Over the weekend I read through the paper that prompted KC Johnson's column in Inside Higher Ed (discussed by Jim last week). Leaving aside the data issues that the authors raise there, I want to explore the "alternative hypothesis" that the authors there raise as the nondiscriminatory explanation for the observed data. I will follow the convention of referring to all non-left/liberal ideologies as "conservative," however inapt that may be.
Here's the author's alternative hypothesis, followed by my analysis:
[Continue reading under hidden text]:
We offer self-selection as the likely culprit. First, there may be a rural/urban divide driving the relationship. Conservatives may want to live in communities whose ideological climate is more consistent with their own belief structure. Thus, given the strong correlation observable between the metropolitan density of a particualr county and the mean conservatism of its citizens, it would not be surprising if conservatives, academic or otherwise, prefer to work in smaller, more rural areas. ***
Second, regional selection affects hiring... It is no secret that Midwesterners and, especially, Southerners are more conservative, more religious, and less Jewish than Northeasterners. ***
Third, many conservatives may deliberately choose not to seek employment at top-tier research universities because they object, on philosophical grounds, to one of the fundamental tenets undergirding such institutions: the scientific method.... Furthermore, cultural conservatism, as revealed in antipathy toward gay rights, the women's movement, and abortion rights (among other things), has been shown to stem in large part from an embrace of Christian fundamentalism as a dominant worldview. Fundamentalism, by definition, is an absolutist, "faith-based" allegiance to a particular dogma, the veracity of which is considered beyond question or argument. Such worldviews are (again, by definition) antithetical to the philosophy of science, which permotes reason and evidence as the determinants of truth. Challenging entrenched dogma is the essence of science.... In other words, the faith-based reasoning of Christian fundamentalism (and by extension, of most socio-cultural conservatives) is essentially incompatible with the mission of contemporary researc universities. So, in sum, we are suggesting that the relationships [identified] might be a spurious function of self-selection based ona fundamentalist antipahy toward the scientific method and other approaches to revealed "trust"--precisely the busines of "top-tier" research universities.
The authors offer no evidence to support these explanations (they appropriately protest that the authors of the original study have not made their data available to researchers). On examination, however, none of them seem to be persuasive and available evidence tends to contradict their proffered hypothesis. Consider each element in turn.
1. Rural/Urban Divide: The authors argue that conservatives prefer to live in more rural areas and so will be disproportionately found at schools in such areas. If this is so, then an easy test of the hypothesis presents itself--my beloved alma mater Dartmouth College is one of two rural institutions in the Ivy League (along with Cornell). It is also in New Hampshire, historically the most conservative of the northeastern states. If the authors are correct about a rural preference, then we should expect to find more conservatives on the Dartmouth faculty than at urban-situated Ivy League schools, such as Harvard, Columbia, Penn, or Brown. A recent examination of party identification, however, finds that 66% of Dartmouth professors are Democrats and the ratio of registered Democrats to Republicans is 12.5 to 1, comparable to the ratios at urban Ivy League and other elite institutions. This is just one data point, of course, but it casts doubt on the idea that conservative academics have an unusually pro-rural preference. On the other hand, the University of Chicago, a rare institution with a historic reputation for a conservative presence (Strauss, Law & Economics, etc.) suggests that there are counter-examples in the opposite direction as well.
2. Regional selection: The authors argue that the South is disproportionately conservative, and as a result, conservatives might self-select for southern schools. Again, the testable hypothesis here is that elite schools in the South should attract a disproportionate number of faculty conservatives relative to similarly-prestigious schools in more liberal areas of the country. The only report I have seen on this is of Duke University, unquestionably one of the most prestigious institutions in the South. According to one survey of eight of Duke's humanities departments, however, Duke has 142 registered Democrats and 8 registered Republicans (17.75 to 1). The Duke History Department pitched a 32-0 Democratic shutout. Again, this is just one sample and only of humanities departments, but it lends little support for the conclusion that conservative professors are disproportionately drawn from and prefer to remain in the South. More generally, although I have seen no data on this, do we really think that rural midwestern small colleges (the three factors the authors identify) like Oberlin, Grinnell, and Kenyon are overrun with conservative faculty?
3. Willingness to Apply the Scientific Method: The fact that the authors would propose this explanation seemingly with a straight face is evidence of the straw-grasping going on here. But I will treat it as a serious hypothesis for the sake of argument. The authors argue that conservatism draws from fundamentalist religion and this makes them hostile to the scientific method. Again, this is subject to testing. If conservatives are hostile to the scientific method, then the disparity in ideology should be greatest in those fields where the scientific method is strongest--mathematics and hard sciences. By contrast, the gap should be narrower where the scientific method is least relevant, such as in the humanities.
Again, the available data don't seem to support this hypothesis. In fact, the ideological gap is much narrower in those fields where the scientific method is most relevant. Consider this summary of one such study:
I looked at the study itself and found the following for Liberal/Conservative identifications:
Mathematics: 4.1 to 1
Physics: 6 to 1
Chemistry: 2.2 to 1
Biology: 4.4 to 1
Computer Science: 2.8 to 1
Engineering: 2.6 to 1
Economics: 1.4 to 1
By contrast, humanities and social science departments (except economics, which is probably more scientific than most of the other social science departments) have the following divides:
English Lit: 29 to 1
History: 7.7 to 1
Philosophy: 16 to 1
Theology/Religion: 16.6 to 1
Political Science: 40.5 to 1
Sociology: 8.6 to 1
Psychology: 10.5 to 1
So, yes, a gap does exist in the sciences, and it is a rather large one, but not nearly as profound as in the humanities and social sciences (aside from economics).
So, the ideological divide is much narrower in the fields where the scientific method is used (including Economics), and widest where it is most absent.
More fundamentally, the authors are simply incorrect to argue that the religious beliefs of conservatives make them unable to apply the scientific method. As Larry Summers learned the hard way, the modern-day Lysenkoism of political correctness is a faith just as strong, if not stronger, than religious faith. Summers great heresy was to suggest that certain faith-based tenets of the modern academy could be subjected to testing by the scientific method. Make no mistake about it--the orthodoxy of political correctness is just as at odds with the scientific method as traditional religious belief. The only difference, of course, is that traditional religion is shunned in the modern academy, while political correctness is the academy's official religion.
The author's blind spot to this point is almost laughably provided by their own analysis of the situation--"cultural conservatism, as revealed in antipathy toward gay rights, the women's movement, and abortion rights (among other things), has been shown to stem in large part from an embrace of Christian fundamentalism as a dominant worldview." Give me break--if antipathy to these issues comes from religion, from where does the support for them come? Are the authors seriously suggesting that "science" supports gay rights or abortion rights and that opposition is faith-based? Again, we are back to recognizing that support for these rights is grounded in politically-correct faith just as the antipathy to them is as well.
The authors also claim, "It is difficult even to imagine ideological discrimination occurring at the point of hiring." This naivete again demonstrates the authors blind spot occasioned by their own narrow world view. There are a multitude of ways in which ideological discrimination manifests itself in hiring. The most obvious is simply the degree of skepticism that incumbent faculty apply to a given scholar's work. If they disagree with the ideological conclusions of the work, they approach it with greater skepticism and a higher burden of proof, and thereby it is easier to conclude that the analysis is flawed or incomplete. Again, this difference is reflected in the fact that the more subjective subjects (Philosophy, History, English, etc.) have greater ideological disparities than less-subjective subjects where standards of scholarly rigor are better-established and have an independent integrity that separate the craftsmanship of the field from the conclusions. (Some have observed that within political science itself, for instance, the more scientific quantative researchers have less ideological bias as well because of the indepdent standards of analysis applied there.) Other biases easily creep in as well--does the candidate do work that is "relevant" to the interests of the department, "collegiality," or common research interests. To say that "it is difficult to see how ideological bias" could creep into hiring is simply naive and perhaps just evidences the lack of self-awareness by the researchers themselves.
Even if this is self-selection, this is not necessarily responsive--when the elite academy is confronted with other examples of "underrepresented" interests, they do not simply throw up their hands and complain of a shallow talent pool. Instead, at Columbia for instance, the diversity committee is "tasked with finding ways to strengthen the pipeline bringing women and minority students into the University's undergraduate, graduate, and postdoctoral programs" and not merely take what the pipeline produces.
Based on my personal experience having known many bright students in many different fields (I meet many undergraduates and graduate students through IHS Seminars at which I lecture), I would say that if there is self-selection here, it is of precisely one type--libertarian and conservative students self-select out of pursuing an academic career because they are well aware of the political obstacles that will be placed in their way. They know that they will confront ideological bias at every stage of their careers--grad school, grad school mentoring to help get jobs, entry-level jobs, and tenure. Given the numbers reported by Klein and others, they are clearly acting rationally in refusing to invest 5 years of their lives to get a PhD to try to roll this stone up that hill.
Finally--and I'm less confident about this--conservatives may be disproportionately turned off by the fundamental "unseriousness" of the modern academy. Conservatives may simply prefer the real world, with its mechanisms of accountability, merits-based determinations, and focus on solving real problems. The emptiness and triviality of so much modern scholarship (especially in the humanities) and the marbeling of every element of academic culture with the burdens and distractions associated with running the modern university--political correctness and its restraints on free inquiry, the whole edifice of the diversity machine and all it carries with it. The upward struggle to persuade colleagues to judge job candidates fairly and on the merits, rather than through the ever-present lens of political orientation and identity/diversity politics.
So the self-selection, if there is one, may be more along the lines of Michael Barone's distinction between "hard" and "soft" America--perhaps conservatives are more prone to self-select into the "hard" America of the private sector, where accountability is stronger and individuals are more likely to rise or fall on their own individual merits, rather than trying to survive in the bizarre ecosystem of the modern academy.
In conclusion let me add a thought--it seems utterly absurd that people are still making uninformed armchair speculation about the causes of the prevailing ideological imbalance in the academy. Is it self-selection? Conservatives are greedier? Conservatives are dumber? When it comes to addressing the issue of other "underrepresented minorities" on college campuses, the record overflows with high profile blue ribbon panels of leading scholars and administrators. No stone is left unturned and no penny left unspent to try to determine why women are "underrepresented" in teaching math and science, or the underrepresentation of minorities. I think maybe it is time to take even a small percentage of those tens of millions being spent at places like Harvard and Columbia and perhaps do a study of the causes of the ideological disparity in the academy, rather than simply speculate and pontificate. At the very least, such a study would eliminate some of the more preposterous hypotheses (such as the idea that conservatives generically like money more than liberals or that conservatives lack the intellecutal frame of mind to succeed in academia).