conservatives and keynesians and chemists – oh my!

We’ve often discussed the sources of professors’ political beliefs. I’ve argued multiple times that the evidence points to a story of self-selection. Political scientist Steve Teles wrote me the following note, which he has given me permission to cite:

What about selection due to anticipation of discrimination? That is, I would like to be a social scientist, but I believe that the field has pervasive discrimination, either directly (“we don’t like conservatives”) or indirectly (“we think those topics are not the kinds of things that elite members of our discipline look at”). If that belief came to be common among conservatives, it would be very powerful in leading to the ideological reproduction of a discipline–even if the belief was factually incorrect. I would note that conservatives only deepen that belief by saying there is a lot of discrimination (just as their statements that litigation is out of control lead to overcompliance with regulation by firms–see epp’s excellent making rights real, which you must

Good point.

There are at least three plausible theories of political beliefs in the academy. (a) Discrimination – liberal professors refuse to admit/hire/promote/cooperate with conservatives. (b) Self-selection – people who are conservative do not try to go into academia for financial or personal reasons (e.g., lower pay). Steve raises: (c) Conservatives anticipate discrimination so the self-select.

When I judge these theories, I think they should be able to account for the following facts:

  1. Overall, the median professor is liberal.
  2. The median professor is liberal even in areas that are a-political, like math or chemistry.
  3. Conservatives have done very well in some disciplines, like business, economics, and the law.

(a)-(c) all explain #1. However, they do not equally explain#2 or #3. For example, the standard discrimination hypotheses does not explain #3. The economics profession, for a long time – from the 1930s to the 1970s or so – was strongly dominated by Keynesians. Even today, the median economist is a liberal Democrat. Yet, economics has notable free market proponents who teach at leading schools. Same for law – many of the leading conservative jurists have had teaching positions at top law schools. These empirical observations are strong evidence against the basic discrimination hypothesis. Anticipated discrimination also explains #1, but fails at #2. For example, are we to believe that young math students avoid research mathematics because liberal professors can spot their politics?

To be fair, we should approach this as a decomposition of variance issue. Self-selection, I believe, probably accounts for the lion’s share of the variance. Anticipatory discrimination probably has some explanatory power in some social sciences and humanities. Between self-selection and anticipation, there probably isn’t a whole lot of room left for outright anti-conservative discrimination.


Written by fabiorojas

July 18, 2011 at 12:58 am

14 Responses

Subscribe to comments with RSS.

  1. Perhaps worth keeping in mind that point #1 is actually a bit more interesting than you’ve made it out. In basically every discipline the median prof is a liberal but in some disciplines (i.e., most of the social sciences and humanities) this is also true of the 90th percentile. This in turn provides some analytical leverage. Basically, you’ll get a bad explanation if you think that exactly the same things are going on in STEM as in sociology.



    July 18, 2011 at 2:32 am

  2. I’m willing to admit that variance explained by discrimination might be higher in soc, but why did conservatives do well in econ and law, which were once (and still are to some extent) liberal fields?



    July 18, 2011 at 4:04 am

  3. Looking at variation by discipline, it seems to me that conservatives pick fields in which there is more opportunity for jobs outside academia even if they remain in academia. This explains their odd (relative) preponderance in nursing, for example.



    July 18, 2011 at 1:59 pm

  4. There is a huge difference between discriminating against an individual, because they are conservative, and liking/disliking particular types of research which happen to be correlated with values or political ideology.

    It would be interesting to see how political ideology lines up with style and content of research within disicplines.


    Michael Bishop

    July 18, 2011 at 4:05 pm

  5. It seems that one would have to add only the slightest bit of nuance to the “anticipatory discrimination” hypothesis to give it more weight. Namely, conservatives anticipate discrimination differently across fields. From this list, a plausible mechanism is that they run the following heuristic “If business related, ok to proceed.” This might work two ways. First, liberal colleague who have to engage with the practical aspects or professional workforce of these fields, and thus are potentially exposed to those conservative elements in them, might be more sympathetic to conservative colleagues’ viewpoints as a result of this exposure. Second, professional acquaintances of the conservative academic might not subject that person to any serious cocktail party ribbing of her career choice because of their prior interactions with, and possible respect for, the academic community which contributes to their field.

    This raises a generalizable question. From where are these missing conservative academics anticipating discrimination? I gather it’s most often said it will come from liberal academic colleague. But, I’d be worried about my conservative non-academic acquaintances. If these people have no access to what my professional life is like, and are likely to be skeptical about my value to communities they care about (i.e., profit-making ones) I’m likely to think very hard about entering academia.


    Mark H.

    July 18, 2011 at 9:14 pm

  6. Re: conservatives doing well in econ and law. My guess is that this has something to do with conservative foundations plowing their money into these fields because of their policy influence (as Teles demonstrates for law). Both were more left-liberal in the 1950s but slowly moved to the right in the decades to follow. I think there might be a certain tipping point in social science fields where there is enough of a conservative influence to signal to others that the water is warm and its safe to jump in (again, this only applies to the perception of discrimination whether real or not).



    July 18, 2011 at 11:12 pm

  7. @Joschmccabe: Financial support for conservatives in law & econ doesn’t get you very far, either. Why should liberal professors be more to nice to conservatives/conservative topics just because it’s funded? No amount of funding would make me vote for a creationist biology professor.



    July 19, 2011 at 2:01 am

  8. Fabio,

    True, and if anything ideological sources of funding should provide an easy (and arguably justified) grounds for criticizing the people who take it, though I think this stigma probably ameliorates the benefit of the resources rather than making them actively counterproductive.

    Still, I think you’re right in hinting that we need to engage with the content of the ideas. Certainly, public choice theory, originalist jurisprudence, and law and economics analysis are a bit more congruent with the dominant paradigms in economics and law (as they existed c. 1970) than creationism is with the neo-Darwinian synthesis in EEB. Similarly, in sociology we have some cases where ideas that were originally seen as conservative ideology but were nonetheless fundamentally congruent with the sociological perspective (e.g., the benefits of growing up in an intact married household and the possibility of second-order effects from policy interventions) have been accepted into the discipline’s scholarly consensus, albeit perhaps more slowly than they might have otherwise been.



    July 19, 2011 at 2:32 am

  9. @ Gabriel

    But the hesitation sociologists had regarding ideas like the benefits of growing up in an intact married household had less to do with the actual idea than the interpretations and discussions that followed from the conservative side.

    Conservatives often made the logical jump that because B is related to C, therefore D. Intact marriages are related to positive outcomes for children, therefore families that don’t fit the “ideal” model cause problems for children. Though the causal argument is not quite as neat as some of the interpretations suggest. The problem, as we sociologists know, is that there are a whole set of contextual/societal factors located in A (the set of factors that come prior to B or are working simultaneously with B), which help to explain the relationship between intact marriages and outcomes for children. Some of these factors have been specified, but others have not yet been identified in our models. So the arguments, conflicts and debates between camps had a lot to do with underlying assumptions.

    With all this talk about how to predict whether certain fields are “liberal” or “conservative” or why some fields are more “liberal” or “conservative” than others, I wonder if we should take a step back, and define our terms. (My apologies if this was done in earlier discussions on this blog, as I am relatively new).

    But, what do we mean when we say “liberal” or “conservative.” It seems we are taking the definitions of these terms for granted in our discussions. If we have clearer definitions of the terms, then we might be able to identify the underlying assumptions of individuals in each camp, and better understand why they might be attracted to some fields more than others. Maybe because there are sets of theories in each discipline that jibe better with the things we take for granted as researchers.


    Scott Dolan

    July 19, 2011 at 1:11 pm

  10. With all of that said, I meant to say: self-selection does seem plausible.


    Scott Dolan

    July 19, 2011 at 1:42 pm

  11. Scott,

    If you’re trying to say it might be spurious just spit it out. Speaking of which, I’m taking it you’re not a fan of McLanahan and Sandefur (and the related literature).



    July 19, 2011 at 3:21 pm

  12. I can’t say completely whether it is spurious, but was merely trying to point out (as you know) correlation is not causation. Nor can I say that I am a fan or not a fan of the work. From what I know of McLanahan and Sandefur and the related literature (which is cursory at best), they deal with many of the antecedent and intervening variables at work in explaining the correlation between family structure and child outcomes.

    But I was just using the example you brought up to make my point about underlying assumptions and interpretation of findings. What we look for and how we interpret findings has a lot to do with the theoretical approaches and paradigms we use as lenses to see and discuss the data. I think these approaches and paradigms differ within and across disciplines (i.e. the standard Intro to Soc differences between structural-functionalism and conflict approaches). I think individuals are attracted to disciplines and theoretical frameworks which fit into their specific worldviews (though granted sometimes our worldviews change as we are exposed to new information).

    And the only reason I brought all of the above up was to make the argument that we should be defining the difference between a liberal and a conservative if we really want to say that some disciplines are more conservative/liberal than others. I, for one am not certain, it’s as clearly black and white as we are making it seem.

    But if we want to treat it at such, define the terms. With definitions in hand, it might be easier to identify why some individuals are drawn to some disciplines (and specifically certain theoretical frameworks within each discipline) rather than others. I would venture a guess that the attraction to econ and law, to take two examples, is not just about remuneration, but also about self-selection based on a particular view of the world and how it works.


    Scott Dolan

    July 19, 2011 at 3:49 pm

  13. And apologies for long-winded-ness ;-)


    Scott Dolan

    July 19, 2011 at 4:08 pm

  14. Um, conservatives haven’t done that well in law, except by comparison to fields where they’re almost entirely excluded. At most top law schools, the number of conservative faculty can be counted on one hand. And to point out, as you do, that “many of the leading conservative jurists have had teaching positions at top law schools” is perfectly consistent with there being widespread discrimination against conservatives — sure, a handful of the very best are allowed to make it, but acceptance of conservatives is not at all modal.



    July 26, 2011 at 2:58 am

Comments are closed.

%d bloggers like this: