FROM ISSUE NUMBER 29 ~ FALL 2016 GO TO TABLE OF CONTENTS

TEXT SIZE A A A

Real Academic Diversity

GERARD ALEXANDER

Campus activism in 2014 and 2015 reached levels that reminded some observers of the student protests of the 1960s and '70s. But while some things sounded familiar — radical student demands, angry outbursts, and administrators caving to pressure — recent developments are no repeat of the famous events of the Vietnam era. That earlier activism was mainly focused outward, off campus, as students tried to influence national and international politics, including a controversial war. Those protestors articulated views that were already spreading through American society.

The recent activism appears very different. It is primarily focused inward, on campus issues such as admissions practices, faculty hiring, and curricular and other university policies. And it is disconnected from most Americans of all political stripes, who struggle to understand the mindsets being revealed and the agendas being pursued on campuses. Most controversially, activist students, often aided and echoed by sympathetic faculty, have moved to sharply regulate speech. They have tried to disinvite and ban speakers. They have deemed some verbal expressions to be threatening and even to be forms of violence. And they have declared certain viewpoints and topics to be not just offensive but harmful. This risk of harm is what justifies "safe spaces" to shelter vulnerable students, sanctions on people making disturbing statements, and "trigger warnings" on academic materials, which enable students to either brace themselves or avoid subjects altogether.

Critics from across the political spectrum have bemoaned these developments, partly by mocking fragile "snowflake" students, but mainly by insisting on the importance of free speech. That is understandable, since free speech is a cornerstone of a liberal-democratic society. But the focus on free speech misses a more fundamental problem on campus: the insufficient diversity of viewpoints. Greater diversity is important for two reasons. First, free speech has the most value when conversations take place between people offering genuinely differing products in the marketplace of ideas. Second, the lack of viewpoint diversity may be one reason why threats to free speech have become so powerful at universities in the first place. Intellectual diversity may help protect free speech, just as free speech helps protect diversity of thought.

AN ABBREVIATED SPECTRUM

The viewpoint diversity that matters most concerns the university's permanent population. Faculties boast diversity in demographic factors like gender, race and ethnicity, regional and international background, and more. But American universities notoriously harbor very limited diversity in political viewpoints. Some progressive professors persist in denying this fact, but it has been well known for decades that American university faculties have long stood significantly to the left of the American public.

There is some variation across campuses and especially across disciplines. The biggest distinction is between the heavily tilted liberalism or progressivism of the humanities (as in literature, religious studies, history, philosophy) and most of the social sciences (politics, sociology, psychology) and the greater balance, or at least lesser imbalance, in a very small handful of fields like economics and business. Faculty dispositions have been measured by surveys and through publicly available voter-registration data. For example, Stanley Rothman, S. Robert Lichter, and Neil Nevitte used a 1999 survey of faculty to conclude that nearly three-quarters of American professors located themselves on the left side of the political spectrum while only 15% self-identified as right of center. While liberal identification was most pervasive in the humanities and social sciences, it also dominated among natural scientists and even in fields like engineering and business. In certain disciplines, 5% or fewer of faculty — in some cases, basically none — were right of center. These included English literature, philosophy, religious studies, political science, history, and sociology.

More recently, Neil Gross and Solon Simmons, on the basis of a 2006 survey, argued that U.S. faculties are more moderate than critics suggest, but acknowledged the heavy preponderance of left-of-center views and the scarcity of straightforwardly conservative ones. Studies of individual fields come to similar conclusions. For instance, a study by Northwestern law professor James Lindgren found that less than 15% of law professors identified as Republicans in 1993-94. Things may have become even more imbalanced since then. A recent study by Bill von Hippel and David Buss found that the share of active social-psychology professors who are right of center may be well under 5%. Psychologist Jonathan Haidt concludes that, "in the 15 years between 1995 and 2010 the academy went from leaning left to being almost entirely on the left."

This, of course, does not translate into uniformity of views on campus. There is heterogeneity within the overwhelming majority. The campus left includes moderate and mainstream liberals, progressives, and diverse self-styled radicals. And even the most politicized faculty have research and teaching interests that range from environmental sustainability to gender and ethnic studies to old-fashioned economic redistribution, which can make for vigorous debate. But on-campus discussion takes place within an abbreviated portion of the American intellectual and political spectrum, a portion which not only substantially over-represents left-of-center perspectives, but which comes close to excluding the center-right and conservative views that make up a large portion of the off-campus American spectrum of thought. This deficiency in viewpoint diversity is a problem for two reasons. The first concerns scholarly research, and by extension teaching. The second concerns the danger posed to free speech.

CONFIRMATION BIAS

There are good reasons to believe that the lack of viewpoint diversity undermines the scientific-research enterprise that is one of the core missions of universities. The philosophical and political orientations of scholars might not matter to the study of plant biology, basic engineering, and accounting, but they are important to topics scattered throughout the humanities and social sciences that intersect with moral, philosophical, and political values and debates.

This describes huge swathes of what is studied and taught at colleges, such as how people interact in economic settings, the meaning and importance of works of literature, how people vote, what challenges to societies matter most and what the record shows about various ways of addressing them, how we should evaluate different moral and ethical positions, how we should think about legal systems, and which aspects of history are most important or worth exploring. In all these areas and many others, representation by only a portion of the viewpoint spectrum has major consequences for which topics scholars choose to investigate and teach, what findings they come to, and how they evaluate and treat each other's research. Specifically, a lack of viewpoint diversity can undermine the rigor of scholarly research in two ways.

First, viewpoints affect how much scrutiny is brought to bear on the study of specific subjects. Scholars, like everyone else, are susceptible to confirmation bias, which is the tendency to prefer and even seek out information or interpretations that conform to one's pre-existing beliefs. Simply put, people find more plausible — and work harder to confirm — the conclusions they already favor, and greet less congenial conclusions with greater skepticism and scrutiny. Scholarly training tries to help academics combat confirmation bias, which is why professors can admire scholars who change their minds in the face of new evidence. But the bias is not abolished. Here as elsewhere, the best way of addressing the problem is to work with rather than against the grain of it. Confirmation bias has an upside: It motivates skepticism when one's values and views are challenged. As a result, we should expect more rigor when scholars have differing views. This is what frequently happens on campus between, say, moderate liberals and progressives, and between scholars who emphasize economic factors versus cultural ones, as each side examines the other's assumptions and findings with fine-toothed combs.

But for the same reason we logically expect much less scrutiny of assumptions, views, and findings that are shared by left-of-center scholars, assumptions that distinguish them from conservatives. The scholars most likely to scrutinize those assumptions — conservatives of various types — are either missing or too few in number to offer a meaningful counterbalance. Because of this, a field of academic study can be characterized simultaneously by vigorous debate over some issues (the ones that left-of-center researchers disagree about) and a lack of debate over others (the assumptions shared by those scholars). The more that a given field of study is populated by scholars who share important underlying assumptions, the more likely will be the emergence of orthodoxies that go unexamined or at least under-examined. This allows error to arise, endure, and thrive.

Consider an example: Many political theorists in recent decades came to think of conservative religious beliefs as being in tension with democratic deliberation. After all, fundamentalism and the confidence that comes from having God on one's side are logically at odds with the mutual respect and intellectual give-and-take required for civil discourse and joint decision-making. The assumption is clearly plausible. But Jon Shields, a center-right political scientist, was skeptical enough to spend substantial time with pro-life activists, who it turned out were often in the habit of remaining civil when engaging with others and limiting their arguments to ones reasoned from testable facts and shared principles. Peter Steinfels joked in a review that the book in which Shields presented his novel evidence — The Democratic Virtues of the Christian Right — had a title that, to many liberal academics, was "akin to 'The Winning Ways of Serial Killers.'"

Or take another example: It was a long-standing proposition in social and political psychology that conservatives are more intolerant than liberals, perhaps because liberals are more intellectually sophisticated or open-minded. Occasional items in scattered studies ran against this assumption, but it took decades and a group of heterodox scholars who had enough doubts about the claimed "prejudice gap" to test it more creatively. Research by Mark Brandt, Christine Reyna, John Chambers, Jarret Crawford, and Geoffrey Wetherell found that when subjects were asked about more carefully designed lists of "targets" — lists that included groups like the military, Catholics, and Protestants — liberals proved to be just as intolerant as conservatives.

In these and other cases, many scholars were content to operate for years on the basis of assumptions and interpretations they assumed were true but that later, closer scrutiny suggests were flawed and were resulting in erroneous conclusions. Conventional left-of-center researchers could have more aggressively tested those assumptions (and hundreds more like them). But the level of motivation required to do that comes more easily to scholars who do not intuitively share the assumptions to begin with and may even find them objectionable. It is these scholars who are more likely to exert themselves in questioning such claims, and to be energetic and creative in testing them. But there are far too few of them to apply that scrutiny systematically.

Some scholars have come to think of ideology as a form of vigilance against specific moral transgressions. Ideological dispositions also make us vigilant against — or at least skeptical of — specific empirical claims. The more ideologically lopsided a community is, the more vigilance it will display against some claims, but the less it will mount against other, preferred ones. In contrast, the wider the array of viewpoints in a community, the greater the chance that vigilance will be targeted in many directions. If a field of study includes very diverse points of view, few major claims are likely to escape the scrutiny of scholars motivated to see if they can refute assumptions and conclusions they do not like.

As heterodox social-psychologist Jonathan Haidt puts it, "When someone says something, other people should be out there saying, 'Is that really true? Let me try to disprove it.'" This may sound like the mobilization of confirmation bias rather than the combatting of it, but the harnessing of bias can result in greater rigor. This is less to say that objectivity is impossible, and more to say that objectivity is not easily achieved and we need help approaching it. Greater diversity of viewpoints can help rectify this. Haidt is pointing out that the genius of the academy has long been its ability to use the confirmation biases of some members to counteract the confirmation biases of other members. If they all rely on rigorous methods, they hold each other accountable. The fact that we all need critics to improve our thinking was pointed out long ago by John Stuart Mill. The person "who knows only his own side of the case, knows little of that" because he does not know the reasons informing the contrary view and so cannot refute them. Mill insists that we can truly learn those reasons only from "persons who actually believe them....He must know them in their most plausible and persuasive form."

We cannot draw much comfort from the fact that there are some examples of heterodox scholars who challenge reigning assumptions and interpretations in this or that research area. Insufficient viewpoint diversity means that such scholars are simply too thin on the ground to do that effectively in the hundreds or thousands of research areas that comprise the modern social sciences and humanities. And even in those areas in which one or a small handful of challengers are present, it is tempting for the overwhelming majority to dismiss or ignore their work. This happens all too easily in heavily populated parts of the social sciences where it is difficult to keep up with all the published work as it is; it is even easier in the humanities, where standards of evidence are less clear.

And we have strong indications that erroneous assumptions and interpretations have not been sufficiently challenged in recent years. This has been made especially evident by several major failures of replication. A leading characteristic of scientific findings is that they are reproducible by other researchers. Yet it is well-known that scholars overwhelmingly spend their time generating new findings rather than trying to replicate past ones, despite the value of the latter to the overall accumulation of knowledge. Recent efforts to correct this have had results ranging from disappointing to scandalous. Some cases involve outright fraud, such as the Dutch psychology professor Diederik Stapel, who has admitted to fabricating data in dozens of papers. But most involve well-meaning researchers whose findings are simply not reproduced when the same data are examined or when the same experiment is re-run under the same (or nearly identical) conditions. Major reproducibility probes have recently drawn attention in pharmaceutical research and social psychology. One prominent effort to replicate studies in psychology, led by University of Virginia psychologist Brian Nosek, announced in 2015 that it was able to reproduce the results of less than half of 100 experimental and correlational studies published in 2008 in several leading psychology journals.

Reproducibility failures are partly the result of understandable but flawed professional incentives: Scholars get more credit for generating their own findings than for testing those of others. But they are also partly a function of the dangers of insufficient scrutiny. The deficient scrutiny in the reproducibility cases that have received the most attention, for example in social psychology, is not explained by a lack of political diversity in particular. But the failures represent strong evidence that groupthink and lowered skepticism are pervasive in these fields. Much more skepticism is needed than has been on offer. This reinforces the intuition that groupthink and reduced skepticism driven by the lack of viewpoint diversity in particular are also serious problems.

UNASKED QUESTIONS

A lack of viewpoint diversity can undermine scientific research in a second way, in addition to weakening rigor in specific research areas. In an article calling for more political diversity in social psychology, José Duarte, Jarret Crawford, Charlotta Stern, Jonathan Haidt, Lee Jussim, and Philip Tetlock suggest that confirmation bias may have effects that extend beyond how individual projects are carried out; it may influence which topics scholars decide to study to begin with. Ideologies also provide people with intuitions about what is important and how the world works. For scholars, those intuitions help generate research agendas by suggesting what is interesting and what sorts of factors have the most explanatory punch. In other words, scholars may find more interesting those subjects that validate their preferred narratives. Scholars' engagement with topics they already find interesting, in turn, will inevitably push even further to the margins those topics that do not strike them as compelling. The result can be entire areas of potential intellectual investigation that are neglected but that would be researched — and taught — if academia were more intellectually diverse.

For example, academics in several disciplines have invested very heavily in the study of racial and ethnic prejudice. That is understandable. Most of us agree that these forms of bigotry have played an important role in the political development of the United States and a number of other countries. They also fit a liberal narrative in which racism remains a robust and odious presence and drives a great deal of contemporary American politics — a narrative in which liberals understand themselves to be on the side of the angels. Conservatives have long articulated the intuition that, in addition to prejudice, two other psychological and cultural attitudes — envy and guilt — also make important contributions to politics. These attitudes may shape people's views toward diverse societal groups, inform positions on many public policies, and help explain differences between left and right.

Yet there is virtually no academic study of or teaching about either envy or guilt, though they suggest many potentially interesting research questions. For instance, how might envy influence support for tax policies or perceptions of who pays their "fair share" in taxes? If envy is consequential, is it evenly distributed across countries, and across time within given countries? Does guilt play the role in modern progressive politics that some conservative intellectuals have long suggested? These attitudes presumably go unstudied partly because left-of-center scholars simply do not think of them as factors relevant to socio-political life. But it also suggests that those scholars do not engage enough with conservative views, or take those views seriously enough, even to want to test the plausibility of conservative intuitions that have appeared in many countries for centuries.

Or consider race in another context. Most American universities have used racial preferences in admissions for decades. This decision was accompanied by strikingly little research on the effects of these practices on affected members of minority groups. This is presumably because most scholars assume that these policies straightforwardly help their intended beneficiaries. A decade ago, Richard Sander, a law professor at the University of California, Los Angeles, grew concerned about the experience of many minority students at his law school, and translated this into a research agenda about the academic trajectory of law students admitted with strong racial preferences. Such an investigation required prying data out of what proved to be very defensive schools, and also weathering accusations from collegial quarters that only dubious motives could inform the questions he was asking. Sander and a few collaborators have produced a stream of papers that demonstrate that minority students admitted to elite law schools with strong race preferences have graduation and bar-exam passage rates not only much lower than their white classmates, but also lower than their minority peers who attended law schools at which they were more closely matched with the median student. Similarly, minority students admitted to undergraduate colleges with strong racial preferences appear to have higher rates of defecting away from STEM (science, technology, engineering, and math) majors than their minority peers who went to schools where they were more closely matched with the student median.

In sum, it is quite possible that strong race preferences have resulted in fewer and not more minority lawyers, doctors, engineers, and scientists in America. This research has invited rebuttals by scholars who seem motivated to defend affirmative-action admission. The fruitful result is that an entirely new research area now exists that has potentially significant implications for public policy, institutional practice, and social-science theorizing about academic achievement and other topics.

Viewpoint diversity was not absolutely required for this research to appear. A liberal scholar could have spearheaded the "mismatch" project. In fact, a liberal did. But Sander is very unusual; he was willing to entertain serious doubts about a long-standing practice popular among people like himself, question key causal assumptions he had held for years, contemplate politically and intellectually uncomfortable implications, make it a priority to investigate the issue, press the matter when he encountered resistance, shrug off suggestions that he was aiding bigots or was perhaps one himself, and combat institutional foot-dragging in order to access the necessary data. He was also entrepreneurial enough to encourage a few other professors to engage with the topic, resulting in enough research to make the argument more difficult to ignore.

Relying on unicorns like Sander is not a smart strategy for ensuring that our universities generate the broadest array of active research agendas. It makes far more sense to harness the biases of politically diverse scholars whose normal, path-of-least-resistance choice is to question each other's shibboleths. In this way, conservative scholars, who must have the same commitment to rigor as liberal and progressive faculty, can offer their colleagues, and society in general, not just alternative takes on topics that liberals think are important but alternative takes on what topics might be worth studying to begin with. We know what some of these research agendas are because they have been generated by conservative-leaning academics. In many cases, though, the lack of intellectual diversity means that there are some whole lines of inquiry of which we're simply unaware.

FIGHTING CENSORSHIP

Greater viewpoint diversity may be able to do more than improve scientific research. It may be a powerful antidote to recent demands for restrictions on free speech. It is not entirely clear what is motivating the drive to censor speech on campuses. After all, it is possible to hold ideas passionately, as these activists clearly do, without trying to limit participation and expression by others. Two possible motivations for censorship stand out. First, censorious campus progressives may simply be manifesting the illiberal instincts that dominant groups have displayed in many political and institutional contexts. Second, the strongest opponents of free speech often seem to echo thinking in those fields in the social sciences and humanities that emphasize the power of culture and discourses. According to this line of thought, people's mental models and speech have profound influence over outcomes of all kinds, and moralistic activists are therefore justified in policing speech.

For example, expressions of bigotry, and even so-called "microaggressions," may have the power to perpetuate bigotry itself, by reinforcing stereotypes that entrench the majority's sense of privilege and demoralize and undercut the performance of members of minority groups. Activists have similarly claimed that the simple discussion of rape can re-traumatize survivors and that speakers who challenge progressive orthodoxies legitimize viewpoints that are deemed immoral or threatening. These two possible motivations for censorship — confident domination by the campus left and belief in specific theories that emphasize the power of words — are not mutually exclusive and could be working together.

Whatever precisely explains the drive to censor, greater viewpoint diversity can help. The enhanced scrutiny that it brings regularly reveals errors in our thinking and highlights the limitations of any given body of thought. Diversity can thereby help create a climate in which participants conclude that no single current approach has a monopoly on truth and that other intellectual strains can and do generate valuable insights. That is a climate of intellectual humility, in which we each continue to have instincts about how the world works, vigorously mount arguments, and work to identify truths, but at the same time recognize both the fallibility of our current specific assumptions and theories and the potential value of the views of many others.

That is the opposite of the intellectual insularity and overconfidence that is on flamboyant display on campuses today. Today's activists, like others before them, think that views different from theirs are clearly and dangerously wrong. It takes a pronounced intellectual confidence in that judgment to take the further step of believing that it is justified to suppress those competing views. It is not a coincidence that intolerance is radiating across universities from those subfields of the humanities and social sciences in which viewpoint diversity is most absent and rigorous scrutiny is most anemic.

The activists' beliefs about which problems matter most, what causes them, and what to do about them, are obviously debatable. The problem is that, in many cases, the issues are not actually debated. They go unchallenged in too many classrooms, on too many reading lists, in most campus forums, and in many scholarly journals and workshops. In these contexts, interpretations that trend in non-left directions can be as rare as flat-earth views are in astrophysics, and risk being treated in about the same way — despite the fact that non-left worldviews have and continue to make valuable contributions.

Peter Wood, president of the National Association of Scholars, describes a human susceptibility to finding "certainty where there is none." Greater viewpoint diversity would help fight that propensity. Free-speech practices are meant to help protect diversity in thought. But we have to remember that free speech, too, is an outcome that has to be caused. It is not a given. It often needs institutional support to be applied consistently. And above all, participants have to believe in free speech for it to be robust. Greater viewpoint diversity may be a crucial prop in causing and sustaining that belief.

SEEING THE PROBLEM

What should reformers do about this? Some argue that universities should more aggressively seek to hire conservative academics. But these calls usually fail to address the shortage of conservatives enrolled in Ph.D. programs. So long as the political demography of Ph.D. production remains the same, conservatives will simply not have enough credible and well-trained people to make a serious dent in the much, much larger number of liberals and progressives who go on the social-science and humanities academic-job markets every year.

José Duarte and his co-authors, in calling for greater diversity in social psychology, propose another strategy. Maybe the benefits associated with greater viewpoint diversity could be achieved by a forthright acknowledgement of the problem by existing scholars. It could become a part of the institutional culture at universities that limited viewpoint diversity risks undermining the scientific-research mission, that this is a problem that needs to be remedied, and that adversarial collaboration is valuable in research projects.

As Haidt puts it, we need to institutionalize a mentality of trying to disconfirm our claims. Maybe that mentality can be taught. The example of Richard Sander and his research on affirmative-action mismatch shows that liberals can improve rigor and expand research agendas if they are sufficiently heterodox in their thinking. Sander is not alone. Haidt is among a number of other liberal and moderate scholars who bring unusual creativity and commitment to diverse thinking to their research agendas. He has led several largely politically moderate professors in founding Heterodox Academy, a group that advocates for greater viewpoint diversity at universities. (I am also a member.)

But this solution faces challenges, too. For starters, we do not yet know how to change the university culture in this direction. Even if we did, this approach asks the liberal and progressive scholars who dominate university life to constantly swim upstream. It asks them not only to believe that it would be good if they could think like conservatives when it comes time to test their preferred theories and develop research agendas, but also to go ahead and actually think like conservatives. That is asking a lot, and it seems much less likely to succeed than having significant numbers of conservative scholars around, working with and not against the grain of their own biases and intuitions. How to develop sufficient numbers of them is a challenge that conservatives have not met. But things might be different if universities decide this is a problem they should help solve. Schools have applied millions of hours of work to the priority of improving racial, ethnic, and gender diversity. Viewpoint diversity could be elevated to similar prominence and urgency. On-campus climates explicitly or — at least as often — implicitly hostile to conservative ideas could be called out, shamed, and combatted. Schools could make it a mission to encourage more intellectually diverse undergraduates to enter the Ph.D. pipeline.

It may help motivate universities in these directions to appreciate that they would not be the only beneficiaries of a more viewpoint-diverse intellectual community. We already see the dangers to free speech posed by a new crop of campus activists. The threat will only get worse if, after graduating, they assume they can silence and punish their opponents in the great game of American democracy, instead of learning to take them seriously when they should have: in school.

Gerard Alexander is a professor of politics at the University of Virginia.