A new study presents evidence that more than a quarter of sociologists (27.8 percent) would “weigh favorably” membership in the Democratic Party by a candidate for academic appointment, and nearly 30 percent would weigh favorably a prospective candidate’s membership in the ACLU. More than a quarter (28.7 percent) would disfavor hiring a Republican, and 41.2 percent would weigh negatively a candidate’s membership in the National Rifle Association.
The study shows even greater bias against candidates with particular religious affiliations. Substantial numbers of the sociologists surveyed said they would be “less likely to hire” evangelical Christians and fundamentalists if they were aware that a candidate fell into either of those categories. Evangelicals face the barrier that 2 out of 5 sociologists (39.1 percent) are disposed not to hire them. Fundamentalists fare even a bit worse: 41.2 percent of sociologists say they would take such an affiliation negatively into account.
These figures are in dramatic contrast with the claims recently published by Neil Gross and his colleagues in a series of articles that offered evidence that the wide disparity between the percentage of liberals and conservatives on the faculty of U.S. colleges and universities is almost entirely the result of “self-selection.” In “Political Liberalism and Graduate School Attendance: A Longitudinal Study,” Gross et al. analyzed data from a long-term health study that happened to capture information about political leanings. They showed—convincingly—that students who as undergraduates define themselves as liberal are far more likely than students who define themselves as conservative to pursue the doctoral degrees that lead to academic careers. I commented on the study in “Unnatural Selection.” What Gross and his colleagues did not explain is why conservatives are so much less inclined to pursue this career path. Could it have something to do with their recognition that the odds are stacked against them because they would face de-selection by hostile search committees?
Gross and his colleagues published another essay, “Political Bias in the Graduate Admissions Process: A Field Experiment,” which reported on what happened when they sent fake letters of inquiry to directors of graduate study. The letters said the inquiring student had volunteered either in the Obama campaign or the McCain campaign (or neither), and the study found that the directors responded pretty much the same way regardless. Gross et al. took this as evidence that there is no bias against conservatives in the graduate admissions process. I critiqued that analysis in “A Flawed Experiment.” In my view a student’s admission that he campaigned for McCain is a poor proxy for conservatism. (What if he had said, “I did an internship last year in the Washington office of the National Rifle Association,” or, “I took a year off from college to do missionary work for my fundamentalist Christian church”?) I also expressed my doubts that a preliminary letter of inquiry to directors of graduate study would unveil the real biases. The directors generally filter out only the manifestly unqualified. The biases typically affect proceedings later in the process of academic recruitment and by much less obvious mechanisms.
And that’s exactly what this new study shows.
The Collegiality Survey
The new study by George Yancey, a professor of sociology at the University of North Texas, is Compromising Scholarship: Religious and Political Bias in American Higher Education (Baylor University Press, 2011). I follow the scholarship on bias in higher education pretty closely but I had not heard of Yancey until he posted (anonymously but with a mention of his book) a comment on Flawed Experiment essay. Unlike Gross, who has made a specialty of studying the political affinities of academics, Yancey is relatively new to the field. His dissertation explored “the social backgrounds that sociologists tend to share” and how those relate to “social biases in the field,” but the substantial body of his published work before now has been is on matters like interracial dating and marriage, interracial families, multiracial religious congregations, racial diversity on college campuses, and racial identity.
Perhaps because he is venturing into new territory, Yancey proceeds with a great deal of caution. He spends a good deal of time explaining what biases are and why they are hard to study (“individuals who dislike those in other social groups” are often unwilling “to disclose their displeasure.”) He notes that such self-disclosure is especially rare in the world of contemporary higher education in which “tolerance” is regarded as a supreme value and “intolerance has become the new curse word.” To get past this barrier, he devised as a survey instrument a “Collegiality Survey,” that focused on the characteristics that academics find desirable or undesirable in their colleagues. The key question on the survey is as follows:
Assume that your facility is hiring a new professor. Below is a list of possible characteristics of this new hire. Many of them are characteristics that you can not directly inquire of prospective candidates. However, if you were able to learn of these characteristics about a candidate, would that make you more or less likely to support their hire? Please rate your attitude on a scale in which 1 indicates that the characteristic greatly damages your support to hire a candidate, 4 is that the characteristic does not makes a difference, and 7 indicates that the characteristic greatly enhances your support to hire the candidate. If you do not understand the characteristics, then please indicate such with “n/a.”
Each of the characteristics is phrased, “The candidate is…” and the total list consists of: Democrat, Republican, Green Party member, Libertarian, Communist Party member, member of the NRA, member of the ACLU, heterosexual, homosexual, bisexual, transgendered, atheist, Mormon, Fundamentalist, Evangelical Protestant, Mainline protestant, Catholic, Muslim, Jewish, vegetarian, [someone who likes] to hunt, married, divorced, in a cohabiting relationship , single with children, over 50 years old, and under 30 years old.
Yancey first sent to the survey to 1,500 academic sociologists chosen randomly from the Directory of the American Sociology Association. Subsequently he sent the same survey to cohorts of academic political scientists, anthropologists, historians, physicists, chemists, experimental biologists, English professors, foreign language professors, and philosophy professors.
The result is an abundance of data that goes far beyond the usual attempts to correlate declared ideological allegiance with other characteristics. Here we have direct testimony about practitioners’ actual biases. Yancey is nothing if not cautious. He allows that an individual’s declaration of a bias for or against some social category doesn’t necessarily mean that the individual acts on that bias. But the data is all by itself dramatic evidence that social biases are likely to play a huge role in academic appointments.
That’s because of something Yancey doesn’t say—namely, that when academic departments conduct searches, they typically strive to achieve consensus and go out of their way to accommodate those who have strong dissents about a particular candidate. In more than two decades as a college and university administrator, I saw very few recommendations from search committees that were not unanimous. And in my time serving on search committees I frequently saw how one strong-worded dissent could be sufficient to sink a candidate who at the outset of the discussion appeared to have majority support. There is a “group dynamics” in academic decision-making that belies the model that a majority who declare themselves as open-minded and unbiased will win the day.
When a quarter to a half of the members of a field openly admit that they would prefer that people who fit a certain social profile not be their colleagues, you can be sure that very few such people will end up with academic appointments.
Anti-Fundamentalist > Anti-Conservative
Yancey’s book is eye-opening in a lot of ways. His most original contribution may be the demonstration that bias against certain kinds of Christians outstrips bias against political conservatives. That was true not just in sociology but in all the disciplines he examined, and across the social sciences, the humanities, and the physical sciences. (He didn’t examine the professional fields, such as business and education.) The biases were least prominent among physical scientists and most prominent among social scientists.
The single most profound prejudice was among anthropologists towards fundamentalists. As Yancey notes, “more than two-thirds of the respondents indicated that knowing that a candidate was a fundamentalist would negatively influence whether they would hire that candidate.” That’s an astonishing degree of animus. It means that an anthropologist, no matter where he went to graduate school, the quality of his scholarship, his knowledge of the field, the merits of his current research, or his ability as a teacher, would—if he was known to be, say, a Seventh Day Adventist—have virtually no chance of academic appointment in the United States, except perhaps at some small denominational college.
Yancey’s finding about bias against Christian Evangelicals and Fundamentalists closely matches the results of a 2007 study by Gary Tobin and Aryeh Weinberg for the Institute for Jewish and Community Research. In Volume 2 of Profiles of the American University: Religious Beliefs & Behavior of the College Faculty, they reported:
Faculty Hold the Most Unfavorable Feelings toward Evangelicals
Just one group elicited high negative feelings among faculty: Only 30% ranked their feelings toward Evangelical Christians as warm/favorable, with only 11% feeling very warm/favorable, the lowest ranking among every other religious group, and 53% said that they have cool/unfavorable feelings toward Evangelical Christians. Faculty feelings about Evangelicals are significantly cooler than any other religious group, leading Mormons as the least liked religious group by 20%. These negative feelings are noted across academic disciplines and demographic factors.
The “unfavorable feelings” that Tobin and Weinberg uncovered turn out in Yancey’s study to be experienced by faculty members who hold them as a legitimate basis for discriminating against a whole class of people.
I can imagine as I write this that a fair number of my Chronicle readers are adding something to the effect, “And rightly so!” Bias by academics against Christian conservatives does not register very strongly if at all as a transgression against the principles of intellectual honesty and fair-mindedness. These groups are academic pariahs, frequently characterized as stupid, anti-intellectual, doctrinaire, ill-disposed towards the values of liberal learning, and deserving of their ostracism. It isn’t particularly hard to find academics willing to give voice to these attitudes, and in fact Yancey tucks in a chapter in which he quotes some of the invective from sociologists’ blogs.
The prejudices in this case don’t stand up any better than other kinds of prejudices—against blacks, women, gays, etc. There are plenty of smart, well-educated, culturally sophisticated conservative Christians. And those who indulge in cartoon caricatures of conservative Christians or act with bias against individuals who they fit to this category are just as guilty of bias as they would be if they engaged in anti-Semitism, racism, sexism, or homophobia—or indeed ostracism of political conservatives.
Yancey’s book offers a feast of surprising but sturdily documented observations. Female academics, for example, are significantly more biased (or, alternatively, significantly more willing to admit bias) than male academics. This holds true of bias in favor as well as biases against and with very few exceptions applies to every category in the survey. Hard as things are for evangelicals and fundamentalists, they are worse when the faculty committee conducting the search includes women. As good as things are for Democrats and ACLU members, they are better when the search committee includes women. Yancey, of course, isn’t the first to come across the phenomenon that “females are more progressive than males.” He cites a slew of other studies that document this, but his survey gives that finding a new and rather disturbing edge.
Yancey has no real explanation to offer for the greater admitted bias of female academics. My guess is that it has something to do with the prevalence of academic feminism which in many of its versions offers a welcoming rationale for bias. “The personal is political,” and all that. But clearly this is a topic that awaits the illumination of further research.
Compromising Scholarship concludes with chapters in which Yancey reflects on the effects of bias in the social sciences and some suggestions on what might be done to offset them. The main effects he worries about are the loss of good, competing ideas and the loss of public legitimacy, as people rightly conclude that the declarations of social scientists often reflect bias, not science. His corrective measures strike me as anemic. They are basically expansions of programs that originated to combat racism on campus: programs to promote diversity, curb “bias speech,” and raise self-awareness. But it would be asking a great deal to expect a scholar to have a ready solution to a problem as large and as entrenched as the institutionalized biases Yancey has documented here.
When Neil Gross and his colleagues came out with two new papers a few weeks ago, their results were greeted with fanfare in the academic press. I haven’t noticed that yet with Yancey’s book. Of course, Gross was delivering what looked like welcome news: exoneration from the charge of liberal bias. It is perhaps in the nature of that bias that a book showing the bias is real is not likely to get such a jubilant reception. But Compromising Scholarship is a very important addition to our understanding of how bias works in academe. It deserves serious attention.