics and those left-wing than STEM staff. This shows in STEM subjects. Figure 78, for instance, that right-leaning staff in SSH subjects shows that “fairly right” British SSH may perceive their work environments as academics see their departments as half a more politically distant than they really scale point more left-wing than their are, as compared to their conservative “fairly right” STEM counterparts. Among STEM counterparts. The combination of “very right” SSH academics in Britain, the SSH being more left-wing in reality, and it gap is a full point, which means that “very being perceived to be even more so than it right” SSH academics view their really is (arguably due to the more departments as “very left” while “very transparently political content of these right” STEM staff see theirs as “fairly fields), creates a marked sense of left.” Results for “very right” academics dissonance among right-leaning social should be taken with caution, however, as science and humanities academics. only 8 individuals in the data (4 STEM

104

Figure 78. Note: UK YouGov data. Predicted ideology of department, with controls for age, gender, rank, and whether someone is retired or active. Ideology-field interaction is significant. Pseudo-R2=.045.

The situation in the United States survey respondents compared to their own and Canada is relatively similar to Britain, makeup. In combination with the with perceived and sample average considerably higher perceived far-left (“actual”) ideological composition fairly share in America (26% compared to 7% in closely aligned. However, Figure 79 Britain), this indicates that academia in the illustrates that while the size of the far left US – at least in the top 100 departments – is underestimated relative to the sample in is likely to be more left-leaning than in Britain, it is overestimated by American Britain.

105

% Share by Actual Respondents Ideology and Perceived Mean Ideology in Own Department (USA) 70 60 60 50 47 40 30 26 20 21 20 16

10 3 4 0 1 0 Very Left Fairly Left Centrist Fairly Right Very Right Actual Perceived

Figure 79. Note: chart compares the percentage of respondents in each ideological category of self-identity in the sample (“actual”) with the ideological variation in what respondents think is the median in their department (“perceived”).

Comparing SSH and STEM left” and “fairly left,” whereas left-leaning disciplines, the pattern in the US SSH academics see their departments as resembles Britain insofar as more right- “fairly left.” This said, there is little wing American SSH academics perceive difference in the ideological skew in their departments as being somewhat more perception between STEM and SSH left-wing than leftist SSH academics do. academics in the United States, as the two Thus the most right-leaning SSH lines in Figure 80 run parallel with each academics in America view their other. departments as somewhere between “very

106

Figure 80. Predicted ideology of department, with controls for age, gender, and race. Ideology-field interaction is not significant. Pseudo-R2=.051.

In Canada, the data look more climate towards people with your political similar to Britain, with a little over 60% of beliefs in your department.” Figure 81 respondents perceiving their department’s shows that 44% of “fairly right” and 63% median academic as being on the left. of “very right” British academics in our YouGov sample perceive a hostile climate Chilling Effects in their department. Notice as well that “fairly left” academics are most In the absence of strong collegial comfortable, with “very left” academics norms against political discrimination, an feeling slightly more hostility, though this ideological skew translates fairly readily is not statistically significant. The UK into a hostile climate for conservative mailout survey shows, similarly, that 17% scholars. The combination of leaning right of far-left, 5% of center-left, 26% of and being in a social science or humanities centrist, and 60% of right-wing academics department, for instance, conditions an report a hostile climate. For UK PhDs, the academic’s answer to the question of numbers are 16-4-9-71, a similar profile, whether “there is a supportive or hostile albeit with an even stronger sense of exclusion among those on the right.

107

A Hostile Climate for Your Beliefs in Dept? (Britain)

70% 63% 60% 50% 44% 40% 30% 20% 16% 15% 8% 10% 0% Very Left Fairly Left Centrist Fairly Right Very Right Own Ideology

Figure 81. N=820. Source: YouGov 2020 survey.

In my American sample, centrists centrist American academics report a and those on the right perceive a hostile climate dovetails with earlier data considerably higher level of hostility for suggesting US universities may tilt further their beliefs than is the case in Britain (see left than in Britain. Figure 82). The fact that nearly 4 in 10

A Hostile Climate for Your Political Beliefs? (USA) 80% 69% 71% 70% 60% 50% 40% 36% 30% 20%

10% 5% 3% 0% Very Left Fairly Left Centrist Fairly Right Very Right Own Ideology

Figure 82. N=803. Source: US Mailout survey.

However, our US sample is British one, in that it is drawn from top considerably more exclusive than the 100 schools and is more SSH than STEM

108

in composition compared to the UK Generally, both centrists and “very left” YouGov sample (but not the British online respondents report higher hostility than the mailout sample). Canadian data are also “fairly left”, who appear to be the modal heavily SSH, but come from a relatively category in academia. One sample that wider selection of universities. Comparing was not included in Figure 83 is a survey “apples to apples” across the three of 227 of the members of the National countries, restricting to SSH disciplines, Association for Scholars, a pro-free speech and also including two high-response rate academic association. NAS members in Prolific samples of PhD students from the sample break 10% left, 24% centrist Britain and North America, reveals the and 64% right-wing, with 58% voting for pattern in Figure 83. Donald Trump in 2016. Sixty-seven Here the responses look relatively percent of the NAS sample reported a uniform across the board, regardless of hostile environment, including 72% of country, with between 60 and 80% of those on the right – mirroring the results in right-wing academics and PhDs reporting Figure 83. a hostile climate for their beliefs.

A Hostile Climate for your Beliefs? (US-UK-Canada)

90% 80% 80% 75% 80% 70% 73% 70% 60% 60% 50% 40% 30% 20% 10% 0% US SSH Canada SSH UK Opt-in UK YouGov UK SSH N America (706) (260) SSH (221) SSH (235) PhDs (77) SSH PhDs (124) Very Left Fairly Left Centrist Right

Figure 83. Note: sample size in parentheses. In a statistical model of UK among SSH than STEM staff. Left YouGov data in Figure 84, controlling for academics perceive their departmental age, gender, rank, income, and whether political climates as supportive, while someone is retired or active, I find a academics on the right, especially SSH sharper ideological hostility gradient staff, experience a hostile climate.

109

Figure 84. Controls for age, gender, rank, income, and whether someone is retired or active. R2=.195. N=820.

Focusing only on UK academics disciplines experience this. Among STEM who perceive a “somewhat hostile” or academics, the share of right-leaning staff “hostile” climate for people with their who feel hostility is lower, at between a political beliefs, I find that a quarter of quarter and a half. Once again, Figure 85 centrists, half of “fairly right” academics, shows a pattern of greater alienation and all “very right” academics in SSH among right-leaning staff in SSH fields.

110

Figure 85. Controls for age, gender, rank, income, and whether someone is retired or active. Pseudo R2=.138. N=820.

A somewhat similar story holds and those on the right, fewer than half do. with respect to identifying with the culture In addition, among right-leaning of one’s department. Here I collapse academics, the share identifying against “fairly” and “very” right-wing academics the culture of their department (42%) is into one category because of small nearly twice as large as that identifying numbers. As Figure 86 shows, a majority with it (22%). There is little difference in of those on the left identify with their perception between the “fairly” and “very” departmental cultures, but among centrists left-wing professoriate.

111

Identify with Departmental Culture (British academics) 70% 59% 60% 54% 50% 42% 42% 40%

30% 20% 22% 18% 20% 15% 10%

0% very left fairly left centrist right

Identify With Neutral Identify Against

Figure 86. N=820.

American data in Figure 87 show a atmosphere. Again, however, we must sharper ideological gradient than in bear in mind that the US sample is more Britain, with 61% of right-leaning and elite and more SSH in composition than is 33% of centrist academics identifying true of the British data. Finally, data from against the culture of their department. the National Association of Scholars Identification with one’s department is sample show that 71% of the 150 right- also higher among left-wing US academics wing NAS academics in the survey than their British leftist counterparts, identified against their department, hinting at a more closed departmental comporting with other results.

112

Identification with Culture of Department (US Academics) 80% 70% 66% 67% 61% 60% 50% 40% 40% 33% 30% 18% 19% 20% 17% 10% 0% Very Left Fairly Left Centrist Right

With Neutral Against

Figure 87. N=803.

With controls for status and their departmental culture than do their demographic factors in Figure 88, we see peers in the STEM professoriate. Again, that the main difference between British we see a stronger ideological gradient in STEM and SSH academics is on the left of the social sciences and humanities than in the spectrum. Here, the SSH professoriate STEM fields. identify significantly more strongly with

113

Figure 88. Controls for age, gender, rank, and whether someone is retired or active. Income not included as leads to listwise deletion. Ordered logit pseudo R2=.033.

Running a slightly different version identifying with one’s department, with of the model for North America in Figure the effect more pronounced in the social 89 shows that ideology is again a sciences and humanities than STEM. statistically significant predictor of

114

Figure 89. Controls for age, gender, minority, and US or Canada. Pseudo-R2=.102. Older, SSH, right, and SSH x right interactions are significant at .1 level. With no interaction, centrist and right are significant at .001 level.

Among the 77 SSH PhD students authored report on academic freedom I in my British sample, 60% of those found that fewer than 4 in 10 Leave- identifying as right-wing, but just 17% of supporting students would feel centrists and 7% of leftists, said they comfortable expressing their Brexit view identified against their department’s in class compared to nearly 9 in 10 culture. Among the 124 North American Remain-supporting students.87 The results SSH PhDs, the equivalent split is 55-31-7. in Figure 90 show a similar disparity. These numbers are fairly similar, but While 85% of all academics in the indicate a slightly higher share of US and sample said Remain voters would feel Canadian centrists identify against their comfortable expressing their Brexit view departmental culture compared to Britain. to colleagues, just 37% of Leave voting academics said Leave supporters would Expressing Political Views to Colleagues feel comfortable doing so. Narrowing the focus to exclude STEM and retired In the overall British YouGov academics, I find that just 18% of current sample, 78% of academics voted Remain Leave-supporting SSH academics say a and 16% Leave. In my previous co- Leaver would feel comfortable sharing his views with colleagues.

115

Would someone be comfortable expressing their Brexit view? (Britain) 100% 85% 80% 80%

60% 44% 37% 40%

20%

0% Remainer/Nonvoter Leaver

Expressing Leave View to Colleagues Expressing Remain View to Colleagues

Figure 90. Note: UK YouGov data, N=820. Includes SSH and STEM, current and retired academics. Note that “don’t know” values are higher for Remainers when thinking about Leavers, and vice-versa, which affects the results somewhat. Includes retired staff.

The US sample tells an even more would feel comfortable expressing their extreme story with respect to Trump views to a fellow lecturer or professor. All supporters in the 2016 election. In Figure told, this is a staggering disparity, and, 91, just 14% of Clinton voters and a mere whatever Trump’s flaws, offers powerful 3% (1 individual) of the 33 Trump-voting evidence for the operation of Mill’s academics in the sample agree that a “despotism of custom” and the absence of Trump supporter would feel comfortable what Lukianoff terms a “free speech expressing this view to a colleague. In culture” among the professoriate.88 This contrast, over 90% say a Biden supporter limits the range of potential interchange.

116

Would Trump/Biden supporter be comfortable expressing their view? (by own 2016 vote) 97 100 91 90 80 70 60

% 50 40 30 20 14

10 3 0 Voted Clinton Voted Trump

Expressing pro-Trump View to Colleagues Expressing pro-Biden View to Colleagues

Figure 91. Source: US mailout survey, N=683 (33 Trump, 650 Clinton). If we just focus on current Trump- survey methods. Among the 126 Trump- or Leave-supporting academics and PhD voting academics in the NAS academics students in the social sciences and survey, 11% said Trump supporters would humanities in Figure 92, we see a be comfortable expressing their views, consistent pattern of silence, with just 0- 82% said they wouldn’t be, and 7% were 33% of those polled saying that someone unsure. This compares with 87% of NAS on their side would feel comfortable Trump voters who said a Sanders expressing their view to a colleague. By supporter would be comfortable doing so. contrast 83-100% felt that Remain or The portrait this paints is of a profound Biden supporters could express their views chilling effect and the shutting down of openly. Though there are few important opportunities for the exchange Trump/Leave supporters in the surveys of perspectives and ideas across central (sample sizes range from 6 to 26), the data societal divides. tell the same story across countries and

117

Would Trump or Biden supporter be comfortable expressing their view to a colleague? (Leave/Trump voters only) 100% 100% 100% 96% 89% 87%*Sanders 90% 83% 80% 70% 60% 50% 40% 33% 30% 18% 21% 20% 11% 4% 10% 0% 0% UK YouGov UK Opt-in UK Leave SSH US Trump US Trump US Trump SSH Leave Leave SSH PhDs (6) SSH SSH PhDs (6) NAS (127) (22) (19) Academics (26)

Comfortable Expressing Brexit/Trump Comfortable Expressing Remain/Biden

Figure 92. *Note: For NAS sample, question asked whether a Sanders supporter would be comfortable expressing their views, rather than a Biden supporter. Sample size in parentheses.

There is important variation opinion to colleagues. And among Remain according to whether someone is retired supporters or non-voters, just 31% of SSH and whether they are in STEM or the staff think a Leaver would feel social sciences and humanities. Figure 93 comfortable expressing their view, shows how comfort among British suggesting this is not just the perception of academics varies between SSH and STEM Leavers. This paints a stark picture, in staff, and by Brexit vote, controlling for which Leave supporters feel they need to academic rank and demographics. Among keep their views to themselves. This currently employed social science and chilling effect is especially pronounced in humanities faculty, just 18% of Leave the social sciences and humanities, supporters indicate that a Leaver would precisely the places where an open feel comfortable expressing their view. exchange of social and political views is Widening this to include retired SSH vital to the mission of research and faculty, 28% of Leavers say a Leaver teaching. would feel comfortable expressing this

118

Figure 93. Controls for age, gender, and rank (professor or lecturer). Pseudo R2=.04; N = 484.

A more serious level of self- Before delving into the substance censorship is for academics to actively of the qualitative feedback I received, I silence themselves in their research or coded responses as 1, where people had teaching. Thus I asked respondents, “Have refrained from airing views for reasons of you refrained from airing views in politics or ideology, or 0, where a negative teaching or academic discussions, or response was received or complaints had avoided publishing research, because of to do with non-political content. Results possible consequences to your career from for the UK YouGov academic sample in doing so? Please describe your Figure 94 show that self-censorship is over experiences (if any), and how frequently twice as high among right-leaning you have faced them.” academics as others.

119

Have you refrained from airing views in teaching and research? (Britain) 50% 45% 40% 35% 32% 30% 25% 20% 15% 15% 15% 10% 10% 5% 0% very left fairly left centrist right

Figure 94. N=820.

Among two groups of academics, STEM faculty, though the disciplinary namely right-wing academics who effect doesn’t widen among right-leaners. currently teach, and those in the social The model confirms that, even controlling sciences and humanities, over 40% said for status and demographic factors, right- they self-censored. When the overlap of leaning academics self-censor at much these two groups is considered, namely higher rates than leftists or centrists. There right-leaning staff currently teaching in the is slightly higher self-censorship among social sciences and humanities, the share very left compared to fairly left faculty, who have self-censored reaches 50%. but this is not statistically significant. The model in Figure 95 shows that SSH academics self-censor more than

120

Figure 95. Controls for age, gender, and rank. Pseudo R2=.07; N = 484. In North America, self-censorship American social science and humanities appears to be greater than in Britain. academics, and 42% of centrists, say they Figure 96 shows that 70% of right-leaning self-censor in their teaching or research.

121

Have you refrained from airing views in teaching and research? (US SSH) 80% 70% 70%

60%

50% 42% 40%

30% 26% 26%

20%

10%

0% Very Left Fairly Left Centrist Right

Figure 96. Note: N=706.

The pattern in Canada appears higher than for STEM when controlling closer to the United States than Britain, for age, race, and gender. As in Britain, with 47% of centrists and 56% of right- SSH academics self-censor significantly leaning academics self-censoring. Figure more than STEM academics. 97 shows that the red line for SSH is again

122

Figure 97. Controls for age, gender, country, and race. Ideology significant at .001 level and SSH significant at the .01 level. Pseudo R2=.05; N = 1093. A full comparative chart, taking in little variation by ideology. An important all SSH datasets, appears in Figure 98. group of NAS scholars on the right said Though there is variation, the broad they spoke out despite considerable pattern shows a substantially higher level blowback, suggesting that legal-moral of self-censorship on the right, with some support groups like the NAS may have a also taking place among centrists and far constructive role to play in upholding leftists. Centrists in North America seem academic freedom, or that they select for to have an elevated fear compared to their more resilient free speech activists. There UK centrist counterparts. One exception is was no clear age pattern for self- the National Association of Scholars censorship, which occurs among those of survey, where around 60% of 226 all age brackets. respondents said they self-censored, with

123

Self-Censorship of Views in Teaching and Research, SSH only, by Sample

70%

56% 50% 47% 45% 42% 40%

31% 26%27% 23% 23% 19% 21% 20% 16% 17% 12% 12% 13%

VERY LEFT FAIRLY LEFT CENTRIST RIGHT

UK SSH YouGov (229) US SSH (706) Canada SSH (260) N Am SSH PhD (124) UK Mailout SSH PhD (77) UK Mailout SSH (216)

Figure 98. Note: number of observations in parentheses.

my views to myself or make grunting British Testimonials noncommittal noises during discussions which turned political. This would happen A range of comments from our in staff meetings, or often during subject British YouGov survey underscore that seminars for lecturers. It went on for social pressure from colleagues is the years.” – Centrist, Remainer. primary deterrent to expressing views. The following pages outline this qualitative “Yes. Avoided expressing views due to data in detail. Those who wish to bypass bullying in workplace” – Centrist, this portion of the report can move ahead immigration restrictionist. to the following section. Testimonials have been lightly edited for spelling and “Yes. I have to be careful about expressing grammar. my views, because the bigoted politically While many simply replied with a correct, woke, thought police would object “yes” or “no” when asked whether they to my conservative (with a small ‘c’) had “refrained from airing views in views.” – Right Leaver teaching or academic discussions, or avoided publishing research, because of “I voted leave but was scared to reveal this possible consequences to your career from as my colleagues were so aggressive in doing so,” others offered an insight into their attitude.” – Tory, Leaver the climate of self-censorship that prevails for many academics on campus. “While enmity toward even centrists, let alone conservatives, is not overt, the “Totally, as I am not massively left-wing, atmosphere is such that homogeneity of and most of my colleagues were. In the opinion and sociopolitical views is implied interests of harmony and a comfortable and expected. For all the protestations of working environment, I used to just keep diversity and inclusion from the ‘woke’ squad, the one thing they most singularly

124

do not want is diversity of opinion.” – “Been told leavers are fascists” – Centrist, Tory, Remainer Leaver

“Yes, I tend to remain quiet. Letting my “I have had research ideas that I have not leave views out in the open would pursued as I think they would have definitely harm my reputation in the uni.” negatively impacted my career.” – – Tory, Leaver Centrist, Tory, Leaver

“Yes. I am a liberal conservative and the “Certain areas are off limits in academia vast vast majority of my colleagues at and you risk your job and reputation if you university are very Left wing. I support go against the herd. The days of academic individual responsibility, thrift, freedom are long gone and it is suicide to independence, hard work, low taxation and express views that differ from those of the small government. My colleagues opposed liberal thought police.” – Centrist, all of these. I have long been a euro sceptic Remainer and support leave- there were some colleagues who shared this view but were “I have ambivalent feelings about the very loath to say so publicly. The Hijab, however, I would not voice them in hypocrisy of the champagne socialist my department for fear of being colleagues was breathtaking. As was their considered Islamaphobic, rather than it open hostility to anyone who held views being an issue of patriarchy, which should different to their own.” – Labour, Leaver be fought against.” – Centrist, Labour, Remainer “Yes. No specific instances but generally I ‘guard my tongue’ except with people I “I don’t think my views on trans issues are fully trust.” – Tory, Leaver entirely in line with the current orthodoxy in academia.” – Centrist, Labour, “1) I hinted that Leave was not so bad as it Remainer seems to 2 colleagues who then accused me of being a leaver and were very angry “Yes. It’s hard to put a number on it in and abusive. 2) I told someone I had voted relation to frequency but there is definitely Leave and they called me a racist.” – a strong view within social work academia Centrist, Leaver that some views (left-wing) are acceptable and some views (right-wing) are not. “Yes, refrained from talking about leave Social work as a discipline sees itself I vote because the atmosphere is that this is think not really as an academy in which an inappropriate view to hold. This being debate and free-thought is the aim but as the case since the referendum.” – Centrist, an advocacy group for marginalized Tory, Leaver groups (which is a laudable aim, don’t get me wrong - but it’s very different from “As a Conservative voter I would not traditional views of academia in my share my political views within the opinion).” – Leftist, Labour, Remainer workplace. I am certain I would have been regarded with hostility from a number of “Agreement is often assumed by ardent fellow employees. I have seen this happen Remainers, and I usually avoid expressing with other employees. The ability to my views unless asked directly. discuss and explore political beliefs Agreement is also usually assumed by seemed to have disappeared.” – Centrist, management for neoliberal employment Tory, Leaver policies (incl. redundancies): I spoke publicly against these and was ‘frozen out’

125

as a result (i.e. moved to another, non- conflict that does not lead to new leadership role). Agreement is also often knowledge or growth, and so I say nothing assumed with various fashionable ‘right- about any of this, while hearing far left on’ ideas about, e.g., sustainability, views very, very regularly, as though decolonizing the curriculum, etc. I others didn’t exist at all, even as a generally avoid such discussions. I am possibility.” – Right, English, US aware of a colleague who has been complained about (anonymously) for her “I frequently bite my tongue around the so-called ‘TERF’ views – I support her lunch table with my colleagues, and I have fully. She only discovered my similar refrained from publishing articles at views because I do not self- censor what I outlets like Quillette or ArcDigital on publish (though I do tend to self-censor in controversial topics, because of fear of fact-to-face interactions at work unless cancellation/public shaming. It’s just not asked a direct question about my views).” worth it.” – Right, Political Science, – Leftist, Labour, Remainer Canada

North American Testimonials “The few conservative professors I know have told me to keep my head down. American and Canadian There were even active attempts to cancel respondents told a similar story, in which two gay professors in my department who considerations of social pressure from publicly resisted the pronoun issue. I tend colleagues, and to a much lesser degree to speak fairly openly in seminars, but I’m students, was the primary motivation. good at building relationships and I However, this was not unconnected to typically speak my mind only after career opportunities as colleagues often establishing a reputation as a thoughtful control departmental resources and and considerate colleague. However, there opportunities, something especially is a crushing weight of assumptions in the important for younger staff. Moreover, in air around me. Trying to make any the US and Canada, concerns about argument about my position usually political correctness were prominent not requires deconstructing so many just on the right, but among centrists and assumptions that my interlocutors no even around half of those identifying as longer even notices as assumptions. As for “fairly left” (among those who reported damage to my career, yes, I have self-censoring). frequently wanted to make written (rather The few conservatives in the than verbal) contributions but I don’t sample had the highest incidence of because I don’t want to leave a written feeling censored. Most cited social record right now. In my teaching I am pressure from colleagues: trying to co-opt the ‘rhetoric of cunning’ that comes from Critical Pedagogy in “Yes though not for career reasons so order to smuggle my viewpoints into my much as social reasons, not trying to stir curriculum.” – Right, English, US the pot. There is no room in the academic setting as I know it to express even “I definitely keep my views to myself and casually anything less than the most adopt a posture of non-conflict with my extreme left liberal views. I do not bother colleagues and students at all times.” – to try to have a conversation. I just smile Right, Psychology, US and nod, every day, and say nothing. I avoid confrontation, and err on the side of “I generally try to avoid any discussion of being as polite as possible…I don’t want personal politics, as I don’t think it’s to hurt anyone, or be hurt, or create productive” – Right, Political Science, US

126

“When colleagues invoke vague notions of “I generally wouldn’t share my views with equity to justify their contentious colleagues, as I think I’m probably more positions, I’ve refrained from asking for conservative than most/all of them.” – more information or questioning the Right, Political Science, US studies that they contend make their points indisputable. I don’t want to seem hostile “I am a closeted Conservative; I feel I to equity as a force in departmental can’t express my views because of the discussion. Also, on issues such as strongly leftist environment in organized labour and Indigenous rights universities.” – Right, Modern Languages, I’m a leftist by broader Canadian standards Canada but a right-winger in my department, and so avoid saying much that would give my “Daily experience. Discrimination against position away. To be clear, I often avoid conservatives is very real.” – Right, asking follow-up questions when Music, US colleagues make sweeping and contentious statements on these subjects as if their Centrist responses were very views were obvious and avoid stating my similar in tone to those of conservatives: own.” – Centrist, History, Canada

“The legal academy in general, and my “I frequently (maybe once every other faculty in particular, is very Left. week per class) do not say things that are Teaching and publishing are not major true for fear of career assassination from issues for me because I believe in colleagues who are much further left than I separating my personal beliefs from my am.” – Centrist, Sociology, US teaching and my analysis anyway. Open meetings, whether faculty meetings or “This spring my department released a academic gatherings, are where centrists statement in support of George Floyd that like myself, and certainly conservatives, attributed all racism and oppression to will not bother speaking up to offer Euro-American imperialism during the different opinions because they will be past 500 years. I am familiar with the ostracized socially and professionally” – evidence from ancient civilizations around Centrist, Law, US the world. That evidence shows oppression, domination, and othering are “Refraining from airing views generally characteristics of humanity and civilization occurs on Twitter. I would not, for no matter where and when. Rather than example, criticize the book ‘White engage in heated Zoom discussions, I Fragility’ or suggest that defunding the skipped the department meetings and then police is a silly idea on Twitter. These just didn’t vote when the statement came topics are less likely to come up, for me, at up for a vote.” – Centrist, Sociology, US work. However, post George Floyd, there will be a reading group where one of the “I disagreed with a statement that the suggested books is Kendi’s ‘How to be an university was systematically racist, but anti-racist.’ I am uncomfortable with this was the only person in the department and book, but I would be reticent to raise feel this has hurt my reputation and objections at work. We’ll see what standing in the department even though we happens (will I object?) as the reading had no discussion on the what the group materializes.” – Centrist, statement meant and whether it was true. Psychology, Canada The university personnel committee received at least one training session each year about implicit bias, but it would

127

plainly have been unspeakable to wonder I see more importance in policies being whether affirmative action and diversity carried out and this is why I overall hires might balance such biases out. Our support him right now. I definitely have to leftist chair wanted to give a session for be careful who I tell, because it is our first-year graduate students on automatically assumed I am a horrible supporting BLM. I questioned whether a person due to the way I view the conservative student would feel able to presidency right now even though I am voice their contrary opinions at the start of someone more than willing to discuss their graduate career and obviously caused differences and be friends with someone the chair irritation. These sorts of incidents who believes differently. All in all, very happen only a couple of times a year, but frequently I worry and have to refrain they poison the atmosphere for me.” – from convos with other professors.” – Centrist, Classics, US Right, Neuroscience PhD, US

“Since the explosion of ‘antiracism’ in the “I served on a hiring committee, and past few months, I’ve heard a lot of others on the committee quite frequently antiracist ideas that are just...bad. I would cited a candidate’s gender (female) or race be uncomfortable offering my candid (underrepresented minority) as a reason to assessment because ‘you’re hire them. I believe the level of bias shown either…antiracist or racist’ according to is morally wrong, and probably illegal. I the new dogma. I’ve been spending more avoided expressing this view. More time on the phone with my brother, generally, I avoid discussing political because we can have open discussions topics with colleagues because there is no about these things.” – Centrist, upside to doing so. I do not yet have Psychology, US tenure, and you never know you you’re going to upset. Regretfully, I do not have “I never air any view that is not strongly faith that people in academia are capable left/progressive out of fear of of separating political views from their retaliation/‘prosecution’; this happens at evaluation of a person’s work.” – Centrist, nearly every faculty meaning when Economics, US matters of teaching and student support are discussed, or when personnel matters are In some cases, respondents discussed.” – Centrist, Linguistics, US mentioned that this pressure narrowed the range of their academic inquiry: Career-Related Concerns “My department is very liberal, and often In a large minority of instances mixes social causes into how we run the where subjects elaborated on their department. I don’t know that I necessarily experiences, concerns bled over from oppose some of these actions, but my trepidation about peer pressure to fears inclination is to question and investigate if that their words could have career they are necessary. But I hold my tongue consequences. This affected their choices because everyone else seems so sure, and I about what to research, whether they am not certain that my less than gung-ho should comment in extramural outlets like approach would be welcome. I have also Twitter, and what to teach: avoided undertaking at least one research project because of the potential fallout-I “Absolutely I have refrained. I am by no think the project is very reasonable, but means an ‘in your face’ type political from what I’ve read I doubt my person and I am someone who is very administration or department would aware of Trump’s bad character; however, support pursuing research that challenges

128

(or even has the potential to challenge) ability to keep my job.” – Right, English, their moral/political views. I hope I’m US wrong, but at present I’m too unsure of my position to try.” – Centrist, Sociology, “Do not speak of politics, because I am a Canada member of the precariat. Would like to keep my job.” – Fairly left, Psychology, “It is harder to publish research if the Canada results could be considered ‘harmful’ to marginalized groups (for example, studies “Because I fear repercussions and know that do not find evidence of bias/unfair for sure that my colleagues discriminate treatment, but instead find differences in against people who do not share their risks/behaviors). Similarly, it is always a views, on issues where I differ, I keep risk to talk about such information in quiet. This is true particularly on matters class.” – Centrist, Sociology, US of race, immigration, and the Middle East (everyone is pro-Palestine and I am pro- “Frankly, even the publication of my Israel). Not in my research nor in my survey responses would be enough to teaching (I am not in a contemporary field) seriously damage my career. I’m not even but in meetings and policy discussions, I that – I’ve never voted Republican in have 100% remained silent because I am anything but a local election. But I have certain of retaliation. The biggest issue is major disagreements with the prevailing that I believe ALL LIVES MATTER and thinking that comes from critical race that it is racist to say otherwise and single theory, or deconstructions of gender, and out some races as more valuable than saying so would be enough to tank my others. But if I were to say this out loud I career.” – Centrist, Music, US would be completely excoriated and banned and there would be serious “Waited until after promotion to full consequences. As a Humanist, I am not professor to publish some of my research allowed to say anything because what I findings.” – Centrist, Sociology, US actually think would be censured severely. So I keep my mouth shut and only vote “Currently in my department, Marxist+ when it is a secret ballot. My department further left is the king; therefore a solid issued a statement after the BLM protests centrist like me is considered conservative. that I absolutely disagree with but I have I am up against all sorts of political, social, to see it up on our website as if it and economic barriers with my research represents me. It is infuriating and they because I work in business anthropology.” will not listen to reason. I am no racist and – Centrist, Anthropology, US in fact I think THEY are the racists. They make race a judgement point in any Staff who had not acquired tenure decision; I think fairness demands that we were especially skittish about revealing treat ALL people fairly as MLK said, views that did not conform to progressive according to ‘the content of their norms: character, not the color of their skin.’ I hate how racist my department is but I “I have refrained from airing my views cannot say a word or I’d be run out, totally because I know that I will be bullied, persona non grata, and my students would mocked, and abused on social media and face bad consequences, and my research in person. Additionally, I don’t have would lose funding.” – Centrist, English, tenure, so being honest about my US conservative views is dangerous to my

129

“I am in the field of Middle Eastern “Yes, because of the tenure track system Studies which, alas, is highly politicized. people are afraid to express their views For many years, I did not voice political and risk their tenure.” – Fairly left, views in academic settings.” – Centrist, History, US History, US To a greater extent than their “At first I did not, but once I began to be British counterparts, North American left- politely asked to remove myself from wing academics voiced concerns over conversations because my views are having their words misconstrued by their heterodox, I began to just stay quiet more own ideological group, which might make often to avoid the hassle with people. those of their own political stripe mislabel There is A LOT of research that is not them. A significant number of center-left published for political reasons and A LOT academics voiced worries about intra-left of research that is published for political pressures: reasons. This is completely wrong, and while I have not suffered from this yet, I “I have avoided discussion in teaching so know that this is an ongoing discussion as not to be mis-categorized as especially in my field of political science reactionary.” – Fairly left, Natural where people are often investigating Resources Science, US politically controversial topics.” – Centrist, Political Science, Canada “The big issue for us these days is racism versus other kinds of discriminatory “I am very wary about how (and with conduct – disability, anti-Semitism, etc are whom) I talk about trans issues. I know not taken as seriously (if acknowledged at people who have lost their jobs over this, all). It’s a matter of just how progressive simply because they talked about ‘Bruce you are, with intolerance for nuance or Jenner’ (‘deadnaming’) and challenged the diversity of views. I’m more senior now so ideas behind trans rights to enter sex- braver.” – Fairly left, Law, Canada segregated spaces, change birth certificates, etc.” – Fairly left, History, US “Yes. I used to teach a Gender Differences course in the late 90s. We were really able “Yes, I have refrained from airing my to talk openly about male/female views because I know that I will be differences. Now, it’s impossible to not bullied, mocked, and abused on social say the wrong thing. I have definitely media and in person. Additionally, I don’t made what I say more ‘vanilla’ than I used have tenure, so being honest about my to in the past. We say that education conservative views is dangerous to my should be about exploring difficult topics ability to keep my job.” – Right, English, but when people misconstrue things that US are said it can turn one apathetic.” – Fairly left, Communications, US “Yes, often because I am contingent (aka NTE) faculty who can be hired and fired at “Yes, I am frequently appalled at the whim.” – Fairly left, Music, US entitlement I hear from my colleagues but keep this to myself. They frequently sound “I also cannot express my objection to far as immature as the students they teach on left movements and university policies. It political issues and engage in ridiculous will certainly get me fired.” – Right, shouting matches over these views.” – Economics, US Fairly left, History, US

130

“Yes. As pre-tenure, there have been someone might not approve of.” – Fairly discussions in which it was clear that my left, English, US opinions could affect my relations with colleagues.” – Fairly left, Musicology, “Yes, my workplace is toxic, and it seems Canada difficult to speak up about without facing retribution in terms of resources.” – Fairly “While overall, my department seems left, Environmental Science, US welcoming of dissenting opinions, including political views, I have the “I shifted my scholarly focus away from impression that certain perspectives are Native American writers/history because taboo. Some involve criticism of left wing of hostile comments at conferences.” – perspectives, e.g. any critique of Fairly left, English, US affirmative action or anything that is not 110% supportive of minority “I would be concerned to speak against empowerment. But our field also punishes even the most extreme left wing ideals.” – certain forms of left-wing advocacy, Fairly left, Media and Communications, especially related to unionization, US advocacy for equitable pay and tenure rights, and open critiques of older faculty “I am sometimes thinking twice before with biased views (e.g. sexist views).” – saying things when I’m teaching or writing Very left, Music, US things in a paper that I believe can be construed as politically incorrect.” – Fairly “I was asked to modify a class exercise on left, Economics, US personal safety in low income neighborhoods because students viewed “Avoided sharing my opinion at our the association as racist.” – Fairly left, weekly diversity and inclusion training Geography, US meetings.” – Fairly left, Psychology PhD, US “Yes. My network is very left. I consider myself also very left, but I disagree with “We were encouraged (strongly) to attend them on many issues. I do not feel like I certain ‘optional’ workshops to help the belong in my left network and I do not BLM movements with political actions think they are left as they claim to be. This and while I agree 100% with the cause, I has gotten significantly worse during the do not like my workplace pressuring me to COVID crisis. I find the polarization take political action.” – Fairly left, occurring to be very troubling, and I do Psychology PhD, US not agree with the left on their reaction to this current situation.” – Fairly left, “Absolutely it would be career suicide to Sociology, US say anything besides a far-left opinion out loud in front of other faculty members in “Yes. A few times – more since this our college.” – Very left, Sociology, US spring. I have not raised points where I disagree with ‘cancel culture’ and aspects “Yes – any opinions that could be seen as of ‘post-colonialism.’” – Fairly left, right of centre (especially wanting Rhetoric, US controlled immigration) would damage career.” – Very left, Journalism PhD, UK “Yes. Less a matter of refraining from “I feel I will be judged if fellow faculty asserting opinions than of being picked on members know my partner’s views.” – for any particular formulation of those Fairly left, UK, Sociology ideas – that is, using language that

131

Transgender Issues be ruined. Lists are apparently kept of women who air such opinions and they are A second major area of concern on prevented from advancing in their career. I the left emerged from the worries of the was also very frightened to see Dr. Lowrey significant number of gender-critical received death threats, rape threats on scholars about running afoul of approved social media which in some cases seemed pro-trans orthodoxy. Indeed, 6 UK to be encouraged by faculty.” – Fairly left, YouGov respondents, 5 on the left and 1 Anthropology, Canada centrist, mentioned self-censoring their views on the trans issue, representing 10% Leftist Complaints Against Conservatives of complaints from leftist respondents. In the North American data, 11 respondents, Those who self-censored due to all on the left (5 “fairly left” and 6 “very right-wing pressure worried mainly about left”) mentioned self-censoring on administrators or students at religious and transgender questions. This likewise conservative universities, or about threats represents around 10% of the stated from the off-campus right. Some cited reasons for self-censorship provided by Middle East/Israeli politics as a third rail. leftist academics, among the most Student evaluations were noted as a common complaints provided by leftists particular concern for some: who self-censored. Here again we see evidence from a wider sample of the “As I teach at a Catholic university, I am powerful forces arrayed against gender- careful in what I say about abortion.” – critical perspectives in the academy: Fairly left, Political Science, US

“I avoid talking about trans and queer “As a teacher (PhD student) I try not to issues with colleagues.” – Very left, US say anything in front of my students that could be perceived as hostile to their “I am fairly left wing but have avoided beliefs, unless it is something the uni doing research or talking about gender and already takes a stand on (e.g. transphobic trans issues because I anticipate censure comments would be shut down).” – Fairly and exclusion from the left. (it would be left, English, Canada bad for my professional relationships with colleagues even to raise certain “I avoid political discussions with younger questions.)” – Very left, Philosophy, US students because I know the university is mostly conservative, though my “Yes, academic discussions on trans department is not.” – Fairly left, women have become dangerous at my Psychology PhD, US university and one faces serious career consequences for airing them. Dr. “I used to work at a very conservative Kathleen Lowrey was fired from her role Christian college in the USA, and was as Associate Chair for saying women do regularly told not to publish or discuss not have penises and being gender critical. certain topics (e.g., abortion, evolution, There is no room for debate on this subject gay rights, etc.)” – Fairly left, Philosophy, at my university. If you even have a Canada question about trans women or reject gender reassignment surgeries for children “I worry that my students may be more (e.g. phalloplasty, mastectomy of children conservative than I assume and what I under the age of 18) etc then you are present as fact from my expertise may be called transphobic. I have been warned not interpreted as lefty ‘political opinions.’ I to air any such opinions or my career will worry that speaking against Trump

132

policies when they come up in the context faculty air their views in class, they of class discussion could cause complaints typically cause a backlash that is an from students/parents.” – Very left, Art annoyance to everyone. Their conservative History, US undergraduate students tell their parents, who tell rich alumni donors, who tell their “Yes, not frequently but I avoid extremely legislators, who then bother the far left opinions in case they have administration, which can do little because repercussions but also so as not to alienate of legal protections for academic freedom. students.” – Very left, English, US For many reasons I try to separate my activism from my work. One further “I am careful not to discuss my political reason is that I don’t tend to like the beliefs in undergraduate teaching because ‘academic left’; they typically talk a lot I don’t want to alienate right-wing and ‘cancel’ each other, but rarely do students.” – Very left, Communications, anything of political significance.” – Fairly US left, Linguistics, US

“Many of my students are more Some leftist professors perceived conservative than I am, so I need to be threats from outside the campus to careful about what types of theories and academic freedom. These include which theorists I introduce in which order, comments expressing fear of the activities or else students might take action against of Turning Point UK in monitoring left- me.” – Very left, English, US wing lecturers in Britain:

“I teach at a religiously affiliated “I recently had an incident where a student university that tends to have a more expressed his concerns about the lack of conservative student body, therefore I try right-wing views expressed by lecturers to keep my political views out of the and this person also criticized my conversation, and instead focus on department as being left-wing. I am not research and what the science says.” – afraid of what the university might do but Fairly left, Psychology PhD, US rather I might find myself the subject of a report to Turning Point, or outed on twitter “Previously worked at a private Baptist for what I’m teaching.” – Labour, college, teaching gender, sexuality, race & Remainer ethnicity, social problems, and cultural diversity courses – many students were “Not at the moment, but I am concerned openly combative to even broaching topics about Turning Point UK.” – Labour, that challenged their beliefs and would Remainer penalize me in evaluations for teaching what is considered canon in my discipline “With organizations like Campus Watch, as inherently political or heretical. and the Proud Boys, and other White Experienced at least with two or three Supremacist, misogynist groups, I am students in each course taught (4/4 load always very careful to keep my own over 2 years), if not more.” – Very left, activism out of the classroom. Of course, I Sociology, US am also careful to create a space where students who are vulnerable because of the “I avoid airing my (left leaning) political their immigration status, gender, race, views in teaching and academic class, disability, etc. are protected from discussions. I work at a public university attacks.” – Very left, Sociology, US and the state that funds it has a very conservative government. When left wing

133

“I once published a pro-life piece that get them dismissed on the basis of their suggested government transfers reduce comments in class, this represents a clear abortions. I was pilloried on blogs and in extra-campus threat to academic freedom. private emails, including threats, by right- Universities should not act upon student or wing activists.” – Fairly left, Political outside actors’ complaints about lecture Science, US content unless the lecturer is blatantly quashing the expressive freedom of Middle East Politics, Israel, and students for politically biased reasons or UK’s Prevent duty to report on discriminating in grading. radicalization were cited by some on the More generally, some left-wing left as reasons to hold back: academics felt they could not express their views in class due to student sensibilities “Prevent makes me very uncomfortable, or pressure to remain neutral, though particularly when discussing urban others felt it was important to let students uprisings and armed resistance during one know where they stood in order to have a of my modules.” – Labour, Remainer good discussion:

“No but as a supporter of rights for “Yes, tend to be wary of expressing Palestinians and an arranger of placements personal politics as management very in the West Bank for students over the against this. Would love to discuss it more summer vacation (linked with their as seems deeply patronizing not to – I course/career choices) am mindful that it teach adults.” – Labour, Remainer causes considerable angst amongst some others.” – Labour, Remainer “I try to avoid making any political statements when teaching.” – Labour, “I avoided doing research on antisemitism Remainer and islamophobia pre-tenure because these are ‘third rail’ topics in my subfield of “It is generally taken that lecturers do not Middle East politics.” – Fairly left, air their personal political views in Political Science, US influencing students. I have had some lively discussions with students over this “I have changed my syllabus to avoid issue particularly in relation to health care. discussions of Israel, because I expect I have had to temper my own views in the conservative colleagues to attack me.” – University setting. When undertaking Fairly left, Philosophy, US research using IPA [Interpretative phenomenological analysis] I had to ask “I am careful not to express my support for the university not to make my work Palestinian rights, as that can lead to false readily available to students as I had to set accusations of antisemitism.” – Fairly left, out my preconceived ideas and biases Political Science, US which included my political beliefs. In an increasing litigious era, lecturers/ These leftist comments raise researchers need more protection as important issues about the boundaries of students seem to use any angle in order to the Prevent duty in the UK, which can bolster erroneous claims.” – Labour impinge on criticism of British or Remainer American Middle East policy, or of Israel. Here it is important to allow maximum “I avoid explicit discussions of my academic freedom within the law. If political views in some of my courses out Turning Point were to instigate Twitter of fear of student retaliation. I lost my first mobs against particular lecturers, or aim to (and only) tenure track job, ostensibly for

134

poor student evaluations. Oddly, two other These final sets of comments show queer faculty were denied either tenure or that an ethic of impartiality holds, at least promotion that same year, ostensibly to some degree, in the classroom and in because of poor student evaluations. I departmental settings. While this could be know several people who expressed seen as a “despotism of custom” that feminist views or were out to their infringes on the expressive freedom of students who had the same experience. academics, one could argue that in these Now I am in a renewable contingent instances, an ethos of neutrality position, and I am VERY careful about counteracts the chilling effect on those what I say in the classroom since I know with dissenting viewpoints, increasing that the ostensibly ‘liberal’ administration support for a free speech culture. A can, at any time, just decline to renew my balance must be struck between allowing contract.” – Fairly left, Anthropology, US maximum freedom for lecturers and colleagues whilst being mindful of Finally, some far-left professors academics’ power (i.e., over grading, class mentioned that they refrained from open atmosphere, departmental environment) to activism. Here is evidence that in some increase the considerable degree of self- departments, progressive advocacy is censorship that already exists among frowned upon: conservative staff and students. Though chilling effects among conservative “Speaking out about issues related to students are mainly due to peer pressure social justice has gotten me reprimanded at and social media rather than worries about my workplace, specifically around race, lecturers marking conservatives down, the gender, and gender identity. I have been latter is not, as we shall see, entirely absent approached by individuals who do not from the classroom.89 know me well, and told that I should be Due to the position of the lecturer less outspoken. Although the department is as an arbiter of a student’s grade, norms relatively liberal, they are less so when it like impartiality arguably expand the comes to individual interaction and degree of liberty in a university even if dealing with their own bias.” – Very left, they limit the expressive freedom of Psychology PhD, US lecturers at the margin – just as norms of debate which limit heckling or frown on “I frequently have to be mindful of how I excessively unruly or opinionated answers express my views on racial diversity in permit a richer exchange of views. What is science. I do not want to be labelled an vital, however, is that collegial norms ‘activist,’ though I work hard towards rather than formal sanctions or mob increasing diversity in science. However, cancellation be deployed: lecturers who the way one must talk about the enduring express their political views in class must effects of white supremacy and racism on not be disciplined, even as it should be black people specifically is sometimes too permissible for peers to criticize or urge uncomfortable for many of my white them to work towards a politically non- colleagues to hear without some discriminatory class atmosphere. softening.” – Very left, Psychology PhD, US

“I have refrained from calling out racist classmates because I wanted to avoid drama. It happens once every few months.” – Fairly left, Healthcare, US

135

Section IIc: Political Discrimination left”). To what extent is the leftist majority likely to discriminate against, or even seek to force out, conservatives? Political Discrimination: Quantitative I broadly follow the methodology Evidence of three existing studies in the United States (Inbar and Lammers 2012; The above data show that many Honeycutt and Freberg 2017; Peters et al. conservative and Leave-supporting 2020).90 These ask people if they would academics, especially in the social politically discriminate in hiring, on grant sciences and humanities, feel estranged applications, and in refereeing a journal from their departments. Half in Britain and article. These studies found substantial 70% in the US say there is a hostile political discrimination. Inbar and climate for their beliefs in their Lammers, in a 96% left- or moderate- departments and a similar number report leaning sample of psychology academics, self-censoring in teaching and research. discovered that 38% were at least Twice as many identify against, as with, “somewhat” likely to discriminate against their departmental cultures. Among a conservative job applicant. The share current staff in the social sciences and willing to discriminate against a paper or humanities who voted for Trump or Leave, grant taking a conservative perspective over 8 in 10 say they are not sure they was 19% and 24%, respectively. would be comfortable expressing their Honeycutt and Freberg, surveying those in views to colleagues (64-97% a wider variety of disciplines, found that uncomfortable, 3-18% unsure). among left-wing academics, 16% would Is this mere perception, or is it discriminate against a conservative paper, grounded in reality? One clue that such 22% against a conservative grant fears are reasonable lies with the fact that application, and 33% against a only a minority of Remain and Biden- conservative job applicant. voting academics thought a Brexiteer or But these authors also found that Trump voter would feel comfortable academics on the left and right expressing their views to colleagues. If, as discriminated against each other in equal I will show, there is evidence for measure: among conservatives, 21% discrimination against conservatives, or would discriminate against a left-oriented even a willingness to expel a staff member paper or grant application, and 32% for their views, then it is perfectly against a left-wing job applicant. Peters et understandable for conservatives to al. (2020), in a 75% left-leaning sample of conceal their beliefs in teaching, research, European philosophers, found that 48% of and collegiate conversation. academics and graduate students would In what follows, I turn my attention discriminate against a right-wing hire and mainly to the second “iceberg” pyramid in 35% would discriminate against a right- Figure 4 (perpetrator’s perspective). This wing grant application. Leftist concerns the leftist and centrist academics philosophers were more discriminatory who comprise about 90% of the than right-leaning philosophers by a full professoriate and largely set the 10-15 points. atmosphere at work. In the social sciences My approach follows much of the and humanities in Britain, about 60% of a thrust of these studies but differs in three conservative’s colleagues will identify as important ways. First, previous studies left (45% “fairly left” and 15% “far left”) contacted academics directly by email, and 30% as centrist. In America and collecting a convenience sample of around Canada, my survey would suggest that 500-600 applicants. Their samples 75% are on the left (including 25% “far achieved a response rate of around 25%,

136

introducing the possibility of self- 2. List 1 contains 3 statements and list 2 selection, though this does not present a contains the same 3 plus a statement on major problem for comparing between whether the respondent would groups in their samples. This is the method discriminate. Subjects are asked how many I use for US and Canadian academics, of the statements they agree with. These though I cast my net wider and have a statements are designed to tap a number of lower response rate than previous studies equally contentious questions, and permit that targeted much narrower subgroups those of different ideological views to find drawn from academic bodies or lists. something to agree with. Because people For Britain, I drew on YouGov’s are not questioned directly about any one panel of those who completed a range of of the items, it is not possible to know different types of surveys for remuneration which individual would discriminate or and just happened to be academics or not, but it is possible to calculate an retired academics. They are thus less likely average level of discrimination across the to be selectively attracted toward filling sample. This anonymity is what allows out a particular survey. Between 61% and people to answer questions free of pressure 76% of YouGov’s panel of professors and to adhere to social norms. lecturers responded, resulting in a sample Respondents are allocated that is more likely to represent the actual randomly into one of the two treatment population of academics than any other groups in Table 3. If there is no study (UK or US) to date.91 I also discrimination, the average number of surveyed British and North American PhD statements people agree with should be the students on the Prolific Academic same in both lists. In this case, assume that platform, where I was able to achieve 86% people agree with two statements of three of the target pool in Britain and 63-72% in in the identical lists. List 1 therefore has an North America. average of 2, so if list 2 has an average Second, I used an experimental score of 2.5, this means that half of survey survey design called a list experiment respondents agreed with statement four across all samples in order to circumvent and are thus willing to discriminate. The the social pressures against admitting to difference between the two lists, .5, would open discrimination.92 List experiments represent the average discrimination level work as follows. Consider Table 3. Half in the sample. the sample receives list 1 and half gets list

How many of the following statements do you agree with? List 1 List 2 Statement 1 (no, yes) Statement 1 (no, yes) Statement 2 (no, yes) Statement 2 (no, yes) Statement 3 (no, yes) Statement 3 (no, yes) “If a known Leave supporter applied for a job at my workplace, I would try to avoid hiring them” (no, yes)

Total: x Total= x + discrimination Table 3. List Experiment Design I can subsequently restrict my or women, and so on, so long as there are analysis to subgroups such as Trump or sufficient numbers of individuals in each Clinton voters, Leavers or Remainers, men to learn something. In this case, I would be

137

looking at, say, all men who took the In the US, 22% of academics survey, comparing scores for men in list 1 admitted they would discriminate against a with those in list 2. Just to be completely known Trump supporter in a direct sure that results are not driven by chance question (“If a known Trump supporter differences in who gets which list, I can applied for a job at my workplace, I would use statistical models to control for major try to avoid hiring them”), while in differences such as age, income, gender, Canada, 26% openly said they would do discipline, work status, political views, and so. Note that American politics receives so forth. considerable coverage in Canada, and A third difference from existing Canadian academics readily understand, studies is that, due to the list experiment, and arguably identify with, American my question is more direct. Rather than political categories. Finally, note that these ask whether people would discriminate on questions were all asked well before the a 7-point scale from 1-Not at all through 4- November election and the Capitol Hill Somewhat and 7-Very Much, mine is riot of January 6, 2021. phrased as a simple Yes/No/Don’t Know. Turning to the results, those in list Thus, I asked British respondents whether 1 in Table 3, who were asked how many they would agree with this statement: “If a among a list of 3 statements they agreed known Leave supporter applied for a job at with, were subsequently asked the my workplace, I would try to avoid hiring discrimination question as a stand-alone them.” Just 10% of the YouGov study said item. The difference between this and the they would avoid hiring the Leaver, 76% revealed discrimination in list 2 helps us that they would not, and 14% replied that understand the extent to which people are they didn’t know. Among Remain and concealing their actual willingness to Left voters, the share willing to discriminate. discriminate inched up slightly to 12%. Smaller datasets are less reliable While these figures are concerning, they for list experiments because there is are considerably lower than the numbers natural variation in answers to the three obtained in two previous US-based emotive but unrelated list questions that studies. Much of this probably derives only dissipates with sufficient cases. from differences in question wording Accordingly, Figure 99 presents results between the 7-point scale and my direct that are significant at the 1% level across yes/no item. Some may also be due to the the five larger surveys, and in only one nature of my sample, which is arguably survey – the smallest – is the difference more representative of the sector as a between the lists not significant. This whole, encompassing lower-ranked suggests that political discrimination universities and not just top research against right-leaning academics is schools. pervasive.

138

Discrimination Against Trump/Leave Supporter in a Hiring 50% 45%** 45% 40%*** 40% 35%*** 35% 32%*** 33%

30% 27%

25% 22% 22%*** 20%

15% 12% 10% 8% 10% 6% 5% 0% US (N=1,308) Canada UK Yougov UK Mailout N America UK PhD (N=463) (N=820) (N=222) PhD (N=397) (N=170) Open Concealed

Figure 99. *p<.05;**p<.01;***p<.001 in chi-squared tests of crosstabulations, testing whether there is a significant difference between the mean score of the 3-item and 4-item lists. Number of observations in parentheses.

Comparing the difference in are approximately twice as high as average score between the 3-item list and academics’ stated willingness to the 4-item concealed list with the average discriminate in the direct question.93 score from the one open question on Among 396 North American PhD students discrimination, I find that the actual level surveyed, 33% said they would of discrimination among North American discriminate against a Trump voter in a academics is nearly twice as large as stated direct question and 35% in the concealed openly, and over three times as large in list condition – the narrower difference Britain. Though this is a dispiriting level between conditions may well be due to the of bias, the level of concealment is a smaller sample, around 200 in each control positive sign insofar as it suggests that or treatment group. most of those who discriminate sense that In the UK, 32% of academics this is problematic and must be kept would avoid hiring a Leaver who applied hidden. By contrast, PhD students seem for a job, whereas only 10% would admit not to be concealing their bias to the same it directly. I therefore assume a degree. This may be due to weaker anti- concealment multiplier of up to three political discrimination norms at the PhD across similar subsequent UK questions on level, though the level of discrimination is direct discrimination. low in the small UK PhD sample. Figure 100 shows the average In North America, the list willingness to discriminate against a Leave experiment reveals that 40% of American supporter across various subgroups in the academics and 45% of Canadian UK YouGov sample. The leftmost bar academics would discriminate against a reveals that 32% of the sample would known Trump supporter. These numbers discriminate against a Leaver. Thirty-

139

seven percent of Remain supporters, 35% previous literature which found centrists to of those identifying as left, 39% of 2019 lie intermediate between those on the left Labour voters, and 44% of female and right in their propensity to academics revealed that they would discriminate. While the differences in discriminate against a Leave applicant. willingness to discriminate between major On a five-point scale, 23% of the political groups are all statistically sample agreed with the statement “I would significant, the male-female difference consider myself an activist.” Fifty-eight only reaches borderline statistical percent of self-described activist scholars significance, so we should not place much would discriminate against a Leaver emphasis on this finding.94 Age, and compared to 20% of faculty who disagreed whether a respondent is a professor or with the statement (i.e., are non-activist). lecturer, active or retired, or in the SSH or Surprisingly, 34% of centrists also STEM sector are not significant predictors revealed that they would discriminate, of discrimination. which runs counter to findings in the

Share Who Would Discriminate against a Leaver's Job Application (Britain) 70% 58%* 60% 50% 44%+ 39% 37% 35% 40% 32%*** 34% 30% 23%+ 20%* 20% 10% 7% 1%* 3%** 0%

Figure 100. +p<.1;*p<.05;**p<.01;***p<.001. P-value on “All” is based on difference between 4-list (including not hiring a Leave supporter) and 3-list (excluding hiring Leaver question). All others are based on regression on total statements agreed with, and whether there is a significant interaction between the variable in question – as compared with its other categories – and the treatment effect (of having an extra question on hiring a Leave supporter on the list). Note that low sample sizes on categories like Right and Leave reduce statistical significance.

The fact that nearly a third of person panel and ten in a 30-strong academics are willing to discriminate department, a number of whom will against a Leave applicant is highly evaluate CVs for a longlist and many of concerning. This means that there will be whom will attend job presentations where one biased member on a three-person the departmental “steer” strongly shapes appointment committee, two on a six- the final choice.

140

On the positive side, just 10% are In the United States, Figure 101 willing to openly admit to discrimination – shows that most of the same academic 17% among activists. More importantly, subgroups that scored highly on political two-thirds of left-wing academics and over discrimination in the UK also emerged: 60% of Remainers would not discriminate activists and leftists are more against a Leaver, even when this view is discriminatory than non-activists and non- completely concealed. Younger staff are leftists. Women are somewhat more no more likely to discriminate than the old, discriminatory than men, but this did not suggesting that the problem is not getting reach conventional statistical significance. progressively worse. This is an important While a 20% minority of right-leaning finding in itself, because it indicates that academics also said they would people are unwilling to admit openly that discriminate against a Trump supporter, it they discriminate on political grounds. is difficult to be confident of this due to This suggests that there is something of a low sample size (which also affects the norm against political discrimination, even small group of pro-Trump academics).95 though it is flouted in practice by an important minority of staff.

Share Who Would Discriminate against a Trump Supporter's Job Application (US) 60%

50% 47% 46% 47%* 43% 41%*** 39%** 39% 40% 34% 30% 20% 22%* 20%

10% -20%+ 2%** 0%

-10%

-20%

Figure 101. Note: “All” column includes American respondents who dropped out before answering ideology questions (N=1220). Subgroup columns based on American respondents who completed survey (N=706). Negative score on Trump supporters (N=33) and positive on Right (N=55) arguably reflects effect of variation in answers to list question from a low sample. + p<.1; *p<.05; **p<.01; ***p<.001. P-value on “All” is based on difference between 4-list (including not hiring a Trump supporter) and 3-list (excluding hiring Trump supporter question). All others are based on regression on total statements agreed with, and whether there is a significant interaction between the variable in question – as compared with its other categories – and the treatment effect (of having an extra question on hiring a Trump supporter on the list).

141

I also asked about people’s one’s academic job. By contrast, left willingness to discriminate against a right- academics meet conservative academics or leaning term paper from a student. Overall, encounter right-leaning content much less there is much less political discrimination frequently. Any negative reactions they in marking, with just 4% of North experience are not meliorated by many American academics openly admitting positive interpersonal contacts with known they would discriminate, rising to 7% in conservatives on the faculty. In short, it is the concealed condition. This number more likely that left-wing academics will masks large differences between American create and maintain a homogeneous (15-16% willing to discriminate in ideological network, impeding concealed condition) and Canadian (-2%, interpersonal contact with those who have i.e., effectively zero) respondents. The opposing views. This latter interpretation American figure, though not high, is would accord with evidence from Britain approaching a concerning level, and that Remain voters and those on the left guidance should be issued against political are now considerably less comfortable discrimination. with Leavers and conservatives than the The most concerning of all, reverse. Those on the left are more likely however, are the results for US and to unfriend online, socially distance from, Canadian doctoral students, with 12% and wish for their children to avoid openly admitting they would discriminate marrying, those on the right.97 Thus we against a right-leaning term paper, rising would expect the left to discriminate to 35% in the concealed condition. against the right more than the reverse. Moreover, doctoral students are as likely Against this interpretation, the fact to give left-leaning term papers a higher that those on the right are a small minority grade as they are to allocate them a lower may incline them to be more self- mark (7% for both). The sample is not conscious and mobilize to favor their own. especially large (338) and there is a This would lead us to predict that the margin of error on list experiments with right-wing minority discriminates against smaller samples. the left-wing majority more. To interrogate this issue, I asked a Do Left and Right Discriminate Equally? series of questions to tap both “hard” and “soft” discrimination. Hard discrimination The list experiment tested for bias is discerned through the question, “In against Leavers only. This begs the choosing a job candidate, I would be question of whether conservatives and inclined to support a known centrist over a liberals discriminate equally, as Freberg known leftist with a slightly stronger track and Honeycutt (2017) suggest, or whether record.” The term “leftist” was then this problem is accentuated in particular replaced with “conservative” to test for parts of the political spectrum, as Peters et anti-conservative bias. This is “hard” al. (2020) find where leftist academics discrimination because it involves giving a were sometimes nearly twice as likely to job to a weaker candidate for political discriminate as those on the right. reasons. The contact hypothesis suggests Soft discrimination – in which that those with greater intergroup contact political affiliation is used as a tie-breaker become more tolerant of difference.96 – is measured by a question which asks: Right-wing academics encounter left- “In choosing between two equally leaning scholars and their writing qualified job candidates, one a frequently. It is difficult to reject a Corbyn/Sanders supporter and another a majority of one’s colleagues and their Leave/Trump supporter, if you had to pick output, and assessing it is a routine part of between them, who would you be inclined

142

to choose?” Respondents could pick one of 50% of right-wing and 40% of left-wing the two, or state “no preference”/”don’t academics preferred their own candidates know.” The last was the most common (as did 38% of Labour voters and 33% of choice. Tory voters, not shown) where credentials are evenly matched. Looking at measures Political Discrimination in Britain of “hard” discrimination, 20% of right- wing and 15% of left-wing academics In the UK, around 30% of Leavers would discriminate against those of the would choose a Leave supporter over a opposite camp who had better records. In Corbyn supporter, and 29% of Remainers addition, numbers (not on chart) show that would choose a Corbyn supporter over a 12% of Tory voters and 13% of Labour Leaver, if both were equally qualified. In voters would discriminate against the other both cases, the majority of academics said side. Twelve percent of Remainers were they would have no preference and would willing to discriminate against still not favor one of the candidates. conservatives with stronger track records Focusing only on currently active than a centrist, while 10% of Leavers SSH academics shows more animus would discriminate against leftists with among Remainers than Leavers: the stronger records than a centrist. YouGov survey found 37% of Remainers The mailout survey of current SSH (N=191), but just 27% of Leavers (N=22), academics shows that 20% of the very few willing to discriminate against a hire from right-leaning academics would the other side of the Brexit divide. The discriminate against a leftist hire, and 12% mailout survey, limited to current SSH of leftists would discriminate against a academics, shows a similarly lopsided right-leaning hire. Thus, if anything, there picture, with just 11% of Leavers (N=19) seems to be somewhat greater bias from but 22% of Remainers (N=142) willing to the right this time as compared to the left. discriminate against a hire from the other One pattern that we also see in the North side when the two are otherwise evenly American data is that discrimination matched. Here the Peters et al. 2020 against concrete populist categories like findings of lopsided bias appear to be Leaver or Trump supporter is greater than borne out for the Leave-Remain divide, against more abstract and non-specific though we shall see that this does not hold “conservatives” or right-wingers – a for the left-right divide. category that includes the less UK results by ideology are controversial libertarians. presented in Figure 102. These show that

143

Job Discrimination, by Own Political Orientation 60% 50% 50% 40% 40%

30% 25% 20% 20% 15% 12% 10% 10% 10% 7% 9% 10% 6% 6% 5% 5% 6% 2% 3% 2% 1% 0% Discriminate Discriminate Discriminate Leaver over Corbynite over against Leaver against against Left Corbynite for Leaver for job Conservative job All Right Center Left

Figure 102.

When I control, in Figure 103, for error that stems from small samples of professor or lecturer status, SSH or STEM, “very right” academics into account. This age, and gender, and look only at the 5- suggests that, following Freberg and point left-to-right question, I find that both Honeycutt (2017), each side is similarly sides engage in approximately similar willing to discriminate against the other in levels of discrimination in hiring. This is hiring. especially so when I take the measurement

144

Figure 103. Pseudo-R2 =.168 in model of discrimination against the left and .144 in model of discrimination against the right. Fairly right is significant at the p<.001 level in model of discrimination against the left while very and fairly left are significant at the p<.001 level in the model of discrimination against the right. Does this mean there is nothing to group is equally biased, but the worry about? Hardly. From my list compositional effect is to slant the playing experiment, I know that 12% of Remain- field against Leavers and conservatives. voting academics say they would This is a similar finding to previous discriminate against a Leaver in a job, but studies. the actual figure is 37%. So, the true levels American results in Figure 104 tell of willingness to discriminate could be up a mixed tale. On the one hand, political to three times higher than stated. discrimination against populist right Second, because Leaver (16%), supporters is worse: 56% of left-leaning Tory (15%) and right-wing (9%) academics said they would favor a Sanders academics are a small minority of my supporter over a Trump supporter when British survey respondents, there is the two had equal merit, and only 25% structural discrimination against them. said the candidates’ politics would not Though individuals discriminate the same affect their decision (in Britain, by in hiring, the collective (“structural”) contrast, 52% of leftist academics said impact of a left majority and small politics would not affect their decision). conservative minority in academia means Moreover, 21% of American right-leaning that there is three times as much academics said they would back a discrimination against Leavers (24.9%) as Trumper over a Sanders supporter, against Corbynites (7.4%), and nearly considerably lower than the 56% anti- twice as much discrimination against Trump soft discrimination coming from conservatives as against the left. Each the left.

145

This might suggest that the However, when it comes to American academic left engages in 2-3 discriminating against a hire who has times as much soft discrimination against better credentials, the net discrimination Trump supporters as conservatives do against conservatives compared to the left against Sanders supporters, backing the is just 12 versus 9%, a more positive Peters et al. 2020 story of disproportionate picture than in the UK. This is partly left-wing bias. On the other hand, 17% of because of the role of centrist American conservatives and 16% of centrists would academics, who resemble conservative discriminate against a leftist hire whereas academics in discriminating nearly twice only 14% of American academic leftists as much against the left as against the would discriminate against a conservative right. By contrast, British centrists align hire. These findings are more in line with closer to the left, discriminating Honeycutt and Freberg’s 2017 finding that disproportionately against conservatives discrimination flows both ways, in equal and Leavers. measure. In addition, this question asks In terms of the net effect of about hiring a “conservative,” rather than discrimination among all academics, focusing on the more charged populist Trump supporters are heavily category of Trump supporter. It may be disadvantaged in soft discrimination that left-wing academics in America make compared to Sanders supporters by a more of a distinction than British leftists whopping 46-2, showing that for two between cultural-populist forms of equivalent candidates, Trump supporters conservatism and others such as fiscal or will be strongly discriminated against. foreign policy conservatism. The Note as well that just 34% of American American left has arguably been more respondents said that the politics of the focused on cultural categories of candidates would make no difference to disadvantage while the British left has their decision, compared to 62% for the traditionally been more materialist and equivalent question in Britain. class-based.

146

Job Discrimination, by Own Political Orientation (US)

60% 56%

50% 46%

40%

30% 21% 21% 20% 17% 16% 14% 14% 11% 12% 8% 9% 9% 10% 6% 7% 4% 2% 2% 0% 1% 0% Discriminate Discriminate Discriminate Trumper over Sanders-ite over against Trumper against against Left Sanders-ite for Trumper for job Conservative job All Right Center Left

Figure 104. N=1308 for “all” and 803 for right, center, and left bars. This is because Trump discrimination question asked much earlier, before many dropped out of the survey.

Canadian results in Figure 105 toward conservatives (18%), the right show a pronounced anti-Trump bias, as in make up just 4% of academics in the America, and a considerable anti- Canadian sample. Meanwhile, Canadian conservative slant, as in Britain. This centrists discriminate evenly between right includes a sharp 62-17 tilt in favor of and left, conferring no advantage on hiring a Sanders supporter over a Trump conservatives. Hence when it comes to net supporter to break a tie, with just 32% discriminatory effects in hiring, saying the candidates’ politics would not conservatives face a penalty of 16% influence their decision. While discrimination in Canada compared to just conservative academics in Canada have a 7% for leftists: worse than the 5:3 ratio in higher propensity (25%) to discriminate Britain and the 4:3 ratio in the US. against a leftist than Canadian leftists do

147

Job Discrimination, by Own Political Orientation (Canada) 65% 62%

55% 51%

45%

35%

25% 25% 22% 18% 17% 16% 17% 17% 15% 12% 12% 10% 9% 8% 7% 5% 6% 5% 1% 1% 0%

-5% Discriminate Discriminate Discriminate Trumper over Sanders-ite over against Trump against against Left Sanders-ite for Trumper for job Conservative job All Right Center Left

Figure 105. N=463 for “all” and 290 for right, center, and left bars. The UK academic mailout survey preference, and none would choose the shows a 21-1 Corbyn over Leave Trump supporter. advantage in the case of a tied set of UK PhDs are more even-handed, candidates, thus a strong leftist advantage. with 46% saying no political preference, However, 69% of UK academics said they as compared to 40% for the Corbyn had no preference among the candidates, supporter over a Leaver, and 6% for the and would not take their politics into Leaver over the Corbyn supporter. account. Even among leftists, 61% of UK Focusing on the subset of British SSH academics in the mailout survey said leftist PhDs, 59% would choose a politics would not affect their choice. Corbynite over a Leaver, but 31% said These reinforce YouGov findings showing they would not have a preference. that most UK academics would not engage Norms against political discrimination in soft discrimination, unlike their US and appear to be somewhat stronger in Britain Canadian counterparts. than North America. While Trump’s These transatlantic differences personal antics do add a distinct dimension largely replicate among PhD students. For compared to Leave, which may account the 338 North American PhD students in for some of the difference, one could our sample, 65% would choose the equally argue that charismatic Leave Sanders supporter over the Trump standard-bearers like Nigel Farage and supporter, 5% would opt for the Trump Boris Johnson also arouse strong criticism. supporter over the Sanders supporter, and Moreover, the powerful partisan only 25% said they would have no polarization in America cannot explain preference. The numbers for SSH PhDs on why Canadian findings resemble those of the left are extreme: 82% would select the the United States. A better explanation is Sanders supporter, 14% said no that partisans in both countries are exposed

148

to similar US media outlets, which PhD students, and British PhD students are interacts with their ideological biases. less biased than North American PhDs. Transatlantic differences in the Though each side discriminates against the propensity to discriminate are likewise other in fairly equal measure, the net effect evident with respect to right-versus-left of academia’s leftward skew is to produce hiring bias. Among North American PhDs, more discrimination against conservatives. 35% would discriminate against a This is especially the case among PhD conservative hire with a stronger track students where the net discrimination ratio record than a centrist while 14% would slants two and a half times against the discriminate against a leftist with a right. stronger record than a centrist. These patterns are summarized in Among British PhDs, 16% – a much Figure 106. Another differentiating factor smaller number – would discriminate between the two sides of the Atlantic may against the conservative hire compared to be a greater skittishness among Americans 7% against the leftist hire. Whereas 42% and Canadians about selecting candidates of North American leftists would with an open political skew – whether on discriminate against a better-qualified the right or left: North American conservative, just 20% of British leftist academics and PhD students are more PhDs would. likely than British equivalents to be biased Overall, when it comes to hiring, against a left-wing hire as well as a right- academics are more even-handed than wing hire.

Soft and Hard Discrimination in Hiring (North America vs. Britain) 70% 65% 3.00 2.5 60% 2.3 2.50 50% 47% 2.00 1.7 40% 40% 40% 1.3 35% 1.50

30% Ratio 16% 1.00 20% 13%

% Who Discriminate Who % 10% 10% 0.50 0% 0.00 North American British Academic North American British PhD Academic PhD Hard anti-Conservative Discrimination Soft Discrimination v Trump/Brexit Supporter Anti-Conservative/Anti-Left hiring ratio

Figure 106.

149

Is there Bias in Refereeing Grant and those on the right discriminating against Promotion Applications and Journal their left-leaning equivalents. Articles? Despite each ideological segment displaying similar levels of bias against the As in other studies, I find clear other, the structural effect of evidence of political discrimination in discrimination in the American academy grants and papers, with the playing field is, as in Britain, slanted against the right. strongly tilted against conservatives. Here Again, this is because of the leftward skew I ask, following previous studies, “If I was of the faculty. Thus the chance of a right- refereeing a grant application, I would be leaning paper facing bias overall among inclined to rate it lower if it took a American academics is 12% compared to politically right-wing perspective,” 5% for a left-leaning paper submission. On altering “right” to “left,” and “grant” to grants, 20% of right-leaning grants face “paper” in subsequent questions. Grants discrimination compared to 9% of left- often have just a 5-10% success rate and leaning grants. For promotions, the ratio is are hotly contested because they involve 12 to 6 (i.e., 2:1). In all cases, the right substantial rewards in the form of prestige faces at least twice as much structural and research resources, and a zero-sum discrimination. In addition, we saw from contest between individuals and our concealed list experiment that actual departments could be leading to an bias against a Trump supporter is nearly elevated level of bias among assessors. twice as high as what is stated in an open American results in Figure 107 question. show that 24% of leftist academics would If the actual rate of discrimination rate a right-leaning grant lower while just is double the amount openly admitted, 16% of right-wing academics would rate a then right-leaning papers, grants, and left-leaning grant lower. However, in promotion applications face a 24-40% terms of papers, right and left discriminate chance of discrimination from any given against each other at a similar rate (13- assessor. Multiplied across 4 panelists or 14%), and for promotion, right-wing assessors, this suggests that academics on academics are somewhat more likely to the right will face discrimination in the discriminate against the left than vice overwhelming majority of paper versa (16% vs. 13%). submissions, grant applications and Weighting for the twofold promotion bids. By contrast, the 5-9% concealment multiplier, this results in 26- overall rate of discrimination against the 48% of American left-wing academic staff left amounts to 10-18% chance of discriminating against a right-leaning discrimination from any one assessor promotion, grant, or paper and 26-32% of given the concealment multiplier.

150

Would Rate Paper, Grant or Promotion Lower, by Rater's Ideology (US)

30% 24% 25% 20% 20% 16% 16% 14% 14% 13% 15% 12% 12% 12% 9% 9% 9% 10% 7% 6% 5% 4% 5% 5% 0% Rate Left Rate Right Rate Left Rate Right Rate Left Rate Right Paper Lower Paper Lower Grant Lower Grant Lower Promotion Promotion Lower Lower All Left Center Right

Figure 107. Note: There are 6 response categories for the grant question, versus 3 for the paper and promotion questions, so these are not strictly comparable. Furthermore, anti-left bias is offset mark it down. This results in net positive by discrimination in favor of the left. For discrimination in favor of those on the left instance, 6% of American academics who signal their political beliefs. would rate a left-leaning paper lower, but In Britain, left-inclined groups 3% would rate it higher, because of its (Remain, Left, Labour) are between three leftist stance. On promotion applications, and five times more likely to rate right- 5% would rate a left-leaning promotion leaning grant applications lower than left- lower than if it did not adopt a leftist leaning grant applications. Right-leaners perspective, but 2% would rate it higher (Leavers and Tories) are between one and because of its leftist outlook. Right-leaning two times more likely to rate left-leaning papers and promotion applications are, by grants lower than right-leaning grants. contrast, only rated higher by 1% of This lower level of discrimination is partly academics, conferring little counterweight due to the fact that only 40% of Leave and to discrimination. Combining the US and half of Tory academics in my sample Canada, 5 of the 55 right-wing academics identify as right-wing. Right-wing in the sample (9% of the total) rate right- academics are six times more likely to rate leaning promotion applications lower left-leaning grants lower than right-leaning while 1 (2%) rates them higher. For grants. Even so, the highest level of papers, 4 rate right-wing papers lower discrimination reported by right-wing (7%) and just 3 higher (5%). Thus, even if reviewers is 18%, compared to 30% for one wanted to get the attention of right- left reviewers. Here I do find the left to be wing reviewers, there is no benefit to more biased than the right. signalling a right-wing orientation. A similar pattern can be discerned In Canada, the pro-left premium is for refereeing papers. Although the greater: 3% on papers, offsetting the 3% average level of bias is much lower here, who would mark it lower. For promotions, this may be due to the different question a left-leaning application gains points from wording for the grant and papers 4% of respondents while just 2% would questions. All told, these figures indicate

151

that while both sides discriminate, the left engage in political discrimination. does so at a higher level. This runs counter However, it is likely that the share who to findings for grants and papers in the would discriminate against a conservative- Freburg and Honeycutt study, but is in line oriented grant application is less than two- with the work of Peters et al. (2020). thirds because, first, the wording of the When weighted for the left’s substantial grant question offers two discrimination demographic advantage within the points (“strongly agree” + “tend to agree”) professoriate, the net result is a substantial that are amalgamated, whereas the paper anti-conservative slant. Across the entire and promotion questions, like that of sample, grants adopting a right-wing hiring, are phrased in a binary “rate it perspective are discriminated against by lower/rate it higher” manner.98 Second, as 22% of academics whereas those taking a the share of biased reviewers rises over left-wing stance only face bias from 9% of 50%, it runs into an increasingly resistant the professoriate. For journal articles, ceiling of meritocratic scholars. A realistic conservatives face discrimination twice estimate would suggest a multiplier of 2 (9%) as often as leftist authors (4.5%). rather than 3.2, thus 44% of reviewers These figures don’t seem high until I would discriminate against a politically consider my multiplier of 3.2, derived right-leaning grant. Nonetheless, this is an from the list experiment where just 10% of astounding figure. UK academics admitted they would Finally, on promotion applications discriminate against a Leaver, but the (see Figure 108), I see an intermediate revealed percentage in the list experiment pattern, with 21% of left-wing reviewers is 32. Applying this multiplier would willing to rate right-leaning promotion suggest that for any given academic who applications lower as compared to 7% of submits a manuscript for publication, there right-wing reviewers who would rate a is a 30% chance each reviewer will left-leaning application lower. Here again, downrate a right-leaning paper. with the multiplier, over 40% of left It would also suggest a close to academics assessing right-leaning two-thirds likelihood that each reviewer of promotion applications would discriminate a right-leaning grant application will against them.

152

Would Rate Paper, Grant or Promotion Lower, by Rater's Ideology (Britain)

35% 30% 30% 25% 21% 20% 18% 15% 15% 11% 10% 6% 7% 5% 3% 3% 3% 0% 0% 0% Rate Left Rate Right Rate Left Rate Right Rate Left Rate Right Paper Lower Paper Lower Grant Lower Grant Lower Promotion Promotion Lower Lower All Left Center Right

Figure 108. Note: There are 6 response categories for the grant question, versus 3 for the paper and promotion questions, so these are not strictly comparable.

A promotion application with a promotion applications lower if they bear a right-wing perspective will, on average, be political cast of any kind. ranked lower. Since there are typically 5- 10 referees (including the adjudication Discrimination by PhD Students panel) for a promotion application, there will be, on average, 2-5 voices in the room Doctoral students are more discriminating against a right-wing discriminatory than academics on these candidate. This may be mitigated to the measures. In North America, 24% of all extent that the applicant is able to name PhD students say they would rate a right- referees, but named referees make up a leaning paper lower, 30% would mark a minority of references – especially when right-leaning promotion application lower, the adjudication panel’s votes are and 33% would rank a right-leaning grant included. A paper is also unlikely to be application down. The level of bias in judged strictly on its merits since most favor of left-leaning papers is also greater journals require at least two referees plus than among academics. Ten percent of all an editor to take a look. This means there North American PhD candidates would is a 60-90% chance of a right-wing paper rate a left-leaning paper higher while 8% being rated lower. would rate it lower, producing a net While 4.4% of UK academic incentive for left partisans to signal their respondents would rank a left-wing beliefs in articles. For promotion promotion application or paper lower, applications, 10% say they would rank 3.3% would rank it higher. Thus it makes left-leaning ones higher while 10% say almost as much sense to advertise one’s they would rank them lower, producing no left-wing beliefs as to conceal them. net discrimination against an openly left- American academics stand out, as leaning promotion. compared to Britons and Canadians, for their willingness to rate papers and

153

In Britain, 9% of PhD students surveys. These figures should be would rate left-leaning papers higher while multiplied by a factor of up to two, though just 4% would rate them lower. For the level of social desirability bias in the promotions, left-leaning applications gain responses is, given the results of our list points from 9% of British PhD reviewers experiment, likely to be substantially and lose them from just 5% of them. higher among academics than PhD Twenty-eight percent of British PhDs students. This does not however take away would rate right-leaning grant applications from the fact that norms of non- lower, 22% would rank a right-leaning discrimination appear to weigh more promotion application lower, and 20% heavily among academics than PhD would mark a right-leaning paper down, candidates. Notice as well that grant bids, with almost none in the opposite direction, which involve substantial monetary leading to high net negative research resources, attract the highest level discrimination. of political discrimination across all Figure 109 summarizes the level of surveys. open anti-conservative bias across five

Openly Admitted Discrimination against Right-Leaning Papers, Grants and Promotions 35% 33% 30% 30% 28% 24% 24% 25% 22% 22% 20% 20% 20% 17% 14% 15% 12% 12% 11% 11% 11% 11% 9% 10% 5% 0% US Canadian UK UK N America UK PhDs academics academics academics academics PhDs (Yougov) (Mailout) Paper Submission Promotion Application Grant Bid

Figure 109. Combining six questions – each discriminatory. Bearing in mind sampling collapsed into a yes/no (coded 0/1) noise, these results seem to indicate that discrimination binary on hiring, academics of different ideological stripes promotion, and refereeing – permits me to discriminate against each other fairly examine which factors predict evenly, as per Honeycutt and Freberg discrimination against left and right. (2017). Controlling for age, gender, SSH or From the evidence, it is unclear STEM, and identifying as an activist, there whether theories about the left’s lack of is significantly greater discrimination from contact with conservatives leading to left academics in Britain and “very right” misperceptions of the out-group, or the academics in the US, with no ideological right’s heightened consciousness about group in Canada standing out as more being an embattled minority, play the

154

bigger part in motivating discrimination. when comparing academics with non- Since both seem to discriminate against academics where I find that ideology and each other at similar rates, a third activism predict political discrimination possibility arises: it may be that partisan just as much among non-academics as bias runs deeply on both sides, regardless academics; even though the work context of contact or professional context. Some in other industries is more politically evidence for this interpretation will follow balanced and less politicized.

Who Discriminates?

Beyond ideology, are any other SSH academics and PhDs don’t factors associated with higher political discriminate more than their STEM discrimination? Running relatively similar colleagues once you control for their more models of discrimination (based on a 5-6 left-wing ideology and higher activism. item composite index covering hiring, It’s also worth noting that in models of promotion, grants, and publication) across discrimination against the left, the only five datasets in Figure 110, what comes variable that matters is being right-wing across is how dominant ideology is when when it comes to identifying who will picking out those who discriminate. discriminate. Other factors, such as age or Compared to centrists, very left academics SSH vs. STEM, do not reach conventional and PhD students discriminate the most, measures of significance. followed by those who are fairly left. Older academics are marginally Once we control for ideology, less likely to discriminate than the young, being an activist only predicts higher but this is only statistically significant in discrimination for the North American one model. In other words, net of academic survey, though the coefficient is ideological differences, we should not in the expected direction in all but one expect more discrimination against survey. Women and minorities are conservatives in the future except insofar significantly more likely to discriminate as recruitment to the profession pushes against conservatives when we control for academia further left than it already is ideology, but only among American and today. Canadian academics.

155

Predictors of Discrimination against Conservatives 0.400

0.300

0.200

0.100

0.000 Standardized Effect Size Effect Standardized

-0.100

-0.200

N America UK Yougov UK Mailout N America PhD UK PhD

Figure 110. Note: For Very Left, all results are significant at the p < .001 threshold. Others that are significant at the p<.05 or lower threshold are Activist, Female, and Nonwhite in the North American Sample, and Older in the UK PhD sample. Outcome measure combines 5-6 binary variables.

Soft Authoritarianism: Chilling Effects of leaning to make it prudent for Discrimination conservatives to hide their views. This substantiates with data the repeated The combination of individual- testimony that there is a climate of level discrimination by both sides with a political discrimination inside the heavy left-leaning majority produces a contemporary university. If conservative high degree of system-level bias against academics wish to have papers accepted conservatives. These findings mean that for publication, to be awarded grants, or to conservative scholars who self-censor are be promoted, it is wise for them to conceal not paranoid, but acting rationally. A their political views. sufficiently large proportion of academics are willing to penalize work that is right-

156

Summary of Discrimination Effects professors might fear that scholarship that appears To summarize the discrimination conservative may be rejected by effects, I constructed an index of seven left-leaning law review editors, and discrimination-related questions for the disparaged or ignored by their data, with four questions on hiring, and colleagues, which will damage one each on refereeing a grant application, their chances for promotions, promotion case, or paper. In the United research money, and lateral States, 74% of SSH academics would appointments. This would explain discriminate against the right, why even [registered Republicans] conservatives, or Trump supporters on at tilt left. Republicans could least one dimension. For Canada, the suppress their ideological views by equivalent figure is 75%. avoiding controversial topics, Discrimination is about 20 points taking refuge in fields that have lower in Britain. Using the UK YouGov little ideological valence, focusing data pertaining to Leavers or on empirical or analytical work, or Conservatives, 44% of the sample would simply writing things they don’t discriminate on at least one of the seven believe.99 questions. This rises to 54% of currently- employed academics in the social sciences The data from my study indicate and humanities, 67% of far-left academics, that a similar dynamic is at play within the and 74% of currently-employed far-left social sciences and humanities more SSH academics. The UK SSH mailout broadly, in both North America and survey shows a slightly lower share who Britain. would discriminate on any dimension: Political discrimination greatly 42% of SSH academics and 66% of far- impoverishes scholarship and teaching by left academics. narrowing the range of research questions With discrimination this pervasive, that are asked, funded, and pursued, and it is little wonder that conservative circumscribing the number of acceptable scholars tend to avoid writing papers that interpretations of the facts. This is signal a conservative perspective, while especially true of research areas that touch left-leaning authors tend to openly on sacred progressive values around race, advocate theirs. A recent study based on a gender, or sexuality. For instance, there random sample of American legal has been comparatively little work on left- scholars, inferring political leanings from wing authoritarianism in social political donations, found that among the psychology, compared to the wealth of small minority of Republican-supporting scholarship on right-wing academics, work could not generally be authoritarianism. The reasons are fairly assigned a political coloring by coders. By clear. Most scholars are progressive and contrast, much of the output of Democrat thus have little interest in examining left- supporters was easily identifiable as wing forms of discrimination and progressive. The authors go on to suggest authoritarianism. Second, investigations of that: left-wing bias or authoritarianism are likely to face more resistance – whether in The most plausible explanation is grant applications or paper submissions – that if the dominant ethos in the top than those focusing on right-wing law schools is liberal or left-wing, forms.100 then Republicans are likely to Collegiality is strongly encouraged conceal their ideological views in in academic departments, and provides their writings. Republican important career and social benefits to

157

academics, including the minority of right- progressives do. This lack of intergroup leaning scholars. Cass Sunstein argues that contact has been shown in other situations the more of a social character an to permit misperceptions to flourish institution or field has, the stronger its unchecked.102 mechanisms of conformity to group norms One measure of social distance is will be, and the lower its performance. In whether an academic would feel such an environment, Sunstein explains comfortable sitting next to someone with a why self-censorship occurs: “Group different view at lunch, be neutral towards members who care about one another’s the idea or not feel comfortable. Eating approval, or who depend on one another lunch together is a common form of for material or nonmaterial benefits, might collegial interaction among faculty and well suppress highly relevant information thus an important barometer of how non- [for organizational functioning].”101 This material factors can alter the attractiveness applies in academia, as it comes to have of a workplace. more of a confessional hue, and one which Figure 111 shows that just 41% of is defined on the basis of political outlook. 2016 Democratic voters would feel Collegiality has its benefits, but one of its comfortable sitting next to a Trump voter, downsides is that it fosters the social 26% said they would be uncomfortable conditions that can sustain a culture of and 30% “neutral.” Trump and Clinton conformity, rather than one that voters are similarly comfortable (61% vs. encourages dissent. 65%) sitting next to a Sanders supporter. Less than half of Trump supporters (48%) Social Aversion to Conservatives and and Clinton voters (25%) said they would Gender-Critical Scholars feel comfortable sitting with a “known proponent of the idea that trans women Political discrimination is not should not be admitted into women’s simply an abstract sentiment, but is rooted refuge centers,” with 21% of Trump in the emotions. One of the “non-material” supporters and 50% of Clinton supporters benefits of collegiality, for instance, is a saying they would not be. pleasant experience at the workplace, Canadian data are similar, with which contributes to well-being. Those 30% on the left uncomfortable with a who politically discriminate, however, are Trump supporter and 36% comfortable. often uncomfortable socializing with those Fifty-four percent of Canadian academics of other political backgrounds. This is said they would be uncomfortable having especially true among progressives, likely lunch with a gender-critical feminist because, as the More in Common report colleague (i.e., who opposes trans women suggests, at high levels of education, they accessing women’s shelters). Here is have far less social interaction with evidence that gender-critical feminists may conservatives than less-educated face even greater levels of discrimination than conservatives and Leavers.

158

Comfortable Sitting Next to Someone of Opposing View at Lunch? (USA) 100% 90% 79% 80% 66% 70% 61% 60% 48% 50% 41% 40% 30% 25% 20% 10% 0% Comfort Sitting Comfort Sitting Comfort Sitting Comfort Sitting Comfort Sitting Comfort Sitting with Trumper with Trumper with Sanders- with Sanders- with Trans- with Trans- (Trumper) (Clintonite) ite (Trumper) ite (Clintonite) Critical Critical (Trumper) (Clintonite)

Figure 111.

The results for the British YouGov admitted into women’s refuge centers,” data are very similar. Figure 112 shows with 30% of Remainers and 12% of that under half of Remain voters would Leavers saying they would be feel comfortable sitting next to a Leaver; uncomfortable sitting next to such a 16% said they would be uncomfortable person. For the UK mailout sample, the and 36% “neutral.” Leavers and corresponding Figures are 12% Remainers are similarly comfortable (60% uncomfortable with a Leaver, 11% with vs. 64%) sitting next to supporter of far- Corbynite, and 30% with the gender- left politician Jeremy Corbyn. Less than critical scholar. Among this group, 54% half of Leavers (49%) and Remainers said they would comfortably lunch with a (34%) said they would feel comfortable Leaver compared to 58% for a Corbynite. sitting with a “known proponent of the This indicates that intra-left animus is also idea that trans women should not be important within academia.

159

Comfortable Sitting Next to Someone of Opposing View at Lunch? (Britain) 100 90 81% 80 64% 70 60% 60 48% 49% 50 40 35%

% Comfortable % 30 20 10 0 Comfortable Comfortable Comfortable Comfortable Comfortable Comfort sitting sitting with sitting with sitting with sitting with sitting with with Trans- Leaver (Leave) Leaver Corbynite Corbynite Trans-Critical Critical (Remain) (Leave) (Remain) (Leave) (Remain)

Figure 112. Across all partisan identities, of the populist left than those of the among North American PhD students, populist right. The anti-populist right skew 39% would be uncomfortable sitting next seems somewhat more marked in North to a Trump supporter, 9% uncomfortable America than in Britain. Though there are with a Sanders supporter, and 59% differences, the positive news is that just uncomfortable with the gender-critical 12-30% of leftist academics said they colleague. For UK PhDs the corresponding would feel distinctly uncomfortable having discomfort figures are 11% with the right lunch with a populist right supporter, even populist (Leave) supporter, 18% for the if only a minority declared themselves far-left supporter, and 54% for the gender- “comfortable” doing so and many gave a critical feminist, suggesting no bias against lukewarm “neutral” response to the idea. right populists. However, in all cases, the American and Canadian SSH left- share saying they would actively feel wing and right-wing academics in Figure comfortable with the idea of sitting next to 113 differ by around 50 points in their a populist left supporter was higher than comfort with the idea of lunching with a the share comfortable sitting with a Trump supporter when I control for age, populist right supporter. gender, and race. However, there is no Overall, gender-critical researchers statistically significant difference between appear to face the highest levels of social SSH and STEM faculty, or between discrimination. There is also a consistent American and Canadian academics, in transatlantic pattern of scholars being their comfort levels.103 more comfortable dining with supporters

160

Figure 113. Note: sample too limited to test “very right” category, which has been combined into the rightmost category. N=1,084. Pseudo-R2=.046.

Looking at the British YouGov somewhat smaller gap than in North results by left-right ideology and America. Nevertheless, there is a lot of controlling for age, gender, professorial variation in comfort level within the left, rank, and SSH vs. STEM in Figure 114 with just under half of leftists saying they shows that those on the left are around 40 would be comfortable sitting with a points less comfortable sitting next to a Leaver. Leaver than academics on the right, a

161

Figure 114. Pseudo-R2=.053. N=820.

There is no obligation to associate into one index with a score ranging from with those one has little in common with. 0, indicating discrimination on no However, there is an important difference dimensions, to 1, discrimination on all 7 between freely associating with like- dimensions. Though ideology is a major minded people and actively excluding cause of political discrimination, even those who are different. In a collegial when I hold this constant, the degree of context, active avoidance of a minority, as comfort with a Trumper/Leaver remains distinct from choosing to associate more extremely important. often with those in one’s in-group, I noted that those on the left split constitutes a form of political fairly evenly between half who are discrimination that can produce important comfortable sitting next to a Trump/Leave chilling and conformity effects. supporter and half who are neutral or Figure 114 showed that ideology uncomfortable doing so. This variation is strongly conditions whether someone is important in separating biased from comfortable sitting next to a Leaver. Does unbiased leftist academics. Thus, even this matter for academic freedom? Yes. To among faculty on the left (the red upper understand why, consider what we noticed line in Figure 115), those who are most in Figures 111 and 112 concerning comfortable sitting next to their comfort at the idea of having lunch with ideological opposites have less than a 50% someone with a different view. The model chance of discriminating at least once underlying Figure 115 asks what predicts against a Leaver, Friends of Israel discrimination against Trump/Leaver supporter, or right-winger in a job, or supporters or conservatives across seven rating a right-leaning article, grant, or questions I have encountered, compiled promotion application lower. Leftists who

162

are not comfortable sitting next to a more important than ideology for Leaver have a greater than 90% likelihood predicting whether someone will of discriminating against them on at least politically discriminate. The academy is one of the seven aforementioned much better off with tolerant leftists than dimensions. In effect, social tolerance is intolerant centrists.

Figure 115. Pseudo-R2=.146, N=820. Very comfortable, as well as leftist, are significant at the p<.001 level, and “very uncomfortable” and “fairly uncomfortable” at p<.01. The pattern is relatively similar in uncomfortable sitting with a Trump North America in Figure 116, albeit at a supporter. Ideology in the US and Canada higher level of intolerance, with around 6 matters almost as much as social comfort in 10 leftists who are comfortable sitting for predicting political discrimination next to a Trump supporter discriminating against the right whereas in Britain against the right on at least one dimension, ideology matters much less than social rising to 85% among those who are comfort.

163

Figure 116. Pseudo-R2=.154. Fairly uncomfortable, as well as leftist, are significant at the p<.001 level, and “very uncomfortable” at p<.05.

Discrimination: Summary or qualifications. The fact that the share who revealed they would discriminate in Political discrimination exists hiring is 2 or 3 times as large as the across society, as several studies of the number who directly admitted to this in an off-campus environment show. However, unconcealed question suggests that the specific problem in the social science committee chairs can highlight the need to and humanities departments of universities avoid political discrimination to panel is that the balance of prejudices is highly members and be on the lookout for lopsided due to the ideological makeup of concealed forms of political bias. staff, and professors’ beliefs are unusually visible in their work compared to other Toward Soft Authoritarianism sectors of society. The key, therefore, lies in These results indicate that a large awareness. If the fair-minded majority of plurality of left-wing and some centrist academics are able to enforce a norm of academics discriminate against non-discrimination on the discriminatory conservatives in aspects of academic life, 30-40%, then structural discrimination can from hiring to promotion to refereeing. be reduced. This would need to encompass This contributes to a culture of self- alertness to biased panellists’ tendency to censorship, hampering conservatives’ tolerate a “dog-whistle” form of freedom to disseminate the full range of discrimination, in which political their ideas in research and teaching, objections are couched in plausibly limiting the chance to bridge society’s deniable language or excessive objections main axis of political antagonism while to other aspects of candidate performance constricting policy solutions. Political

164

discrimination – acting as much through a into the harder “cancel culture” politics of chilling effect and the perpetuation of institutional sanctions or dismissal actions. stereotypes of academia as a leftist The results below show that softer preserve as through overt rejection – may and harder forms of authoritarianism are also help explain why so few connected because “hard” restriction of conservatives choose a career in academia. academic freedom in the name of This distorts the production of knowledge protecting disadvantaged groups is and the effectiveness of the university. connected to “soft” political Notwithstanding this injustice, might a discrimination. This emerges clearly in the thick-skinned conservative manage to analysis in Figure 110 where I modeled survive in academia? Not necessarily. the predictors of discrimination – with this outcome measured through an index of Political Discrimination and Illiberalism political discrimination aggregating the in Academia results of seven questions tapping a willingness to discriminate against Leave Part I looked extensively at support voters or those on the right. for hard authoritarian measures such as Recall that in Part I, we saw that dismissal campaigns or mandatory reading younger, far-left, and activist academics list quotas that restrict academic freedom. are more likely to endorse a campaign to Is there any connection to Part II of this fire a controversial scholar. When political report on the soft authoritarianism caused discrimination is added to the model in by political discrimination? Absolutely. Figure 117 predicting which academics The kind of person that would discriminate will support one of four hypothetical against a conservative entering into, or campaigns to dismiss controversial succeeding in, academia is far more likely scholars from their posts, I find that the to be the kind of person that will endorse willingness to discriminate against those measures to push them out. on the right jumps out as by far the Conservative and gender-critical strongest correlate of illiberalism. This academics are keenly aware of political greatly improves the model’s ability to discrimination, and take steps to self- predict which academics will support a censor their research and teaching to avoid push to fire conservative or Leave-voting drawing unwelcome attention to dissenters. Here we see clear evidence that themselves in a left-dominated soft authoritarianism (i.e., discriminating) environment. Here Mill’s “despotism of serves as a gateway to hard custom,” which chills expression, shades authoritarianism (i.e., backing a firing campaign).

165

Predictors of Supporting a Dismissal Campaign (US and Canada)

Political Discrimination v. Right*** Younger*** Very left*** Activist*** Minority** Female+ SSH Canadian

0.000 0.050 0.100 0.150 0.200 0.250

standardized effect

Figure 117. Note: R2=.179, N=1,093. Reports standardized beta coefficients. Significance at +p<.1, *p<.05, **p<.01, ***p<.001.

The pattern in Britain (YouGov In both North America and Britain, data) is similar, with political younger academics are more willing to discrimination the strongest predictor of endorse cancelling controversial staff, being willing to endorse one of the four even when political discrimination and dismissal campaigns. This finding holds ideology are held constant. Looking ahead up across all five surveys, with political to the future of academic freedom, this discrimination significant among both may be viewed as a concerning academics and PhD students. development.

166

Predictors of Supporting a Dismissal Campaign (Britain)

Political Discrimination v. Right*** Very Left*** Age (younger)** Ethnic minority+ Political Attention Activist Current Staff Soc Sci/Humanities Female

0.000 0.050 0.100 0.150 0.200 0.250 Standardized Effect

Figure 118. Note: R2=.137, N=820. Reports standardized beta coefficients. Significance at +p<.1, *p<.05, **p<.01, ***p<.001. Being a political discriminator in Those on the far left who are the social sciences and humanities has a willing to discriminate are also somewhat elevated effect on a person’s considerably more illiberal than far leftists willingness to endorse the sacking of a who are unwilling to engage in political fellow academic as compared to STEM discrimination. Figure 119 shows that fields. Discrimination is connected to hard those on the far left who are not willing to authoritarianism in both parts of academia, discriminate against a right/Trump but matters more for hard authoritarianism application, paper, or hire are around 25 in the social sciences and humanities in the points less likely to endorse a campaign to US and UK (but not Canada). Despite this fire another academic. Among far leftists relationship, it is vital to bear in mind that who would discriminate, 45% would political discrimination, which contributes endorse at least one authoritarian measure to a chilling effect that stifles academic against conservative academic dissenters. freedom, is considerably more widespread The elevated relationship among far leftist than hard authoritarianism. As a result, discriminators holds independently in both some 70% or more of academics who the American and Canadian academic would politically discriminate do not surveys. support dismissal campaigns.

167

Figure 119. R2=.152, N=1,093. Far Left x political discrimination interaction is significant at p<.01 level.

Likewise, in Britain, far-left fire another academic. Figure 120 shows academics who are unwilling to that among far leftists who would discriminate against a right/Leave discriminate, nearly 3 in 10 would endorse application, paper, or hire are around 20 authoritarian measures against points less likely to endorse a campaign to conservative academic dissenters.104

168

Figure 120. R2=.152. Far Left x political discrimination interaction is significant at p<.01 level.

distinctively hostile environments for Part III Off-Campus conservatives.

Is Academia Different? Comparison with Non-Academics

Earlier we saw that while As I will show, academics may not university professors and lecturers are be more likely to politically discriminate more liberal than professionals in other than other professionals. Indeed, a well- sectors, the political gap is smaller if known American study asking 1000 compared to degree-holders outside people to judge applications for a academia, especially advanced degree- scholarship – where CVs that were holders. Perhaps, therefore, a chilling otherwise identical variously stated the effect on free speech, political applicant had been president of a Young discrimination, and anti-conservative Democrats or Young Republicans club – authoritarianism are parts of professional found that respondents were biased toward life and high culture in elite institutions their own party 80% of the time, more generally. As Andrew Sullivan regardless of merit, but cared little about writes, “We all live on campus now.”105 race. As the authors note, “Political As we’ll see, there is a great deal of truth identity is fair game for hatred. Racial to this hypothesis. Yet it is equally identity is not. Gender identity is not. You apparent that universities stand out as cannot express negative sentiments about social groups in this day and age. But

169

political identities are not protected by race, gender, and sexuality, about 30-35% these constraints. A Republican is of Republican students are reluctant to someone who chooses to be Republican, share their views in class compared to 15- so I can say whatever I want about 25% for Democrat students. While these them.”106 numbers show a substantial chilling effect, A major Cato Institute YouGov they indicate that right-leaning students survey showed that 6 in 10 employees are somewhat less inhibited in expressing with master’s or doctoral degrees who their views than right-leaning academic support the Republicans say that they “are staff. worried about losing [my] job or missing Of students who feared speaking out on job opportunities if [my] political up, 61% worried about how their peers opinions became known.” Almost half of would judge them and nearly half were independents with postgraduate degrees concerned about what their professor agreed, compared with 25% for would think (not entirely an unfounded Democrats. Meanwhile, 88% of Trump- concern in the US context given that 7- voting degree-holders compared to just 15% of academics we asked revealed they 44% of Clinton-supporting degree-holders would mark down a right-leaning term agreed that “The political climate these paper). Forty-three percent felt that their days prevents me from saying things I views might wind up on social media and believe because others might find them nearly 40% worried that a complaint may offensive.” On the perpetrator side, 50% of be lodged against them.109 Here we see “strong liberals” said a business executive that both soft forms of authoritarianism who was found to be donating to the based on peer pressure and hard forms Trump campaign should be fired. Thirty- involving fears of being disciplined work six percent of “strong conservatives” said to produce a chilling effect. the same for an individual donating to the Biden campaign.107 Non-Academic Comparator Study Moreover, it goes without saying that students also feel the chilling winds of To compare my academic sample Mill’s “despotism of custom.” In a with professionals outside the sector on previous Policy Exchange report utilizing many of the same questions used in the a sample of over 500 British students, Tom academic survey, I used Prolific to sample Simpson and I found that only 4 in 10 867 degree-holding employees over age 23 Leave-supporting students said they would residing in Britain.110 The average age in be willing to express their views in class. the sample was 39, with a median Students also split fairly evenly into a household income of around £45,000, with group that supported no-platforming two-thirds homeowners and around 72% controversial writers Jordan Peterson and married, living as married, or formerly Germaine Greer, a group that opposed married. Just 15 are university or college these actions, and an intermediate, lecturers or professors, though some 97 biddable group. Social media rather than work at universities, permitting us to lecturers were identified as the source of compare them with other sectors. Survey views on cancel culture and free speech 108 platform samples tend to lean left and be questions. tech-savvy, and ours is no exception. American research shows that 55% Eleven percent of respondents identified as of students feel that the “campus climate right-wing, 46% as left-wing, and the prevents me saying things I believe.” Fully remainder centrist or unstated. Twenty-one 82% of conservative students said they had percent voted Leave, 73% Remain, with self-censored at least once in class, 6% not voting. This puts the sample of compared to 40% of liberals. On politics, degree-holding non-academics about 5

170

points to the right of my academic sample, Non-Academic Workplaces Are Less but well to the left of the degree-holding Hostile to Expressive Freedom public. Thus, when asked about the views Most conservative and Leave- of the median staff member at their voting professionals experience workplace, the non-academic sample considerably less apprehension in their replied 36% left and 15% right, a 2:1 work environments than their counterparts left:right slant compared to the 9:1 ratio in in academia. Fifty-four percent of my UK YouGov academic survey. These respondents felt a Leaver would feel non-academic professional workplaces are comfortable expressing their views at the clearly more centrist than universities. workplace, with only 16% saying such a Limiting ourselves to the small person would not feel comfortable doing subset of Leave voters in the non- so. Among Leavers, the corresponding academic survey, I find they report a 1.8:1 figures were 56% and 13% – no real left-right balance in their work difference. By contrast, we saw in Figure environment. This figure among non- 90 that, among academics, just 37% of academic Leavers compares to a 7.4:1 current academics of any political stripe ratio reported by academic Leavers (in my think a Leaver would feel comfortable academic survey) when describing their expressing their views, with just 28% of work colleagues’ median ideology. In the academic Leavers saying so. Among social Prolific non-academic data, the subsample science and humanities academics, only of 97 university employees (mostly staff 18% of Leavers felt they would be rather than academics) report, on average, comfortable doing so. This indicates that a 4.8:1 left-right ratio. This is clearly university-educated Leavers feel 2-3 times different from the 2:1 left-right ratio freer to express their views when working reported by those working outside the outside of academia. university sector. The latter seems to be an This is reflected in the views of the accurate representation of non-academic 97 people working in the university sector professional workplaces: YouGov in my non-academic dataset – researchers, Profiles’ 56,000-person representative administrative staff, and around 13 sample shows a 2.1:1 left-right ratio academics. Only 22% said a Leaver would among degree-holders working in non- feel comfortable expressing their views on university professional sectors, which campus, significantly lower than in any of matches that of my Prolific non-academic the other 8 industry categories I recorded. sample. Universities, then, lie considerably In factories, there was no difference in to the left of other professional sectors. comfort level between Leavers and These figures reinforce the picture in Remainers when it came to expressing Figures 49, 50, and 51, that find, using their opinions. In all other sectors, Figure YouGov’s much larger sample of 121 shows that 51-59% of respondents felt credentialed professionals, that British that Leavers would be comfortable sharing academics place significantly to the left of their views and that 76-90% of Remainers professionals in other sectors – even would as well. While there is a chilling though university graduates in most major effect for Leavers in all white-collar professions also lean left and Remain. workspaces, it is two to three times larger in universities.

171

Would Leaver/Remainer Be Comfortable Expressing View? (by Sector, Britain)

100% 90% 87% 88% 84% 81% 81% 90% 76% 77% 77% 78% 80% 71% 70% 59% 58% 57% 60% 60% 52% 51% 54% 50% 40% 30% 22% 20% 10% 0%

Leaver comfortable Remainer comfortable

Figure 121. Note: Limited to degree-holders only. Sample size for number of respondents in each sector provided in parentheses.

The share of conservatives and than professional workplaces outside the Leavers who feel they work in a hostile university. environment is also considerably higher Figure 122 shows that outside of inside academia. Seventeen percent of universities, there is only a modest 8-point non-academic Leavers and 25% of non- difference between Remainers (.08 academics who identify as right of center probability) and Leavers (.16 probability) say their workplaces are at least somewhat in the likelihood of a staff member saying hostile toward people with their political their workplace is a hostile environment beliefs. Among current academic staff, the for their political beliefs. However, among equivalent figures are 43 and 47%, rising the 97 university employees in the sample, to 50% of Leavers and 64% of the gap between Leavers and Remainers conservatives in the social sciences and explodes in size to 54 points, with 4% of humanities. Here again, I find that Remainers, but 58% of Leavers, reporting academic environments are two to three a hostile work environment. times more hostile toward conservatives

172

Figure 122. Pseudo-R2=.06. Leave x University interaction is significant at p<.01 level. When it comes to people reporting mentioned by many, with those working in self-censorship of their political beliefs, politics, law, and nursing saying they were the difference between academic and non- prohibited by codes of conduct from doing academic environments is even starker. so: Indeed, I couldn’t find any difference in the prevalence of self-censorship between “Avoid discussing politics at work as it those on the left and right, or between causes arguments and hostility. Everyone Remainers and Leavers, in my non- is entitled to their views.” academic data. The substantially less politically “Letter to all staff informing them not to hostile character of non-academic bring politics into the staff room or face compared to academic work environments disciplinary. This was in regard to the strongly emerges in the comments 2014 referendum.” provided by non-academics. There were only a handful of people who said that “I have to remain politically neutral in the expressing their political views could workplace due to working for a affect their career. Among those who government agency. I am often told by refrained from talking about politics, most communications on how I should limit my did so out of a mixture of courtesy, political expressions in and out of work.” believing it to be unprofessional to discuss it, or not wanting to cause arguments with “Yes because the nursing code forbids us co-workers. At least a third of the to. I try to steer clear of political comments referred to conflict and discussion at work.” argument as a reason not to talk politics. An official requirement for neutrality was

173

“I refrained from airing views as not “Yes in Northern Ireland it is a difficult professional and could cause tensions.” topic for a number of reasons. You have to be very careful as political opinions can be “Yes, not within my office with immediate interpreted as sectarian. It has always been colleagues but with clients as I work with like this.” a wide range of people with whom I would not discuss politics.” “Have refrained from discussing the Scotland Independence vote but for There were also a broader mix of personal reasons not related to political considerations outside academia, progression.” suggesting a more varied ideological environment: “I have avoided discussing Scottish Independence as this causes tensions with “My boss is very right wing and brings up some colleagues, including those above politics a lot. A lot of the things he says I me in the organization.” don’t agree with, but I keep quiet as I don’t want to start a political debate.” “It is important in my role that I appear neutral in my political opinions. I manage “I am a leave voter and all my friends are my social media channels and have had to remain socialists. I find that they do not bite my tongue on numerous occasions listen to any other point of view except when people have posted views very their own. I have been bullied. My friends different to my own on subjects like will post anti-Brexit on social media Brexit, welfare and immigration. Within calling Brexiteers thick stupid uneducated. the office I can be fairly free about my A friend said to my face that leavers did left-wing politics but at board level I must not know what they were voting for, that be careful as some members of the board pressed my button. Friends have labelled are strong Tory supporters.” me indirectly on their social media. I post nothing political that would upset It is also interesting to compare them…the whole referendum and elections partisans’ comfort with the idea of having has been very divisive and I no longer lunch with a member of the opposing respect my friends since they have not camp. Forty-four percent of Remain- respected me.” voting non-academics would be comfortable lunching with a Leaver, “The Headteacher at my school is a Tory slightly less than the 48% of academics Brexiteer – he knows I am the polar who said they would be comfortable doing opposite, but my conversations are so. Forty-six percent of Tory voters governed by the power dynamic which outside academia would be comfortable allows him to pour scorn on my views, sitting down with a Corbyn supporter, whilst I have to be more respectful about which is 10 points less collegial than in his.” academe, where 56% of Tories said they would be comfortable having lunch with a “Yes frequently, as I work in a political Corbyn supporter. These figures indicate environment for right wing politicians. that intergroup comfort is somewhat lower Although not in a politically restricted outside than inside academia. One factor role, those of us who have more liberal might be the larger pool of Corbyn views are often referred to as ‘lefties’ by supporters among university faculty, managers with scorn.” which may expose Tory academics to greater intergroup contact and an ease with

174

the far left that is less common in other left-wing employees outside academia sectors. were willing to rate the promotion application of a known Conservative voter Willingness to Discriminate lower, which is somewhat below the 21% of left-wing academics willing to rate a Non-academics reported a promotion application lower. somewhat lower willingness to The results of my list experiment discriminate than the academics in my are presented in Figure 123. This again previous survey. In a direct question, 9% reveals considerable concealment of true of Remain voters outside academia were intent, and a much larger number of people willing to discriminate against a known willing to discriminate, as in the academic Leaver in a job application compared to survey. Indeed, 25% of non-academics 12% among Remain-voting professors and were revealed to be willing to discriminate lecturers in my earlier university survey. against a known Leave supporter, about Sixteen percent of university-educated four times the 6% who were willing to admit to this in a direct question.

Revealed Share That Would Discriminate against a Leaver for a Job (Non-Academics, Britain) 60% 55% 50% 40% 37% 31% 25% 26% 25% 27% 30% 22% 20% 9% 10% 5% 6% 0%

Figure 123. N=686 (329 for 4-list and 353 for 3-list).

Figure 124 compares these results on the academic survey. The similarities with those for my academic survey.111 It with academia overshadow the differences shows that the overall share in the non- – indeed, by conventional measures of academic survey willing to avoid hiring a significance I am not certain that the Leaver was 25%, compared to 33% in the difference between university and non- academic survey of professors and university results has not been produced lecturers. Among Remainers, the Figure by chance. This also echoes work cited was 26%, compared to 37% in the earlier from the US showing a widespread academic survey. There was no gender general tendency in the population to split, but a majority of self-described politically discriminate.112 activists were willing to discriminate, as

175

Revealed Willingness to Discriminate against a Leaver (Britain) 70% 58% 60% 55%

50% 44% 39% 37% 40% 37% 35% 32% 31% 30% 25% 26% 25% 27% 22% 20% 20% 20% 9% 7% 10% 5% 6% 1% 3% 0%

Academics Non-Academics

Figure 124. Source: YouGov Academic and Prolific Non-Academic surveys, 2020. Note: N=424 for 4-list treatment in YouGov sample and 329 for 4-list in Prolific non-academic sample. 3-list sample is 396 in YouGov, 353 in Prolific non-academic. Outside academia, there is also The left/Remain skew of these data evidence that discrimination against the means there is structurally more left and right is more evenly matched. For discrimination against Leavers (6.3%) than instance, 12% of Leavers and 16% of Corbynites (4.5%). Yet my sample of Conservatives said they would graduate non-academic workers is almost discriminate against a Corbynite for a job, certainly unrepresentative, given what we whereas just 9% of Remainers and 6% of know from the YouGov industry data on Labour voters outside academia said they the Brexit vote visited earlier, and from the would discriminate against a known way my non-academic respondents Leave-supporting applicant. Compared to describe the typical ideology of their results from my academic survey, workmates. Had my sample been discrimination by the right against the left representative of the political balance in is similar, but discrimination by the left non-academic organizations, I would against the right is 25-50% lower. likely find a similar degree of Moving to the list experiment, I discrimination against left and right, a very also asked a quarter of the sample if they different picture from the situation in would discriminate against a Corbyn universities, where the structural terrain is supporter in a job, and found an overall tilted heavily against conservatives and level of 20%. Meanwhile, 25% said they Leavers. would speak against a Conservative Discrimination rates may be applying for promotion, similar to the 25% slightly higher in academia than outside it, who said they would discriminate against a but this alone is unlikely to explain the Leaver’s job application. The net result dramatically greater hostility that seems to be only a marginally-lower anti- conservatives perceive in the academy. To left bias than anti-right bias. do so, it is necessary to account for the

176

higher visibility of people’s political Brexit views of over half of those they beliefs on campus – notably in the social worked with, compared to 54% among the sciences and humanities where this is 97 staff in the sample working at a manifest in the content of people’s work. university (many in support or research While I did not ask these questions on my roles). The visibility of Brexit beliefs by academic survey, I can compare campus sector is given in Figure 125. Universities and non-campus environments in my non- have significantly higher Brexit belief academic survey. For instance, off- transparency than other sectors. campus, just 27% said they knew the

Share Who Know the Brexit Vote of over Half Their Colleagues (%), by Sector 60% 54% 50% 40% 36% 33% 30% 33%

% 30% 25% 20% 23% 20% 16% 10% 0%

Figure 125. N=867.

Differences were less apparent scrutinized by numerous members of the when it came to being able to discern department during longlisting and Brexit credentials in job applications, shortlisting, at formal job presentations to though those working at a university were the department, and during interviews. three times more likely (6.7%) to say they Given the more politically- could often tell someone’s Brexit position balanced work environment outside from their application than those outside academia, political discrimination does not academia (2.8%). This was not statistically have as great a structural impact as bias significant, however. does within the university sector. That is, While I did not ask the individual-level biases in opposing transparency question in my academic directions cancel each other out to a survey, it is reasonable to surmise that greater extent outside academia, but not political belief transparency is higher inside it. This was also apparent in the among professors and lecturers in the qualitative comments, where far fewer social sciences and humanities. In this part people outside academia feared that their of the university, competition for jobs is politics could have repercussions on their particularly fierce. Reputations travel, and careers. the content of applicants’ public work is

177

This said, there is already a 2:1 member know that they should find work left-to-right tilt among graduates in non- elsewhere?” Use of the term “student” is academic professional settings in Britain, ambiguous here, but was used to try and and there was around 20 points more self- match a question on the academic survey. censorship of Leave than Remain views Between 7 and 8% of respondents among non-academics in our sample. endorsed these efforts, rising to 8-11% of YouGov Profiles data show that the left- left-wing respondents, and 17-20% of to-right ratio is about 2.5:1 among those who identify as far left. The professionals under 24, 2.3:1 for those corresponding figures among academics aged 25-39, 1.4:1 for the 40-59s, and close (from the UK YouGov academic survey) to 1:1 among those over 60. In terms of are 6%, rising to 8% for the left, and 14% support versus opposition to political for the far left. For current academics (also correctness, those under 40 back political from academic survey), the figures are, correctness by a 1.5:1 ratio, 40-59s are respectively, 6%, 9%, and 18%; and for even, with the over 60s net opposed, at current social science and humanities 0.7:1. This age structure likely portends at academics, 10%, 14%, and 20%. This least some change in the British suggests that there is no real difference in professional workplace of the future. “social justice” authoritarian attitudes The US situation may be more between academics and non-academics. acute than in Britain. The Cato Institute The share of those willing to use the survey we encountered earlier showed that disciplinary apparatus of the workplace to 60% of Republican professionals with expel dissenters is around 6-10%, rising to advanced degrees feared for their careers if 20% on the far left. their politics became known.113 In the Though only one of the two future, if the professions shift left due to questions was the same as in the academic increasingly progressive new cohorts of survey, some of the same factors matter graduates (which is already apparent in outside the university as matter inside it: a American political donations, and among willingness to discriminate against medical and legal graduates from top US conservatives and Leavers and, to a lesser universities), the climate for conservatives extent, being far left. As in my academic in non-academic workplaces may begin to analysis in Figures 117 and 118, these resemble that of the universities.114 questions predict whether an employee will endorse expulsion tactics. When Authoritarianism controlling for these drivers of intolerance in Figure 126, those in my non-academic I asked two questions about sample who work in universities are no expulsion in my UK non-academic survey. more likely than those outside academia to First, “If someone in your workplace was endorse taking disciplinary action against known to favor restrictions on ideologically nonconforming staff. immigration, would you support efforts by There are also some differences in the organization to let the person know the non-academic model in Figure 126. that they should find work elsewhere?” Activists were especially likely to endorse Given that a majority of the country feels expulsion whereas this was only this way, this is an especially stringent test borderline significant in the academic of authoritarianism. Second, “If a staff sample. Men were significantly more member in your workplace did research likely to endorse expulsion than women in showing that greater ethnic diversity leads the non-academic sample, whereas there to increased societal tension and poorer was no gender difference among social outcomes, would you support or academics. Younger people were not more oppose efforts by students to let the staff pro-expulsion in the non-academic sample,

178

whereas they very much were in the academic survey concerned diversity or academic survey. Finally, minorities came immigration, whereas there was more out as significantly pro-expulsion. This gender content in the academic survey’s may be because both questions in the non- illiberalism measures.

Predictors of Expelling Conservative Dissenter (Non- Academic Sample)

Minority*** 0.172

Activist*** 0.147

Discriminate v. Leaver*** 0.123

Discriminate v. Tory** 0.110

Male** 0.086

Very Left* 0.085

0.000 0.050 0.100 0.150 0.200

Figure 126. Pseudo-R2=.114. Other controls (not significant): age, university sector, Brexit vote. *p<.05; **p<.01;***p<.001. I also tested a version of the 65% backed the quotas, while only 23% of “decolonization” question with non- those in the center and on the right did. academics: “Please imagine there was a Young people and women were new initiative at work, stipulating that at significantly more likely to support least 50% of staff must be women, and affirmative action, but minorities were not. 33% people of color.” This is quite a Fifty-seven percent of self-described stringent stipulation given that the UK activists and 86% of far leftists who population is only about 15% nonwhite. described themselves as activists backed Answers ranged from publicly expressing affirmative action for women and opposition, through privately expressing minorities. opposition, remaining neutral, privately Figure 127 presents a model with a expressing support, or publicly expressing range of demographic controls, showing support. I had half the sample read a that those on the “very left” of the version of the UC Diversity statement with spectrum who also identify as activists academic content removed to see if this come close to full endorsement of moved opinion. affirmative action (scale point 3) while Overall, 31% expressed support those on the right lean neutral if activists and 32% opposed the measure, with 36% and lean towards being opposed if they undecided. Among those on the far left, reject the activist label – which most on

179

the right do. The diversity statement had academic statement is less workplace- only borderline significance: 33% of those specific than the academic one or because who read it backed the quota plan versus the question people were asked to answer 28% of those who did not. This is a major is more radical, involving a higher difference from the academic question on minority share and hiring quotas rather quotas for the curriculum where the than content quotas, which do not as diversity statement strongly shifted directly affect people’s life chances. opinion. This could be because the non-

Figure 127. Pseudo-R2=.056. Other controls: age, gender, minority, university sector.

Finally, these data show that about wouldn’t do so publicly. As in academia, 30% would privately oppose or support the there is considerable indifference or cross- initiative, about the same as would express pressuring among many, with ideology their views publicly. This shows a higher again a key factor. level of preference falsification among opponents of quota measures outside Empirical Summary academia than inside it: half the opponents of quotas outside academia say they would This report concentrates on two remain silent compared to only a fifth of forms of illiberalism, which I label hard academic opponents saying they would and soft authoritarianism. Its studies were remain quiet about their opposition to designed to elicit both experiences of curriculum quotas or dismissal campaigns. victimization and respondents’ willingness Regardless of this difference, there is little to act as perpetrators. On campus, evidence for the “silent majority” victimization falls mainly on a small hypothesis that most support freedom but

180

minority of conservative and gender- sampling methodologies. However, critical scholars. Nevertheless, a majority findings from all surveys essentially of scholars are involved as perpetrators in conform to previous studies of soft that they are willing to discriminate authoritarianism that uniformly find a) a against the right on at least one of six strong left skew in the professoriate; b) measures, and are either uncomfortable or significant political discrimination against unsure about sitting next to a gender- conservatives; and c) chilling effects and critical feminist. self-censorship. In Table 4, I compare my This work breaks new ground by UK and US surveys with the surveys asking about hard authoritarianism in the conducted by Inbar and Lammers 2012, form of people’s willingness to endorse Honeycutt and Freberg 2017, and Peters et dismissal or enforce compulsory reading al. 2020, from which many of my list quotas; as well as academics’ questions were drawn. Yancey’s (2011) experiences with the university’s data also show nearly identical results. disciplinary apparatus. More recently, similar studies have I then moved on to explore soft uncovered profound political intolerance authoritarianism in academia, introducing and chilling effects in universities in a list experiment to capture concealed Germany, and in Europe more broadly.115 discrimination. The results in this report This is not just an Anglosphere issue. come from a wide array of sources and

2012* 2017* 2020* UK UK US All SSH Cdn % Left/Remain/Clinton voter 85% 71% 75% 78% 81% 83% % Conservative/Leave/Trump 6% 14% 14% 16% 9% 4% voter Left:Right Ratio 14 5 5 6 9 14 Discrimination v 38% 26% 48% 10% 14% 23% Leave/Republican/Conservative for job (admitted) Discrimination v 32% 37% 40% Leave/Republican for job (revealed) Discrimination v Right for 24% 18% 35% 22% 24% 19% grant (admitted) % of Right-Wing Staff saying >50% >50% >50% 45% 50% 78% Hostile Climate for beliefs % saying Leave/Trump supporter 37% 30% 15% would openly express beliefs % saying Remain/Biden supporter 84% 86% 88% would openly express beliefs % Endorsing Dismissal of 8% 9% 7% Controversial Professors/Staff Number of responses (N) 774 618 794 820 235 1650 Approximate response rate from 26% 26% <10% 61- 61- 3% pool of those approached 75% 75% Table 4. Previous and Current Studies of Academic Bias and Discrimination. *Inbar and Lammers 2012; Honeycutt and Freberg 2017; Peters et. al. 2020. UK surveys from YouGov. Table 5 below summarizes the this study, limiting to “apples-to-apples” results from 7 surveys I have fielded in comparisons based on current SSH

181

scholars. With such a consistent density of denied the leftist makeup of the work replicating the same results, those professoriate (up to the 1990s), or who say who refuse to recognize the reality of that the earth is no warmer today than it political discrimination and chilling effects was a century ago. are not dissimilar to those who initially

UK UK US Canada US UK NAS 2020 UK You Mail Mail Mail PhD PhD Mail Non- Gov (SSH) SSH SSH (SSH) (SSH) academic SSH % Remain/Clinton/ 81% 71% 79% 71% 75% 85% 11% 73% Canadian left voter % Leave/Trump/Canadian 9% 5% 4% 6% 5% 6% 58% 21% Tory voter Left:Right Ratio 9 13 15 19 10 12 0.13 4

Discrimination v 14% 6% 24% 22% 32% 8% 7% Leaver/Trumper for job (open) Discrimination v 41% 26% 40% 45% 38% 0%* 26% Leaver/Trumper for job (revealed) Discrimination v Right for 26% 24% 21% 17% 39% 40% grant (open) % of Right-Wing Staff 52% 60% 76% 67% 89% 80% 71% 25% saying Hostile Climate for beliefs % saying Leave/Trump 30% 32% 13% 13% 6% 18% 12% 54% supporter would openly express beliefs % saying Remain/Biden 83% 90% 91% 80% 92% 94% 89% 81% supporter would openly express beliefs Number of responses (N) 235 222 706 260 111 77 227 867 Approximate share of 61- 2% 3% 3% 62- 85% 10% 16% targeted population 75% 72% Table 5. Summary of Survey Results. *Low sample size makes these experiment numbers less reliable.

Findings accumulated over a It is not clear that academics are decade convincingly show that a majority more likely to discriminate on political of conservative academics experience a grounds than professionals in other hostile environment for their beliefs in US, sectors. And while left-wing academics Canadian, and British universities. This is may be slightly more likely to discriminate a rational appraisal of the significant against the right in refereeing and structural discrimination against them in promotion than vice versa, right and left the higher education sector. Around half of discriminate against each other equally in left and centrist academics would hiring. Indeed, prior research shows that discriminate against a conservative or across society, political discrimination is Trump supporter/Leaver in at least some deemed acceptable in a way racial or aspects of academic life. Thirty to forty gender discrimination is not. percent would discriminate against The problem is mainly structural. populist right supporters in a job Conservatives in academia are application. outnumbered at least 6:1 by those on the

182

left, rising to between 9 and 14 to 1 among discrimination threatens hiring, promotion current social science/humanities scholars. and publication prospects. This compares with a ratio of about 2:1 in But hard authoritarianism results in other professions. Furthermore, political more immediate threats of verbal bullying, beliefs are more transparent in universities, online harassment, reputational damage, especially in the social sciences and and being investigated by universities’ humanities, where views can be deduced disciplinary apparatuses with the threat of from scholarly work. This means that even being penalized or fired. Even if if left and right discriminate equally academics don’t lose their jobs or suffer at against each other, and if individuals in the hands of a Twitter mob, the possibility academia are no more likely to that this could happen, in a field where discriminate politically than those in other jobs are extremely scarce, greatly sectors, the net result is a very high amplifies the chilling effect that cripples probability (often a near certainty) that academic freedom for political minorities conservative academics will face like conservatives and gender-critical discrimination. This form of bias is far feminists. more pervasive among the professoriate Notwithstanding the barriers faced than individual-level race and gender by ideological minorities in the discrimination, for which evidence is less professoriate, there is some good news. A conclusive. majority of academics, including a Discrimination leads to self- majority of leftists and many far leftists, censorship, curbing the freedom to would not support a campaign to oust a investigate and debate ideas that is the colleague for doing politically incorrect lifeblood of a properly functioning research touching on race, gender, or academy. With just 10-20% of Trump- and sexuality. Over half of left-wing Leave-supporting academics in the social academics would not discriminate against sciences and humanities willing to air their a Trump or Leave voter in hiring and a political beliefs, the views of half the majority or near majority wouldn’t electorate are effectively being silenced, discriminate in other aspects of academic limiting the kinds of conversations that are life. needed for mutual understanding. Finally, a majority, including most Sunstein’s conformity dynamic appears to on the left, do not support the silencing of have taken hold, restricting viewpoint controversial views. The principal problem diversity in the professoriate. The problem is that only a minority of academics on the is even more serious among academics left actively oppose authoritarian than students, where I found that 4 in 10 campaigns (even privately) that seek to British Leave-supporting students said a force intellectual dissenters from academia Leaver would be willing to express their – even if they don’t support such actions. beliefs in class.116 Fair-minded leftist academics outnumber Soft authoritarianism is the hard-authoritarian left by a factor of compounded by conservative fear of hard two or three (even in SSH fields), and authoritarianism, which is again more offer an important base from which to prevalent in the social sciences and build a future consensus in favor of humanities. This is because in these academic freedom. departments, both those who identify as “far left” and left activists – the groups most likely to discriminate against conservatives and Leavers and express left-authoritarian views – outnumber conservatives 2:1. Generalized

183

Policy Discussion Moreover, a study using the American National Election Study (ANES) shows that voters with higher education levels What to Do?: Policy Directions to consistently express more ideological Safeguard Academic Freedom prejudice than those with lower levels of education.119 As Nicholas Kristof Those who recognize the issue of confesses of his liberal tribe, “We’re fine discrimination in academia face questions with people who don’t look like us, as regarding how and the extent to which long as they think like us.”120 Case argues policy can ensure that the hard- that political prejudice may occur authoritarian minority of faculty and wherever there is a strong ideological bent students who currently encounter little in the workplace, but that political opposition within the academy are not prejudice at universities cannot be likened effective in silencing dissenting voices. to conservative conformity in industries One may likewise argue that it is also like mining: “epistemic injustice at the important to address the soft university arguably deserves special authoritarianism of political attention. This is because the academy is discrimination, which produces chilling putatively the central site for knowledge effects and self-censorship. The pervasive production and dissemination. While discrimination that has become normalized epistemic injustices within the academy in academia represents a clear epistemic are localized to a context, the importance injustice, in which a person’s views are of that context to knowledge production dismissed because of who they are, not the makes them a society-wide concern.”121 content of what they do and say. This report distinguishes between hard and For philosopher Spencer Case, soft authoritarianism; each presents a discrimination against the distinct problem. What policy approaches, conservative/libertarian political minority if any, might be taken? in academia is analogous to the prejudiced dismissal of testimony from blacks, Libertarian Approaches women, Muslims, or Communist Party members in courtroom settings such as those in Harper Lee’s To Kill a The two main policy approaches Mockingbird or the Army-McCarthy are what I will term libertarian and hearings. Even a few cases of injustice interventionist. Libertarians place their such as this would be deemed faith in cultural change, working to unacceptable today. The retort that all convince progressives and administrators conservatives, gender-critical feminists, or that free speech is an important value that Trump voters are ignorant reprobates is has protected the left in the past and precisely the same logic used against continues to do so on selected issues such blacks or communists to unjustly deflate as the study of the Middle East and Israel. their testimony. For, even if someone is Heterodox Academy, founded in 2015, is ultimately wrong, we should not adopt an at the forefront of this intellectual project. uncharitable starting point, prejudging Many center-left and liberal writers are them as holding the least defensible views also on board, such as those who signed of their sect’s extremists, but should listen the Harper’s Letter in defense of free to and assess their arguments rationally.117 expression. Helen Pluckrose and Yascha Importantly, research on Myside Mounk are two important commentators Bias shows that prejudice towards the in- who believe that a liberal left can be mobilized to resist the illiberal left within group is one of the few forms of bias that 122 is not correlated with intelligence.118 elite institutions such as universities.

184

Some libertarian-leaning thinkers, approach would require universities to such as Foundation for Individual Rights adopt the Chicago Principles, or an in Education (FIRE) president Greg analogous statement on academic freedom, Lukianoff, believe that when universities and remove or amend all non-compliant are publicly ranked on free speech, signals speech codes and policies. The leader in can shape consumer behavior and provide this respect is Britain, which announced a incentives for change. Good ideas, comprehensive set of academic freedom practices, and universities can drive out policies in February 2021, including new the bad without the need for government legislation. Endorsed by Prime Minister intervention. A new free speech culture Boris Johnson and drafted by Education can arise.123 Secretary Gavin Williamson’s team, the FIRE helps those who are accused measures center on a new Academic of violating university policies for legally- Freedom “Champion”, who would sit on protected speech with legal advice and the sector regulator. This individual can assistance. In Britain, the Free Speech hear cases, as well as proactively fine or Union (FSU) has recently taken up this instruct universities to amend policies and role, and has worked on behalf of the issue redress. Legal changes focus on accused to send letters to universities, bringing student unions under the rubric of provide some free legal advice, publicize free speech protections, allowing those cases, and help victims to crowdfund legal whose freedom has been infringed to sue campaigns, where appropriate. It is now their universities, and requiring branching into the United States, which universities to ‘promote’ academic should have a positive effect in countering freedom.126 Many of the recommendations campus illiberalism. draw on findings presented in a Policy Exchange report I co-authored in August Interventionist Approaches 2020.127 Academic freedom policies have Though welcoming libertarian also been instituted by conservative approaches, interventionists believe that Canadian provincial governments in only democratic government policy can Ontario and Alberta. These focus mainly on policy documents, however, and appear alter the incentive structure that currently 128 permits hard and soft authoritarianism in to have had limited practical effect. In universities. While building what Greg the US, 11 states have passed laws against Lukianoff terms a “free speech culture” so-called “free speech zones” that only that makes intervention redundant is the protect speech in a few designated areas. ultimate aim, interventionists believe that As recently as January 2021, Iowa, universities cannot reform from within, Florida, and Arkansas have been active in seeking to legislate to protect free speech and universities’ legacy and network 129 effects erect barriers to entry that limit in public higher education. new entrants. A further set of policy proposals Proponents of government for the American context is summarized in intervention, be it federal or state, argue the Goldwater Institute’s 2017 report. that individual rights are more important These seek to audit colleges to ensure that than the autonomy of institutions like they penalize those who disrupt speech universities.124 The role of government and uphold their free speech obligations. Four states have adopted its proposals to intervention would be to proactively 130 enforce the law. For instance, at present, date. Two complementary yet distinct around 9 in 10 American universities policy documents are the James G. Martin maintain speech codes that violate the First Center for Academic Renewal’s Amendment.125 An interventionist “Blueprint for Reform,” and the National

185

Association of Scholars’ “Freedom to and in politically congenial US states and Learn” principles.131 While both Canadian provinces, appears to be moving libertarian and interventionist activity is in the direction of the interventionist increasing, the momentum in Britain, approach.

186

Appendix branches are actually far more similar to each other than to the rest of society. 1. YouGov Survey of UK Academics. Retired and current academics also differ, but here again, there is more commonality The sample was collected on than difference between the two samples. March 27, 2020, by YouGov, whose As a result, while many of my analyses exceptionally large panel of respondents separate current from retired academics, enabled us to achieve an unusually large and SSH from STEM, I also use the full sample of university professors and sample for both tabular and statistical lecturers.132 The sample consists of 820 analysis. Using the larger sample is respondents, 484 of them currently especially important where statistical working as university professors or power is required. lecturers, and 336 who previously worked This study differs from the as academics (nearly all retired). I methods employed in those discussed concentrate more on current than retired discrimination above, with better staff where sample size permits. representativeness. Previous studies The average age of the sample of collected a convenience sample by current academics is 49. The average age emailing academics directly. With in the retired portion is 70. The response rates of around 25%, this respondents are 43% female (45% for introduces the possibility of self-selection. current staff) and 5% from a minority In contrast, YouGov’s panel consists of ethnic background. These data place my participants who complete a range of figures very close to the sector average on different types of surveys for gender, but with a substantial undercount remuneration, some of them active or of minorities and younger academics. retired academics. They are thus less likely Where appropriate, I use YouGov data to be selectively attracted toward filling weights based on official Higher out a particular survey, mitigating the risk Education Statistics Agency (HESA) data of bias. A majority of between 61 and 76% by ethnicity, gender, and age.133 Minority of YouGov’s panel of professors and respondents tended not to differ from the lecturers responded, resulting in a sample average on most questions, though female that is more likely to represent the actual and younger respondents often did. population of academics than any other 134 A little under half the sample study (UK or US) of this kind to date. teaches in the social sciences, humanities, Full question wording and response arts, psychology, or in education (SSH frequencies can be found at: fields). This share is approximately similar https://docs.cdn.YouGov.com/4lwd0ybm5 for current and retired academics. The c/BBResults_200423_Academics.pdf. remainder are scientists, economists, or teach medicine, business, engineering, or 2. UK Academic Mailout Online Survey in applied and vocational (STEM) fields. I focus more on the social sciences and This contains the same question humanities because political wording as the UK YouGov survey. It is considerations are a larger aspect of these based on a survey mailed out to around fields’ conceptual foundation, and are thus 10,000 academics with publicly-available assumed to exert greater influence over the email addresses in Social Science and culture and practices of these disciplines. Humanities departments from the top 100 The difference in attitudes to (of 143 listed) institutions across the UK 135 questions of political bias, academic in September 2020. N=338 for section freedom, and social justice between SSH 1, with N=222 completing the survey. The and STEM fields is significant, but the two UK survey can be compared to the UK

187

YouGov survey as a way to contextualize population sample that required many less- the North American results to see if educated respondents. Where the Prolific accidental and online mailout samples samples are less representative is with differ. Response rates to these surveys regard to gender and to some extent race, were low, so it is important to compare with women and minorities somewhat them with the YouGov and PhD surveys overrepresented. Statistical modelling which are little affected by the problem of allows me to control for these self-selection.136 I find little difference characteristics, however, to examine the between surveys, despite the radically impact of variables of interest such as different sampling methodologies. ideology, net of gender and race. Conducted on online survey platform 3. US and Canada Academic Mailout Prolific Academic, May 1-June 4, 2020. Online Survey N=170. N=78 SSH, 94 STEM.139

This is a survey mailed out to 5. US and Canada PhD Students Survey around 40,000 academics at the top 100 US and top 40 Canadian universities with See above on the UK Survey. Conducted publicly available email addresses in on online survey platform Prolific August 2020.137 It focuses largely but not Academic, July 11-September 25, 2020. solely on Social Science and Humanities N=397 (46 Canada, 351 USA). N=142 departments. N=1,777 for section 1, with SSH, 256 STEM.140 N=1,093 completing the survey (463 Canada, 1,308 USA. Completes: 290 6. UK Non-Academic Survey Canada, 803 USA).138 To compare with the academic 4. UK PhD Students Survey surveys, I also ran a comparable survey, on Prolific Academic, of 867 British I fielded two similarly-worded respondents. This reached approximately surveys to British and North American 16% of the total number of degree-holding PhD students who happen to be working and employed respondents on Prolific. as survey-takers on Prolific Academic, a Conducted online, August 21-22, 2020. high-quality online survey platform. As N=867.141 with the YouGov UK academic survey, these individuals are on the platform 7. National Association of Scholars filling out surveys for unrelated subjects (NAS) Survey, United States and are thus not self-selected for their interest in this subject. The surveys were In addition to these samples, which open for a considerable period of time, all lean strongly left, I sought permission permitting me to acquire responses from to field a partial survey to members of the 86% of UK PhD students and 63-72% of National Association of Scholars (NAS). American and Canadian PhD students on As a mainly right-leaning and centrist the system. association, it offers a distinct window into As a result, these samples are the views of the conservative scholarly unlikely to be affected by selection bias, minority. I received 227 responses, which offering a good snapshot of doctoral represents approximately 10% of the student opinion. While Prolific membership and helps boost my small respondents tend to be young, educated, samples of conservative academics. Based and tech-savvy, PhD students also fit this on a survey mailed out to NAS members. description, so this is not the drawback it May 6-June 12, 2020. N=227 (6 from might be if I were after a representative outside USA).142

188

8. Graduate Student Academic Careers January 5, 2021. N=843 (434 UK, 368 USA, 41 Canada). N= 361 SSH, 482 Conducted on online survey platform STEM.143 Prolific Academic, December 23, 2020-

189

References

1 Glynn, Adam N. 2013. “What Can We Learn with Statistical Truth Serum? Design and Analysis of the List Experiment.” Public Opinion Quarterly 77(S1): 159-172. 2 Acevedo, David. 2020. “Tracking ‘Cancel Culture’ in Higher Education.” National Association of Scholars, accessed January 31, 2021. Available at https://www.nas.org/blogs/article/tracking-cancel-culture-in-higher- education#caseslist. 3 McWhorter, John. 2020. “Academics Are Really, Really Worried About Their Freedom.” The Atlantic. Available at https://www.theatlantic.com/ideas/archive/2020/09/academics-are-really-really-worried-about- their-freedom/615724/. Also see So to Speak podcast, “Academics Are Really, Really Worried” with John McWhorter, by Nico Perrino, January 28, 2021. 4 “Academic Freedom in Our Universities: The Best and the Worst.” 2020. Civitas. Available at https://www.civitas.org.uk/publications/academic-freedom-in-our-universities/. 5 Personal correspondence with Toby Young, Free Speech Union. 2020; Stock, Kathleen. 2020. “Are Academics Freely Able to Criticise the Idea of ‘Gender Identity’ in UK Universities?” Medium. Available at https://medium.com/@kathleenstock/are-academics-freely-able-to-criticise-the-idea-of-gender-identity-in-uk- universities-67b97c6e04be. 6 Wright, Mike. 2019. “Lecturer Accuses University of 'Bringing Back Blasphemy as a Crime' After Dismissal for Criticising Religious Centre.” The Telegraph. Available at https://www.telegraph.co.uk/news/2019/02/28/lecturer-accuses-former-university-bringing-back-blasphemy- crime/; Turner, Camilla. 2019. “University Lecturer Has ‘No Legal Right to Be Anti-PC’ After Claims He Was ‘Hounded out’ by Left-Wing Colleagues.” The Telegraph. Available at www.telegraph.co.uk/education/2019/01/12/university-lecturer-claims-hounded-left-wing-colleagues. 7 “A Letter on Justice and Open Debate.” 2020. Harper’s Magazine. Available at https://harpers.org/a-letter-on- justice-and-open-debate/. 8 Karran, Terrence and Lucy Mallinson. 2017. “Academic Freedom in the UK: Legal and Normative Protection in Comparative Context. Report for the University and College Union.” University of Lincoln. Available at http://eprints.lincoln.ac.uk/id/eprint/26811/. 9 Note that the NAS is expressly concerned with questions of academic freedom, thus there may be a self- selection effect in the data. 10 Costello, Thomas H., et al. 2020. “Clarifying the Structure and Nature of Left-Wing Authoritarianism.” PsyArXiv Preprint, p. 35. Available at https://psyarxiv.com/3nprq/. 11 The index adds the four items and divides by 4, resulting in scores of 0, .25, .5, .75 and 1. 12 Results would suggest that in a larger sample, there would be a small effect of being “fairly left” on supporting a dismissal campaign. 13 Social media activity only available for UK YouGov sample. 14 Dreher, Rod. 2018. “Against Moralistic Therapeutic Totalitarianism.” American Conservative. Available at https://www.theamericanconservative.com/dreher/larry-chapp-moralistic-therapeutic-totalitarianism/; Lukianoff, Greg and Jonathan Haidt. 2017. “How Trigger Warnings Are Hurting Mental Health on Campus.” The Atlantic. Available at www.theatlantic.com/magazine/archive/2015/09/the-coddling-of-the-american-mind/399356. 15 “2020 College Free Speech Rankings.” 2020. FIRE. pp. 19, 40. Available at https://www.thefire.org/research/publications/student-surveys/2020-college-free-speech-rankings/2020-college- free-speech-rankings-view-rankings/. 16 Downs, Donald Alexander. 2020. Free Speech and Liberal Education. Cato Institute. pp. 134-141; Knight Foundation. 2018. “Free Expression on Campus: What College Students Think About First Amendment Issues.” 17 Kuran, Timur. 1995. Private Truths, Public Lies: The Social Consequences of Preference Falsification. Harvard University Press. 18 Sunstein, Cass. 2019. Conformity: The Power of Social Influences. NYU Press: p. 149. 19 Samuel, Juliet. 2020. “A Victory for Freedom at Cambridge Shows the Woke Mob can Only Win if we Let Them.” The Telegraph. Available at https://www.telegraph.co.uk/news/2020/12/11/victory-freedom-cambridge- shows-woke-mob-can-win-let/. 20 Kaufmann, Eric and Tom Simpson. 2019. “Academic Freedom in the UK.” Policy Exchange. Available at https://policyexchange.org.uk/publication/academic-freedom-in-the-uk/. 21 Clark, Cory. 2021. “How We Empower Political Extremists,” Psychology Today. Available at https://www.psychologytoday.com/ca/blog/the-antisocial-psychologist/202101/how-we-empower-political- extremists. 22 McAndrew, Siobhan, Dave O’Brien, and Mark Taylor. 2020. “The Values of Culture? Social Closure in the Political Identities, Policy Preferences, and Social Attitudes of Cultural and Creative Workers.” The Sociological Review 68(1): 33-54.

190

23 Smith, Christian. 2014. The Sacred Project of American Sociology. Oxford University Press: p. 82. 24 Hawkins, Stephen, et al. 2018. “Hidden Tribes: A Study of America’s Polarized Landscape.” More in Common. Available at https://hiddentribes.us; Juan-Torres, Míriam, Tim Dixon, and Arisa Kimaram. 2020. “Britain’s Choice: Common Ground and Division in 2020s Britain.” More in Common. Available at https://www.britainschoice.uk. 25 Abrams, Samuel. 2018. “Professors Support Free Speech.” American Interest. Available at https://www.the- american-interest.com/2018/04/18/professors-support-free-speech/. 26 Note that I do not have access to the raw data to model these relationships. 27 Bloom, Paul. 2017. Against Empathy: The Case for Rational Compassion. Random House. 28 Redden, Elizabeth. 2017. “Gender Trouble.” Inside Higher Ed. Available at https://www.insidehighered.com/news/2017/12/20/professor-resists-departmental-attempt-add-female-author- class-reading-list-sake. 29 “2020 College Free Speech Rankings,” 2020, FIRE. 30 “UC Statement on Diversity.” 2019. UC Davis Diversity, Equity & Inclusion. Available at diversity.ucdavis.edu/uc-statement-diversity. 31 Batty, David. 2020. “Only a Fifth of UK Universities say They Are 'Decolonising' Curriculum.” The Guardian. Available at https://www.theguardian.com/us-news/2020/jun/11/only-fifth-of-uk-universities-have- said-they-will-decolonise-curriculum. 32 Hillman, Nick. 2016. “Keeping Schtum? What Students Think of Free Speech.” HEPI Report. 85: 235-237; “The First Amendment on Campus 2020 Report: College Students’ Views of Free Expression.” 2020. Knight Foundation/Gallup. p. 2. 33 Stanovich, Keith. 2020. “The Bias that Divides Us.” Quillette. Available at https://quillette.com/2020/09/26/the-bias-that-divides-us/. 34 See Kaufmann and Simpson, 2020. 35 Ellis, John M. 2020. The Breakdown of Higher Education: How it Happened, the Damage it Does, and What Can be Done. Encounter Books; Lukianoff, Greg and Jonathan Haidt. 2018. The Coddling of the American Mind: How Good Intentions and Bad Ideas Are Setting Up a Generation for Failure. Penguin Books; Mac Donald, Heather. 2018. The Diversity Delusion: How Race and Gender Pandering Corrupt the University and Undermine our Culture. St. Martin's Press. 36 Stokes, Doug. 2020. “The Campus Grievance Industry.” The Critic. Available at https://thecritic.co.uk/issues/september-2020/the-campus-grievance-industry/. 37 Ellis 2020, p. 146; Lukianoff and Haidt 2018. 38 “Statement on Freedom of Speech: Regent House ballot.” 2020. Cambridge University. Available at https://www.cam.ac.uk/news/statement-on-freedom-of-speech-regent-house-ballot. 39 Sunstein, 2019. 40 The increase in political polarization over four decades is charted by Iyengar, Shanto and Masha Krupenkin. 2018. “The Strengthening of Partisan Affect.” Political Psychology 39: 201-218; Sobolewska, Maria and Robert Ford. 2020. Brexitland: Identity, Diversity and the Reshaping of British Politics. Cambridge University Press; “Lessons in Resilience: Canada's Digital Media Ecosystem and the 2019 Election.” 2020. Public Policy Forum. Available at https://ppforum.ca/articles/lessons-in-resilience-canadas-digital-media-ecosystem-and-the-2019- election/. 41 Iyengar, Shanto and Sean J. Westwood. 2014. “Fear and Loathing Across Party Lines: New Evidence on Group Polarization.” American Journal of Political Science 59(3): 690-707. 42 Yancey, George. 2011. Compromising Scholarship: Religious and Political Bias in American Higher Education. Baylor University Press. 43 Inbar, Yoel and Joris Lammers. 2012. “Political Diversity in Social and Personality Psychology.” Perspectives on Psychological Science 7(5): 496-503. 44 Honeycutt, Nathan and Laura Freberg. 2017. “The Liberal and Conservative Experience Across Academic Disciplines.” Social Psychological and Personality Science. 8(2): 115-123. 45 Peters, Uwe, Nathan Honeycutt, Andreas De Block, and Lee Jussim. 2020. “Ideological Diversity, Hostility, and Discrimination in Philosophy.” Philosophical Psychology 33(4): 511-548. 46 Inbar and Lammers 2012, pp. 500-501. 47 Honeycutt and Freberg 2017, pp. 118-119. 48 Ibid., p. 119. 49 Peters, et al. 2020, Figure 5. 50 Karran and Mallinson 2017. 51 Lipset, Seymour M. and Everett C. Ladd Jr. 1972. "The Politics of American Sociologists." American Journal of Sociology 78(1): 91.

191

52 Hamilton, Richard F. and Lowell L. Hargens. 1993. “The Politics of the Professors: Self-Identifications, 1969–1984.” Social Forces 71(3): 603-627; Rothman, Stanley, et al. 2005. “Politics and Professional Advancement.” Academic Questions 18(2): 71-84. 53 Klein, Daniel and Charlotta Stern. 2009. “By the Numbers: The Ideological Profile of Professors.” In Maranto, Robert, et al. 2009. The Politically Correct University: Problems, Scope, and Reforms. Rowman & Littlefield: pp. 18-26. 54 Gross, Neil and Solon Simmons. 2007. “The Social and Political Views of American Professors.” Working Paper presented at a Harvard University Symposium on Professors and Their Politics, p. 41. 55 Duarte, José L., et al. 2015. “Political Diversity Will Improve Social Psychological Science.” Behavioral and Brain Sciences 38. 56 Abrams, Sam. 2016. “Professors Moved Left Since 1990s, Rest of Country Did Not.” Heterodox Academy. Available at https://heterodoxacademy.org/blog/professors-moved-left-but-country-did-not. 57 Langbert, Mitchell. 2018. “Homogenous: The Political Affiliations of Elite Liberal Arts College Faculty.” Academic Questions 31(2): 186-197. 58 Langbert, Mitchell, Anthony Quain, and Daniel Klein. 2016. “Faculty Voter Registration in Economics, History, Journalism, Law and Psychology.” Econ Journal Watch 13(3): 422-51. 59 Halsey, Albert H. 1992. The Decline of Donnish Dominion: The British Academic Professions in the Twentieth Century. Oxford University Press: ch. 11 (Appendix I). 60 Carl, Noah. 2018. “The Political Attitudes of British Academics.” Open Quantitative Sociology and Political Science; Hanretty, Chris. 2016. “Is the Left Over-Represented Within Academia?” Medium. Available at https://medium.com/@chrishanretty/is-the-left-over-represented-within-academia-90b1cbe00e8a. 61 Matthews, Dylan. 2020. “One Pollster’s Explanation for Why the Polls Got it Wrong.” Vox. Available at https://www.vox.com/policy-and-politics/2020/11/10/21551766/election-polls-results-wrong-david-shor 62 Shields, Jon A. and Joshua M. Dunn Sr. 2016. Passing on the Right. Conservative Professors in the Progressive University. Cambridge University Press: p. 2. 63 van de Werfhorst, Herman G. 2020. “Are Universities Left‐Wing Bastions? The Political Orientation of Professors, Professionals, and Managers in Europe.” The British Journal of Sociology 71(1): 47-73. 64 Nakhaie, M. R. and Barry D. Adam. 2008. “Political Affiliation of Canadian University Professors.” Canadian Journal of Sociology 33(4): 882-84. 65 Lord Ashcroft. 2017. “How Did This Result Happen? My Post-Vote Survey.” Lord Ashcroft Polls. Available at https://lordashcroftpolls.com/2017/06/result-happen-post-vote-survey/. 66 Lancee, Bram and Oriane Sarrasin. 2015. “Educated Preferences or Selection Effects? A Longitudinal Analysis of the Impact of Educational Attainment on Attitudes Towards Immigrants.” European Sociological Review 31(4): 490-501. 67 Carl, Noah. 2017. “Lackademia: Why do Academics Lean Left.” Adam Smith Institute. Available at https://www.adamsmith.org/research/lackademia-why-do-academics-lean-left; Carl, 2018. 68 Langbert et al., “Faculty Voter Registration,” figures 4-5. 69 Langbert, Mitchell, and Sean Stevens. 2020. “Partisan Registration and Contributions of Faculty in Flagship Colleges.” National Association of Scholars. Available at https://www.nas.org/blogs/article/partisan- registration-and-contributions-of-faculty-in-flagship-colleges. 70 Ellis, 2020, pp. 26-36. 71 Tilley, James and Geoffrey Evans. 2014. “Ageing and Generational Effects on Vote Choice: Combining Cross-Sectional and Panel Data to Estimate APC Effects.” Electoral Studies 33: 19-27; Peltzman, Sam. 2019. “Political Ideology Over the Life Course.” Available at https://dx.doi.org/10.2139/ssrn.3501174. 72 Ellis 2020. pp. 26-36. 73 Lipset and Ladd 1972, p. 84. 74 Gross and Simmons 2007, pp. 29-30. 75 Lipset and Ladd 1972, p. 90. 76 Academics who finished their careers in academia and retired after 65 are actually somewhat more leftist than academics their age who stayed on into their old age, beyond 65. 77 This said, there is no evidence that this selection effect is stronger in SSH than in STEM disciplines. 78 Among whites, age was not significantly associated with voting Republican in 2016 (N=512) or for the entire 2000-2016 period (N=1,116). 79 Woessner, Matthew and April Kelly-Woessner. 2009. “Left Pipeline: Why Conservatives Don’t Get Doctorates.” In Maranto, Robert, et al. 2009. The Politically Correct University: Problems, Scope, and Reforms. Rowman & Littlefield: pp. 38-59. 80 Ibid., p. 50. 81 Gross, Neil and Ethan Fosse. 2012. “Why Are Professors Liberal?” Theory and Society 41(2), 151. 82 N=18 right-wing SSH PhDs and 23 right-wing SSH master’s students.

192

83 Gross and Fosse 2012, p. 159; Gross, Neil. 2013. Why Are Professors Liberal and Why do Conservatives Care? Harvard University Press: p. 162. 84 Simpson and Kaufmann 2019; Surridge, Paula. 2016. “Education and Liberalism: Pursuing the Link.” Oxford Review of Education 42(2): 146-164; Woessner, Matthew and April Kelly-Woessner. 2009. “I Think My Professor is a Democrat: Considering Whether Students Recognize and React to Faculty Politics.” PS: Political Science and Politics 42(2): 343-352; Strother, Logan, et al. 2020. “College Roommates Have a Modest but Significant Influence on Each Other’s Political Ideology.” Proceedings of the National Academy of Sciences 118(2): e2015514117. 85 Gross and Simmons 2007, p. 41. 86 Those who answered “don’t know” (1.8% in SSH, 3.2% in STEM), were added to the centrist figures. 87 Kaufmann and Simpson 2019. 88 Lukianoff, Greg, and Adam Goldstein. 2020. “Law Alone Can’t Protect Free Speech.” Wall Street Journal. Available at https://www.wsj.com/articles/law-alone-cant-protect-free-speech-11597269089. 89 On peer pressure, see Larson, Jennifer, Mark McNeilly, and Timothy Ryan. 2020. “Free Expression and Constructive Dialogue at the University of North Carolina at Chapel Hill.” UNC Chapel Hill. Available at https://fecdsurveyreport.web.unc.edu. 90 Inbar and Lammers 2012; Honeycutt and Freberg 2017; Peters et al. 2020. 91 It is impossible to calculate a specific response rate for any given survey because the company uses a system called “active sampling.” This means that respondents don’t get invited to take part in a specific survey and get allocated only to the survey that needs them most when they enter the survey system. In this case, the response rate was likely a lot higher than the usual 33% take-up, because the types of respondents (older middle-class people with degrees) are more likely to respond. 92 For more on list experiments, see, for instance, Gonzalez‐Ocantos, Ezequiel, et al. 2012. “Vote Buying and Social Desirability Bias: Experimental Evidence from Nicaragua.” American Journal of Political Science 56(1): 202-217. 93 Here it is important to note that the list experiment appeared early in the survey before political questions led some to discontinue. The sample sizes are thus larger - 1308 Americans and 463 Canadians - than later, when we look at discrimination by political affiliation. 94 Chi-Square p-value is .069. 95 N=33 Trump voters and 55 right-wing academics. 96 Allport, Gordon W. 1954. The Nature of Prejudice. Addison-Wesley Pub. Co.; Pettigrew, Thomas F. and Linda R. Tropp. 2006. “A Meta-Analytic Test of Intergroup Contact Theory.” Journal of Personality and Social Psychology 90(5): 751-783. 97 Prescott-Smith, Sarah. 2019. “Brexit has Caused More Arguments Than the General Election.” YouGov. Available at https://yougov.co.uk/topics/politics/articles-reports/2019/12/09/brexit-has-caused-more-arguments- general-election; Carl, Noah. 2020. “Who Doesn’t Want to Hear the Other Side’s View?” Medium. Available at https://medium.com/@NoahCarl/who-doesnt-want-to-hear-the-other-side-s-view-9a7cdf3ad702. 98 Existing studies use a 7-point gradation from “1-not at all” to “7-very much” so that anyone above 1 is classified as discriminating – I would thus expect even less of a multiplier as respondents can express very mild discrimination. 99 Chilton, Adam S. and Eric A. Posner. 2015. “An Empirical Study of Political Bias in Legal Scholarship.” The Journal of Legal Studies 44(2): 277-314. 100 Conway III, Lucian G., et al. 2018. “Finding the Loch Ness Monster: Left‐Wing Authoritarianism in the United States.” Political Psychology 39(5): 1049-1067; Harper, Craig. 2020. “Ideological Measurement in Social and Personality Psychological Science.” PsyArXiv Preprint, p.14. Available at https://psyarxiv.com/wpsje. 101 Sunstein 2019, p. 88. 102 Hawkins et al. 2018; Pettigrew and Tropp 2006. 103 Note that Canadians are highly familiar with American politics due to the high penetration of American news and television media, so can readily relate to American politics categories. 104 This interaction was not significant in the mailout and PhD datasets. 105 Sullivan, Andrew. 2018. “We All Live on Campus Now.” NYMag Intelligencer. Available at https://nymag.com/intelligencer/2018/02/we-all-live-on-campus-now.html. 106 Iyengar and Westwood 2014; Klein, Ezra and Alvin Chang. 2015. "’Political Identity is Fair Game for Hatred’: How Republicans and Democrats Discriminate.” Vox. Available at https://www.vox.com/2015/12/7/9790764/partisan-discrimination. 107 Elkins, Emily. 2020. “Poll: 62% of Americans Say They Have Political Views They’re Afraid to Share.” Cato Institute. Available at https://www.cato.org/publications/survey-reports/poll-62-americans-say-they-have- political-views-theyre-afraid-share. 108 Kaufmann and Simpson 2019.

193

109 “Campus Expression Survey.” 2019. Heterodox Academy. Available at https://heterodoxacademy.org/campus-expression-survey/; “2020 College Free Speech Rankings,” 2020, FIRE. 110 Roughly a dozen academics are included in my sample. 111 Note that the list experiment for this sample is only about half the size of the academic sample (N=202 vs 424) as I had a few other experimental treatments included as well, so there is slightly more noise in the results (i.e. 5% of Leavers revealed to discriminate against a Leaver). 112 Iyengar and Westwood 2014. 113 “Cato Institute Summer 2020 National Survey.” 2020. Cato Institute. Available at https://www.cato.org/sites/cato.org/files/2020-07/Political%20Expression%20Survey_2020_B_0.pdf. 114 Bonica, Adam, et al. 2020. “Ideological Sorting of Physicians in Both Geography and the Workplace.” Journal of Health Politics, Policy and Law; Bonica, Adam, et al. 2017. "The Political Ideologies of Law Clerks." American Law and Economics Review 19(1): 96-128. 115 Norris, Pippa. 2020. “Closed Minds? Is a ‘Cancel Culture’ Stifling Academic Freedom and Intellectual Debate in Political Science?” Available at SSRN https://ssrn.com/abstract=3671026; Revers, Matthias and Richard Traunmüller. 2020. “Is Free Speech in Danger on University Campus? Some Preliminary Evidence from a Most Likely Case.” KZfSS Kölner Zeitschrift für Soziologie und Sozialpsychologie 72(3): 471-497. 116 Kaufmann and Simpson, 2019. 117 Case, Spencer. 2020. “Political Conviction and Epistemic Injustice.” Philosophia: 1-20. 118 Stanovich 2020. 119 Henry, P.J. & Jaime L. Napier. 2017. “Education is Related to Greater Ideological Prejudice.” Public Opinion Quarterly, 81, 930-942; Stevens, Sean. 2018. “Education is Related to Greater Ideological Prejudice.” Heterodox Academy. Available at https://heterodoxacademy.org/blog/research-summary-education-ideological- prejudice/. 120 Kristof, Nicholas. 2016. “A Confession of Liberal Intolerance.” New York Times. Available at https://www.nytimes.com/2016/05/08/opinion/sunday/a-confession-of-liberal-intolerance.html. 121 Case, 2020. 122 Pluckrose, Hellen and James A. Lindsay. 2020. Cynical Theories: How Activist Scholarship Made Everything about Race, Gender, and Identity—and Why This Harms Everybody, Pitchstone Publishing (US&CA); “Why 14 Critics of ‘Social Justice’ Think You Shouldn’t Vote Trump.” 2020. Areo Magazine. Available at https://areomagazine.com/2020/10/27/why-fourteen-critics-of-social-justice-think-you-shouldnt- vote-for-trump/. 123 White, Ken and Greg Lukianoff. 2020. “What's the Best Way to Protect Free Speech?” Reason. Available at https://reason.com/2020/08/04/whats-the-best-way-to-protect-free-speech-ken-white-and-greg-lukianoff-debate- cancel-culture/. 124 Adekoya, Remi, Eric Kaufmann, and Tom Simpson. 2020. “Academic Freedom in the UK.” Policy Exchange. Available at https://policyexchange.org.uk/publication/academic-freedom-in-the-uk-2/. 125 “Spotlight on Speech Codes 2020.” 2020. FIRE. Available at https://www.thefire.org/resources/spotlight/reports/spotlight-on-speech-codes-2020/. 126 See “Higher education: free speech and academic freedom,” www.gov.uk. Available at: https://www.gov.uk/government/publications/higher-education-free-speech-and-academic-freedom 127 Adekoya, Remi, Eric Kaufmann and Tom Simpson. 2020. “Academic Freedom in the UK.” Policy Exchange. Available at https://policyexchange.org.uk/publication/academic-freedom-in-the-uk-2/. 128 “Freedom of Speech on Campus: Annual Report to the Ontario Government by the Higher Education Quality Council of Ontario.” 2020. Higher Education Quality Council of Ontario. Available at: https://heqco.ca/wp- content/uploads/2020/12/FINAL-HEQCO-Free-Speech-on-Campus-Annual-Report-to-Government-December- 2020.pdf. 129 Ellis, Lindsay. 2021. “No ‘Social Justice’ in the Classroom: Statehouses Renew Scrutiny of Speech at Public Colleges.” Chronicle of Higher Education. Available at https://www.chronicle.com/article/no-social-justice-in- the-classroom-new-state-scrutiny-of-speech-at-public-colleges. 130 Kurtz, Stanley, James Manley, and Jonathan Butcher. 2017. “Campus Free Speech: A Legislative Proposal.” Goldwater Institute. Available at https://goldwaterinstitute.org/wp- content/uploads/cms_page_media/2017/1/30/_Campus%20Free%20Speech%20Paper.pdf. 131 See Martin Center Staff. 2020. “Blueprints for Reform.” James G. Martin Center for Academic Renewal. Available at https://www.jamesgmartin.center/2020/12/blueprints-for-reform/ and “Freedom to Learn: Amending the Higher Education Act.” 2021. National Association of Scholars. Available at https://www.nas.org/storage/app/media/Reports/Freedom%20to%20Learn/Freedom%20To%20Learn%202021. pdf. 132 YouGov’s sampling and weighting methodology can be found at “YouGov | Panel Methodology.” YouGov. Available at https://YouGov.co.uk/about/panel-methodology/.

194

133 “Figure 6 - All Staff (Excluding Atypical) by Equality Characteristics 2018/19.” 2019. HESA. Available at https://data.gov.uk/dataset/452fa2dd-72e2-4de3-9e91-25be38dec27d/higher-education-staff-statistics-uk-2018- 19. 134 See footnote 92, above. 135 Selected from https://www.ukuni.net/universities. 136 Response rates are between 2 and 4% of those eligible in the UK and North American academic surveys. 137 See “Ranked: The Top 100 US Universities.” 2020. Top Universities. Available at https://www.topuniversities.com/where-to-study/north-america/united-states/ranked-top-100-us-universities; “Top Universities in Canada | 2020 Canadian University Ranking.” 2020. UniRank. Available at https://www.4icu.org/ca/. The top 100 in the US is a considerably more exclusive list than the top 40 in Canada. 138 Available at https://www.dropbox.com/s/2020gcp7rdog4fc/NA1.pdf?dl=0. 139 Available at https://www.dropbox.com/s/c6j3e7rqh753oms/UK PhDs.pdf?dl=0. 140 Available at https://www.dropbox.com/s/5xjnft83jz8wmw5/NA%20PhD.pdf?dl=0. 141 Available at https://www.dropbox.com/s/jkhi80qx9f08ojy/Nonacademic.pdf?dl=0. 142 Available at https://www.dropbox.com/s/vd6cbsryvb1tkkd/NAS.pdf?dl=0. 143 Available at https://www.dropbox.com/s/1zg776oyxiw94us/acadcareer.pdf?dl=0.

195