Conservatives are the REAL victims of discrimination

Underrepresentation and discrimination in the academy

In a recent barpod episode,

Jesse Singal

and Katie Herzog discussed the underrepresentation of conservative scholars in academia, particularly in the social sciences. For example, they cited survey results from 2011 which suggested that only 9.4% of academics in social and personality psychology self-identified as moderate or conservative and that nearly 40% of participants said that “in choosing between two equally qualified job candidates for one job opening, [they] would be inclined to vote for the more liberal candidate (over the conservative)”. While the left-leaning tilt on campus is certainly not new, “[w]hen the Carnegie Foundation conducted its faculty survey in 1999, it found that a mere 12% of professors were conservatives, down from 27% in 1969”, indicating that it’s only gotten more extreme over time. In line with those results, Pippa Noris has found that “the youngest cohorts of political scientists are significantly more liberal-left in their values than the oldest cohorts.”

Jonathan Haidt has argued that what he views as a “statistically impossible lack of diversity” with respect to political ideology among social psychologists, and the potential downsides of such a lack for the quality and breadth of the research they produce, justifies affirmative action for conservatives. But there’s disagreement about the degree to which this underrepresentation reflects real discrimination vs. self-selection (due to perceived discrimination) vs. differential desires or abilities. While personality differences (and potentially differences in average IQ) between liberals and conservatives would be expected to contribute to underrepresentation, as Stephen Teles notes “The magnitude of the difference between the number of faculty on the left and right is simply too large to be accounted for by inherent differences. Furthermore, the decline of conservatives in academia (albeit from a low base) raises the problem of explaining change with a constant.”

As Jesse and Katie discussed in the episode, the debate around ideological diversity on campus rhymes remarkably well with that around racial diversity (both on campus and in the workplace), but with the scripts largely flipped. Here we have conservatives and moderate academics noticing a significant difference between the representation of a group in academia vs. in the broader population and suggesting that this is evidence of a problem, potentially one which requires affirmative action to redress. Meanwhile, others on the left explain this underrepresentation as resulting from differences in group attributes. It’s likely true that much of the underrepresentation of conservatives in academia is due to a “pipeline problem” where conservatives are less likely to be in the applicant pool in the first place, and thus, as Teles says “academics may have a willingness to discriminate, but because they are presented with so few conservatives at the hiring stage, they rarely get the opportunity to do so.” But as he also notes:

Defenders of an employment status quo typically move from finding that there is little discrimination to a claim that any differences are the result of underlying group attributes. Many scholars who are repelled by such claims in the case of racial minorities are remarkably comfortable making them when it comes to conservatives.

Still, as Teles explains, a combination of perceived discrimination (leading to self-selection out of the hiring pool), a preference against working in overwhelmingly liberal environments (where conservatives feel social pressure to falsify their preferences), and the feedback loop between job opportunities over time and “what is "hot" in the discipline” could be sufficient to explain the severe underrepresentation even if there were little to no explicit discrimination. Jon Shields suggests how such a process could be self-reinforcing:

[…] even if all professors were impartially handing down their discipline's body of knowledge — indeed, even if they went out of their way to welcome conservative students — we would still expect conservative undergraduates to be less drawn to majors that have been shaped by progressive interests and interpretations. And that means that the leftward tilt of the social sciences and humanities is self-reinforcing. As the left has often appreciated in other contexts, the most intractable inequalities are often sustained because of structural causes rather than personal animus.

In any case, trying to explain and understand the source of the underrepresentation is not necessarily the right place to start. As with racial and gender representation in various industries, it’s difficult to precisely tease out the degree to which the status quo is a result, of explicit discrimination, perceived discrimination, socialization, subcultural norms and values, or group-level differences in personality traits or abilities. But shouldn’t we first start with whether any given instance of underrepresentation is really a problem or not? Once we conclude that it is a problem, we need to decide how big of a problem it is and how hard we should work to fix it. This is the point at which it most obviously becomes useful to understand the precise causes: so that we can consider what may be done to address the situation and the potential costs of changing the status quo.

The desire to frame underrepresentation as resulting from explicit (or at least systemic) discrimination is motivated by the fact that most people agree (at least in theory) that discrimination is wrong, even when it’s done against their outgroup. If underrepresentation is a matter of explicit discrimination, it’s also seen as easier to change, and the group advocating for such change is assumed to occupy the moral high ground. But one can equally advocate for change on the basis that the status quo is suboptimal with respect to the organization achieving its goals or mission without any need for claims of unfair treatment. 

I think such an argument is easier to make for conservative underrepresentation in the academy than it is for something like the underrepresentation of women in tech. Sure, we could say that women will influence product design, or bring new ideas to the table in ways which allow the product to better serve the overall customer base, but such arguments are pretty weak for a number of reasons. First, the workforce at a tech company will almost surely be smarter on average (and in many other ways different) from their typical consumer, so we’d have to explain why women in particular must be equally represented while other differences are not a problem. Second, the work done at such a company doesn’t often rely in any direct way on gendered experiences, so it’s often unclear why gender would be a particularly good way to measure diversity of thought where it matters. And third, women are already 25-30% of the workforce in tech, which is high enough that any additional diversity benefits would likely be very small.

But even when it comes to academia, it’s not trivial to argue that conservative underrepresentation (absent discrimination) is a problem. First, as with the women in tech example, we should expect that academics will not be representative of the general public in many ways (which may include their political ideology), and as Teles concedes there are reasons to expect that liberals will be overrepresented in academia since:

[…] conservatives are — almost by definition — lacking in the psychological characteristics that predict high achievement in academic subjects, such as openness to experience and a willingness to question received understandings. Conservatives are, after all, biased toward conserving knowledge, whereas academia tends to reserve its laurels for creative, conceptually path-breaking producers of new knowledge.

But still, even if conservatives will always be underrepresented, we’d still have to explain why they are even more underrepresented now than they were in the past. Second, as with the example of women in tech, there are disciplines where an academic’s personal political ideology shouldn’t directly bear on her work or the questions she chooses to investigate. That said, it’s hard to imagine that political ideology doesn’t affect the research produced within the social sciences. And on the third point, the extremely low levels of self-identified conservatives within the social sciences relative to the levels at which women in tech are underrepresented makes the argument for the benefits of increasing diversity at the margin much easier to buy.

Conservatives and liberals likely have different interests such that we could expect greater diversity to lead to a wider variety of research questions being asked or methods uncovered in certain disciplines. But the role that academics play, not as collaborators, but as one another’s evaluators seems to me to make a stronger case for diversity within academia. The scientific method works best when scholars are motivated to sharply criticize and even attempt to disprove one another’s work. Doing so at the academic level requires specific training and knowledge in addition to high levels of competence, such that outsiders may not be able to adequately serve this role. And without the appropriate credentials, their work may not be taken seriously even if they are capable. If the Overton window within a discipline is exceedingly narrow, it’s not hard to imagine how certain assumptions might go unquestioned, how empirical results that confirm existing biases might go unchallenged and how this could lead to certain questions appearing “settled” when they in fact have not been appropriately scrutinized. 

I’d expect that the left-leaning environment on college campuses and the broader social world in which academics exist is simply unattractive to most conservatives, and that this, in addition to personality differences and the self-reinforcing evolution of disciplines in which liberals are overrepresented, probably explains much more of the underrepresentation than explicit discrimination does. But regardless of whether anyone is being unfair to potential conservative hires, the extremely low level of their representation within the social sciences, the reduction in that representation over time, and the potential for certain assumptions and results to go unchallenged in such an environment seems to me to justify concern.