A recent edition of the estimable journal Perspectives on Psychological Science (September 2012) ran something you almost never see in academic literature: an exploration of political bias in academe. In fact, the journal devoted more than half of that issue to six papers exploring the prevalence of political bias in the field of experimental psychology and its effects upon research.
The issue of bias in the profession has been on the table since a now-famous meeting of the Society of Personal and Social Psychology in 2011. Social psychologist Jonathan Haidt asked how many conservatives were in the room—and three hands went up in the group of about 1,000 people. Now a journal is pursuing research that helps explain that tiny showing.
The Perspectives papers were, in my view, too short to investigate the problem in any real depth; however, they certainly raised the salient issues.
The lead-off paper (by Yoel Inbar and Joris Lammers, both professors of social psychology at Tilburg University in the Netherlands) reports on the authors’ survey of personality and social psychologists in academia (a professional category I will call PSPs). The authors conducted two surveys aimed at determining how liberal PSPs are on average, how accurately PSPs can observe the ideology of other PSPs, and how inclined leftist PSPs are to be hostile toward or discriminate against perceived rightist/conservative PSPs.
On the first issue, the respondents were on average disproportionately on the left. On economic issues, 63.2 percent rated themselves to the left of moderate (i.e., 1, 2 or 3); on foreign policy issues, 68.6 percent were to the left of moderate; and on social issues, a massive 90.6 percent were to the left of moderate.
On the second issue, Inbar and Lammers found that their respondents overestimated the liberalism of their colleagues. This led them to focus their second survey on determining whether the reason why conservative PSPs don’t recognize other conservatives is that conservative PSPs hide their views out of fear of discrimination.
In that survey, PSPs were asked if they perceived a hostile climate in the field towards researchers of their political orientation, whether they were reluctant to express their beliefs to colleagues out of fear of retaliation, and whether they thought colleagues would discriminate against them if their political beliefs were known.
The composite of the three questions showed quite telling results. Conservatives felt they faced more ideological hostility than did moderates, and far more than did liberals.
The authors conclude that the reason it is hard to identify the relatively few conservative PSPs is that, “Hostility toward and willingness to discriminate against conservatives is widespread.” Their data show that one in three PSPs would discriminate against conservatives in hiring, and many also said they would discriminate against their conservative colleagues in publication and grant applications.
Three of the other papers broadly supported the Inbar/Lammers paper. Richard Redding (professor of law and of psychology at Chapman University) argues that the Inbar/Lammers study adds to the already extensive literature documenting bias against conservatives in the field of psychology in particular and the academy generally.
Redding surveys that literature and notes that the pervasive discrimination against conservatives starts as early as in application for grad school. He argues that this probably accounts for any “self-selection” of conservatives away from the field. He notes that the ratio of liberals to conservatives in the field went from 3 to 1 in the 1960s to 20 to 1 today.
Redding then asks why, if diversity is so important in psychology that it is central to the American Psychological Association’s own stated ethical code, this doesn’t incline the profession to take steps to include conservatives in the field. After all, the same arguments given for demographic diversity—viz., that a person’s ethnicity or culture is important for his identity, that people who are different ethnically or culturally are often victims of discrimination, and that they bring different experiences and ideas that invigorate research—apply as much or even more to ideological diversity.
This is especially true when you consider that psychologists do a lot of research that bears on public policy—such as programs promoting diversity, such as affirmative action—and that psychologists are no less subject to confirmation bias than anyone else. Redding notes, “It is no coincidence that virtually all the research and commentary by psychologists is liberal leaning or the Society for the Psychological Study of Social Issues advances liberal rather than conservative causes….”
Having conservative scholars critique research that appears to support liberal policy positions will make it stronger and the reverse would be true as well. And it is likely that conservatives would address research issues that liberals would avoid, and devise fruitful new ones as well.
A second supporting article is by Philip Tetlock (professor of psychology at the University of Pennsylvania). Among other points, he alludes to those whom he cheekily calls “system justifiers.” These are writers who claim that conservatives are so scarce in psychology because they are by nature “anti-science” and therefore avoid scientific disciplines. (This is a view pushed by Chris Mooney in his recent writings).
Tetlock responds by saying that this is a sweeping generalization (in that it fails to distinguish religious conservatives from national security conservatives and libertarians), it is unsupported by substantial evidence, and it isn’t applied to (say) Marxist PSPs, who are likely to be ideologically resistant to historical evidence on the failure of state-planned economies and the heritability of IQ.
Tetlock adds, “No one knows how many research programs social psychologists have failed to launch” because of the ideological bias in the field.
He doesn’t give any examples of the impact of ideological bias on research, but I would suggest a couple. Conservative and libertarian psychologists would be more likely to investigate the psychological impact of racial affirmative programs on the white males who are discriminated against (as sociologist Fred Lynch did in his 1991 book Invisible Victims: White Males and the Crisis of Affirmative Action). And they would be more likely to look at the downside costs of immigration and ethnic diversity, as sociologist Robert Putnam did in Bowling Alone: The Collapse and Revival of American Community).
Finally, Lee Jussim (professor of psychology at Rutgers University) trenchantly challenges those who object to the Inbar/Lammers paper on “purely scientific” grounds: “What could disconfirm your belief that social and personality psychology are objective fields whose conclusions are entirely untainted by political biases?” Given that the social sciences often attract people who want to affect policy, a convincing answer is unlikely.
Jussim gives an example of evident liberal bias. A 1983 study (by J. M. Darley and P. H. Gross) that purported to show that stereotypes function as self-confirming hypotheses—which fits into the progressive narrative of pervasive “inegalitarian” bias in our culture—was cited nearly 800 times in later papers. Yet a 1995 paper (by R. M. Baron, L. Albright and T. E. Malloy) reporting two failed attempts at its replication was cited less than 30 times in later papers.
The two pieces criticizing the Inbar/Lammers survey are by Linda Skitka (psychology professor at the University of Illinois at Chicago) and Deborah Prentice (psychology professor at Princeton). Skitka rejects the Inbar/Lammers piece for sample selection issues (such as for not excluding graduate students from the sample) and the framing of the questions (such as asking whether the respondents would discriminate against a grant application for taking a conservative perspective—but not asking whether they would against one that took a liberal one). She accuses the authors of confirmation bias.
Prentice also holds that the Inbar/Lammers piece fails to provide solid evidence that conservatives are underrepresented in the personality-social psychology profession due to discrimination, when it might simply be that personality-social psychology attracts people with the “goals and values” that liberals have.
If psychologists were 95 percent male (parallel to the 20-to-1 ratio of liberals to conservatives according to Richard Redding), one doubts that Ms. Skitka and Ms. Prentice would ask for a statistically robust survey before concluding that discrimination against women was involved.
I commend Perspectives on Psychological Science for addressing the issue of bias in its field. Given the prevalent use of psychological research in justifying various progressive liberal public policy measures, the fact that there is pervasive political bias in the field is a real problem.
However, I don’t expect to ever see any of the top journals in my own field (philosophy) raise the same issue, despite the fact that philosophers consider themselves bound above all others to raise deep and disturbing issues.