Liberal Academic Self-Selection

April 2011

Few aspects of faculty demographics generate more attention than their politics. Why is it, many want to know, that professors are far more likely than the general public to be liberal? Many theories have been put forward, including the view (much discussed in conservative circles) that academe is hostile to conservatives and tries to either weed them out or convert them.

Two studies being released today provide more evidence that bias is not the cause -- and the studies provide some additional evidence to back the theory (put forward last year by one of the authors of the new work) that "self-selection" is the primary reason so many academics are liberal. In brief, the self-selection idea holds that some professions have become "typed" in American society in various ways that may relate to gender or class but could also relate to politics. Academe is seen as more liberal, so liberals are more likely to identify being an academic as something to which they aspire. The argument is significant because it does not contest the lopsided political nature of many faculties, but also suggests that higher education is open to those conservative scholars who want careers there.

One of the new studies was an "audit" of the reactions of graduate program directors to initial inquiries from potential graduate students who said something to indicate their political leanings. The research found no evidence of bias.

The second study used a longitudinal database that had information on how thousands of individuals thought about politics and the launch of their careers. This study found that those who pursue academic careers are far more likely to be liberal than conservative -- again countering the idea that conservatives are being turned away from doctoral programs, or that a leftward shift is a price of success in Ph.D. programs.

"These studies together make a very strong case that most of the liberalism among professors is the result of self-selection," said Neil Gross, associate professor of sociology at the University of British Columbia, who is among the authors of last year's self-selection work and both of the new studies. (Gross works in Canada, his work on these topics is focused on the United States.)

While Gross sees the new studies providing important backing for the earlier research, not everyone (as he is the first to admit) is likely to be convinced. The studies are only being released now and have yet to be widely reviewed. But Peter Wood, president of the National Association of Scholars, said that they don't rule out bias. He said one of the studies suggests that self-selection is part of the reason that faculties lean to the left, but that there is no way of knowing from available evidence that liberal academics are not discouraging conservatives from joining them.

The Audit

The study of how graduate directors respond to inquiries was conducted by Gross; Ethan Fosse, a graduate student at Harvard University; and Joseph Ma, an undergraduate at the University of British Columbia.

Posing as undergraduates getting ready to apply to doctoral programs, they sent e-mail messages to graduate program directors in top departments of sociology, political science, economics, history and English. The inquiries were similar in describing their academic preparation, their undergraduate institutions, and their interest in applying. Some of the e-mails made no mention of politics, but some mentioned having previously worked on either the Obama or the McCain presidential campaigns.

The political references were brief, and followed by a phrase about the campaign work having been "a learning experience," so as not to suggest that the candidates were ambivalent about academic careers. (The authors acknowledge that working on the Obama and McCain campaigns might not be perfect proxies for liberal and conservative. "We worried that a stronger conservative prompt, such as being a George W. Bush supporter, might -- if claims about the extent of hostility to conservatism in academe are true -- lead some respondents to question the legitimacy of the e-mail," they write.)

The researchers then had independent (and politically mixed) observers rate the responses from the graduate directors on frequency, timing of replies, information provided, emotional warmth and enthusiasm. In a few cases, the researchers found "traces" of a political impact, but "no statistically or substantively significant evidence of bias."

The paper notes a number of limitations to their study. Most significantly, the test looked at an initial stage of contact between candidates and departments, not the crucial admissions decision, when bias might also surface. Further, they note that all of the publicity over alleged political bias might make graduate directors censor themselves and not reveal their biases.

At the same time, however, the authors cite "research on stereotypes and social biases in general, as well as on political bias and the associated affect specifically [that] suggests that, when present, biases operate primarily in the domain of automatic cognition. Since responding quickly to prospective student e-mails is, in the language of 'dual process' models in psychology, more a matter of heuristic than systematic processing, one would expect political biases likely to affect a range of judgments to show up in our results." In sum, the authors write that "if political bias toward graduate students were robust in the fields we studied, our methodology would very likely have detected it."

The paper notes the ethical issues involved in deceiving the graduate program directors, but argues that they are justified. "[P]eople on both the right and left consider the issue of political bias and discrimination in higher education to be an important topic -- conservatives think it exists and is unfair, liberals tend to deny it but worry about the effects on academic freedom of conservative allegations." Further, the paper says that "an audit study (requiring deception) is one of the best ways of gaining empirical traction on the matter" and "that it is not asking much of subjects to respond to two e-mails (though we realize it is asking something); and that there are few risks to subjects from participating in the study."

The Longitudinal Study

The second study is by Fosse, Gross and Jeremy Freese, chair of sociology at Northwestern University. This study makes use of the Add Health database, which was created to track the long-term health behaviors of 90,000 adolescents, but which also includes questions about political orientation and educational/career plans. The authors realized that although this database was not created to examine the question of why professors tend to be liberal, it had the potential to provide some answers. At various points (well after adolescence) the participants were asked questions about political views and about whether they were headed to graduate school.

One "wave" of questions took place when the respondents were aged 18-26 and another when they were 24-32. The idea was to determine whether political orientation during the period when future professors are likely finishing their undergraduate educations might be a factor in whether they subsequently were enrolled in graduate school. (Those who did not complete a bachelor's degree were excluded since they could not go to graduate school.)

What the study found was that those seeking a doctoral degree in the 24-32 age group were clearly more liberal than the population as a whole. In the entire pool, 35 percent identified as liberal or very liberal, while 49 percent of those seeking doctorates did so. In the entire pool, 23 percent identified as conservative, while only 18 percent of doctoral seekers did so.

These figures match (generally) data on the political leanings of young professors. "These numbers strongly suggest that much of professorial liberalism is indeed a function of who goes to graduate school: filling job openings in academe with a random draw from the pool of graduate students would still produce a distinctly left-leaning occupation," the study says.

Further, other data show that while a significant minority of those studied became more liberal in their doctoral programs, so did a significant number of those who didn't go to graduate school. In addition, doctoral students were slightly more likely than those who stopped their education after their bachelor's degrees to become more conservative than they had been earlier in their lives. These findings generally cast doubt on the idea that professors are liberal because they are socialized that way in graduate school.

There are some inconclusive results in the analysis about whether some personality traits (not linked to politics) may make some people more likely to go to graduate school. But notably, the study found no relationship between either materialism or early marriage and a disinclination to go for a doctorate. (These findings rebut theories put forward by some observers that conservatives' desire for more money than young professors tend to earn, motivated either by greed or family obligations, explains the scarcity of right-wing academics.)

Self-Selection or Bias?

Whatever the results of these studies, many people remain bothered by the lopsided nature of professorial politics. The annual meeting of the Society for Personality and Social Psychology this year was dominated by a talk charging that the disciplines represented in the organization may have a bias against conservatives, The New York Times reported.

Jonathan Haidt of the University of Virginia polled the audience of 1,000 scholars and asked by shows of hands how many of them identified themselves in various political ways. He found that about 80 percent called themselves liberals, a few dozen said that they were centrists or libertarians, and only three said they were conservatives. "This is a statistically impossible lack of diversity," Haidt said, given that 40 percent of Americans identify as conservatives.

Wood, of the National Association of Scholars (a group that has criticized what it considers liberal bias in academic life), found particular fault with the study of graduate admissions directors. He questioned whether McCain was a good proxy for conservative leanings, saying that the senator "is virtually nobody's idea of a standard bearer of conservatism."

But more broadly, he said that "they were looking for actionable bias in entirely the wrong place." He argued that most graduate program directors would only discourage someone who was "manifestly inappropriate" by virtue of a poor academic record, no prospect for a bachelor's degree or similarly clear lack of qualifications. Bias in admissions, he said, would more likely come later, when admissions committees vote on candidates.

Wood called the longitudinal study "the far more substantial paper," and he said it does establish a role for self-selection in the political make-up of the professoriate. But he said that there is still "a gap" in the thinking that this rules out bias. Even if self-selection is "a primary driver for the liberalism" of faculty members, that does not mean bias does not exist, he said. "They have set up a false opposition."

It also may be the case, Wood said, that faculty members have a responsibility for the self-selection going on. "There's a kind of chicken and egg problem that they have done their best to avoid," Wood said. "Given today's university, you'd have to be pretty obtuse not to understand that going into the field is going to mean a lifelong association with professions that are dominated by liberal individuals and liberal ideologies."

Gross noted that the studies acknowledge their limits, but said that it was significant that at the same time that more evidence is emerging for the role of self-selection, efforts to find systematic bias (such as the inquiries to graduate directors) were coming up empty. He also acknowledged that nothing in the research he and others have done denies that some conservatives may feel that academe is "unwelcoming" to them as a profession.

But "unwelcoming," he noted, "is still a self-selection story, as opposed to an exclusion story."

One response by those who want to see more political balance in the make-up of faculties, Gross said, would be to take these studies and make the point to talented, right-leaning students that the door is open to them. More efforts could be made to create and sustain conservative intellectual efforts, he said.

But even if that happens, Gross predicted that claims about liberal bias in academe would not go away. "That line of argumentation serves a pretty important function for the conservative movement," he said. Modern conservatism has cast itself as a populist movement, Gross noted, and populism "requires a bashing of elites." Conservatives have a tough time bashing economic elites, Gross said, "so there's been a strong need to find alternative elites to bash."

From William F. Buckley Jr. on, the solution has been "to focus on cultural elites," such as faculty members. Gross stressed that he was not suggesting "a conspiracy" by conservatives, but just stating the reality that this line of argument is one that has been made for decades, with considerable success. "It provides a sort of collective identity for educated conservatives, that sense of feeling excluded," he said. The challenge posed by his research, he said, is that it suggests that choice is the primary reason liberals are more common in academe than are conservatives.