t will surprise few that college campuses in America lean left. But does it matter? Research suggests that the left’s takeover of academia is hurting the quality of new science produced on campuses. And new evidence hints that students’ educations may be harmed as well.
What are the Dangers for Science?
One of the primary missions of universities is the creation of new knowledge through scientific research. Is a political imbalance a threat to this mission? New York Senator Daniel Patrick Moynihan once warned that “Ideological certainty easily degenerates into an insistence upon ignorance.” And Jose L. Duarte, Jarret T. Crawford, Charlotta Stern, Jonathan Haidt, Lee Jussim and Philip E. Tetlock argue that political dominance threatens the self-correcting tendencies of science. They use the field of social psychology to illustrate three main dangers of ideological dominance: skewed topic selection, biased methodologies and misinterpreted results.
Skewed Topic Selection
University faculty have unparalleled freedom to determine what topics they will research. Even if potential topics hold differential appeal to researchers of differing political leanings, this aspect of academic freedom will ensure that all relevant topics will be adequately researched, so long as the faculty is ideologically diverse. But, if a field is politically imbalanced, academic freedom can lead to distorted research topic selection. For example, left-leaning researchers may be more likely to investigate prejudice against women and minorities, while right-leaning researchers may be more likely to investigate prejudice against Christians. If left-leaning researchers are dominant, prejudice against their favored subpopulations gets explored extensively, while prejudice against other subpopulations is neglected.
Methodologies can also be warped by political uniformity when biased views are embedded in theories and assumptions. Duarte et al. cite a survey in which participants agreed that the phrase “the efficacy of hard work” was a “rationalization of inequality.” Another group regarded disagreement with the statement, “If things continue on their present course, we will soon experience a major environmental catastrophe” as denial of ecological reality. As Duarte et al. point out, “the core problem with this research is that it misrepresents those who merely disagree with environmentalist values and slogans as being in ‘denial.’”
The combined effect of skewed topic selection and biased methodologies often yields unreliable conclusions. For example, as the researchers point out, “a long-standing view in social-political psychology is that the Right is more dogmatic and intolerant of ambiguity than the Left,” but this conclusion is probably the result of only looking at topics where the right is more dogmatic than the left.
The end result of these three issues is a skewed research environment and less reliable scientific output.
One-Sided Political Dominance of a Field May Hurt Students’ Careers
But what about the educational mission of universities? Given their extensive education, faculty members should be capable of presenting and teaching views with which they personally disagree. But this is much less likely to occur in fields that are dominated by one political party. Presenting a conservative view is much easier if you interact with conservatives or have to contend with their ideas during the research and publication process. If a field is tilted heavily towards progressives, it is all too easy to remain in a progressive bubble and rarely engage with conservative views at all. This can lead to unsympathetic caricatures.
Consistent with this hypothesis, new evidence suggests that politically imbalanced academic fields may harm students’ career prospects. In 2018, Mitchell Langbert compared the voter registrations of faculty at top liberal arts colleges and found that Democrats outnumbered Republicans by more than 10 to 1. But the results varied considerably by academic field. In engineering, there were 1.6 Democrats for every Republican, but in sociology there were almost 44.
I compared the ratio of Democrat to Republican professors in a number of fields with the share of students graduating from a program that passed a debt-to-income test called Gainful Employment Equivalent. This test is based on the old gainful employment regulations that the Obama administration imposed primarily on for-profit colleges (updated to account for differences in the definitions of programs, earnings and debt—see here for more details). Gainful Employment Equivalent determines whether the Obama administration would have tried to shut a specific program down if it were offered at a for-profit university.
The results are quite startling.
Academic fields in which Democrats outnumbered Republicans by less than 6 to 1 fared well, with over 80% of students graduating from programs that passed the debt-to-earnings test. But, once you exceed a 6:1 ratio, performance declines quickly. In philosophy, Democrat professors outnumber Republican professors 18:1, and only 45% of students graduate from a program that passes the debt to earnings test. The pattern continues as fields become more politically dominated by Democrats. Art has a 40:1 ratio and 16% of students graduate from programs that pass, and religion has a 70:1 ratio and only 4% of students graduate from programs that pass. Sociology is a relative outlier, with a ratio of 44:1 and passage rates of 50%. But, while its passage rate is better than expected, given the dominance of Democratic professors, it still falls far short of the 80% passage rate of the more balanced fields.
These results only show correlation, not causation. There isn’t yet enough data to determine whether being more politically imbalanced causes fields to fail the debt-to-earnings test, or whether fields that fail the debt-to-earnings test just happen to also be politically unbalanced. Or perhaps some unrelated factor or factors is driving both results. For instance, fields with low ratios and good performance on the debt-to-earnings test tend to be more quantitative. Perhaps there is stronger demand for graduates of quantitative fields in the labor market and perhaps people attracted to quantitative fields tend to be more ideologically balanced. This explanation, which is consistent with both Robert Nozick’s numbersmiths vs. wordsmiths distinction and Joseph Schumpeter’s theory of how education erodes support for capitalism, could explain the correlation pictured in the chart, without implying causation.
But, for the students involved, it doesn’t much matter whether these results are due to correlation or causation. If they value their career prospects, students should be wary of severely politically imbalanced fields. (For the benefit of students set on choosing one of those fields, the Texas Public Policy Foundation has created a web tool at https://www.texaspolicy.com/college-earnings-and-debt/, to allow students to find programs that pass the debt to income test, even in riskier fields.)
Can Anything Be Done?
If Democratic dominance of a field lowers the quality of research (probable) and of education (possible), what should be done? Further research to distinguish correlation from causation would certainly be valuable. But there are enough warning signs that the left’s takeover of academia is harming the quality of higher education that we should at least try to decelerate the takeover, by fighting bias against conservatives in the faculty hiring process. But is that possible?
In some fields, the answer is no. Since existing faculty are the primary evaluators of applicants for new faculty positions, when political imbalance reaches high proportions, a field’s leftward drift is virtually unstoppable. Suppose that everyone is either a Democrat or a Republican, and that we have a five-member search committee in a field like social psychology, which has a 14:1 Democrat to Republican ratio, according to Duarte et al. A Republican job candidate would have an over 70% chance of facing an all-Democrat search committee, which doesn’t bode well for her chances. And it gets worse. 82% of Democratic social psychologists “admitted that they would be at least a little bit prejudiced against a conservative” job candidate. This means that a conservative job candidate in that field would have less than a 0.001% chance of encountering a search committee none of whose members were prejudiced against her.
It is difficult to see how a field can fix itself when a candidate with minority views faces such a small chance of not being actively discriminated against. Either change will be imposed on these fields by external forces, or the fields will continue their leftward drift.
It is astonishing that so many Democratic social psychologists admitted prejudice against conservative job candidates. Readily admitting blatant bias indicates a frightening degree of confidence in its acceptability. Such obvious bias threatens to destroy social psychology’s credibility on any topic that is remotely political.
But not every field exhibits this degree of bias. Those that don’t should be protected from the rapidly increasing numbers of diversity, equity and inclusion policies that can interfere with the faculty hiring process. Diversity, equity and inclusion are admirable goals to the extent that they seek to increase diversity by helping subpopulations overcome unfair barriers. The problem is that some subpopulations are more equal than others. While diversity, equity and inclusion administrators seek to tear down barriers for some groups based on race, class and gender, these same administrators are erecting barriers to ideological diversity.
The most visible of these is the growing trend of requiring diversity, equity and inclusion statements for faculty hiring and promotion. Such statements would be unobjectionable if they merely required new faculty to confirm that they will not discriminate against legally protected groups. But, in reality, these statements are being used as ideological litmus tests. As Jerry Coyne notes, the statements are graded and to pass “you have to swear fealty to an ideology.” John Cochrane calls them the new “loyalty oaths” and Abigail Thompson wisely observes that the “idea of using a political test as a screen for job applicants should send a shiver down our collective spine … Imposing a political litmus test is not the way to achieve excellence.”
The University of California at Berkeley’s recent faculty hires illustrate the danger. Typically, a candidate for a faculty position would have her research assessed by a departmental search committee. As we just saw, this faculty-driven process can replicate and entrench political dominance. But Berkeley has greatly exacerbated this process. In what they have admitted is a “dramatic change” from the typical hiring process, diversity, equity and inclusion administrators have been given veto power over which candidates the departmental hiring committees are able to consider. The scale of the resulting purge would make Stalin blush. Of 893 nominally qualified candidates, 679 were eliminated solely due to insufficiently woke diversity, equity and inclusion statements. In other words, Berkeley used a political litmus test to eliminate over three-quarters of the applicant pool.
So far, allowing such intrusions by diversity, equity and inclusion administrators into the faculty hiring process has been apparently voluntary at the departmental level (though the fact that some departments are already agreeing to such ideological litmus tests is disturbing). But diversity, equity and inclusion statements are increasingly being required across campuses, and, if Ronald S. Sullivan Jr. is correct that “angry demands, rather than rigorous arguments, now appear to guide university policy,” campuses are just a few student protests away from making these litmus tests mandatory.
Policymakers should prevent the imposition of these litmus tests on departments that prefer to choose their faculty based on their qualifications rather than their ideology. However, no one should be under the impression that this would be a panacea. It is more akin to the advice, if you find yourself in a hole, stop digging. That’s certainly a good first step, but how do we climb out of the hole altogether? That’s a much tougher question.