Affirmative Action in Social Psychology?

Social psychologist Jonathan Haidt has attracted some controversy recently over his call for affirmative action in social psychology. Haidt polled his colleagues over their political affiliation during a lecture and found that only a tiny minority identifies as conservative. Of course, as he well knows, this isn’t strong evidence for the claim that social psychologists are overwhelmingly ‘liberal’ (in the American sense of that word), but the available data would suggest that this is overwhelmingly likely to be the case. If this is correct, social psychology would be unrepresentative of the general population (given that around 40% of Americans identify as conservatives). Hence Haidt’s call for affirmative action: aim to have at least 10% of the membership of the professional organization be conservative.

But why might the fact that an organization, any organization, is unrepresentative of the population be a matter for concern? The mere fact of being unrepresentative can’t be a particular problem. If the organization representing social psychologists is not unrepresentative as far as IQ goes, then something has gone very wrong with the field (similarly, if the audience is not far more statistically sophisticated than average, something has gone wrong). Academia is an exclusive club, and membership is restricted to the relatively few who can pass the stringent tests it imposes. Of course not academics are very smart (even more obviously, not all very smart people are academics), nevertheless, academia selects for intelligence.

Merely being unrepresentative is therefore not a cause for concern. What is? I think there are two possible grounds for concern, one more general (applying to any organization) and one more specific to an academic organization.

The general concern is that the unrepresentativeness might reflect discrimination. Haidt points out that social psychologists are often concerned about underrepresentation of women and minorities, presumably because they think it is likely that this underrepresentation is the result of discrimination. Is it likely that there is discrimination against conservatives in social psychology, and that this explains the imbalance in political affiliation?

It is not implausible that there are (probably quite subtle) pressures that bias toward liberalism. These pressures might either tend to cause students to gradually become more liberal in their views (as a consequence of something like attitude polarization) or make those who are more conservative seek less unfriendly environments. Even without the imposition of illegitmate pressures, one would expect people to tend toward the political views of those they associate with and respect, due to our tendency toward social conformity. It is also likely, though, that attraction to a field to like social psychology is correlated with being liberal. Part of the reason is that conservatives tend to be higherin just world beliefs, and those who score higher on this scale are less likely to be interested in group membership and its characteristics. Given, however, that social psychology does not seem very different from other fields of academia in its political leanings, we should look for another explanation.

Part of the reason is likely to lie in the fact that recently the Republicans have been an anti-intellectual party. Only 48% of Republicans accept that global warming is happening (compared with 87% of Democrats). This is not merely a phenomenon of the base: Chris Mooney has spoken, rightly, of a Republican War on Science. Given the pressures against intellectual activity on the right, the political bias of academia might be explained. Further, unlike members of the groups whose underrepresentation has triggered concern among social psychologists (Haidt mentioned efforts made by the organization to increase representation by “ethnic or racial minorities, first-generation college students, individuals with a physical disability, and/or lesbian, gay, bisexual, or transgendered students”), conservatives are not in general a discriminated against group. They are wealthier and more powerful than average, not less. Given this fact, we ought to wait for (and to seek) better data on the causes of liberal political affiliation among academics, before becoming unduly worried.

A second reason to worry about underrepresentation is more specific to an intellectual organization: it is that sameness of view causes ‘group think’; leading to lower quality output or less creativity from the group. Given Haidt’s own longstanding claim that conservatives have a richer social ontology than liberals – they understand the social world along dimensions to which liberals are blind, especially sanctity, loyalty, and authority – they might generate hypotheses which liberals would not. However, though liberals might not understand the world in these terms, it seems that they certainly understand that others do. In fact, liberals are very sensitive – it might be argued oversensitive – to signals of authority and group loyalty. Of course, liberals might be less likely propose explanations that presuppose that the sacred is real, but – even if one does not see that fact as an unalloyed benefit – it seems that for most explanations of social phenomena the hypothesis that subjects believe that the sacred is real is co-extensive with the hypothesis that they are right in seeing things in this way. It is only on some particular hypotheses, according to which God intervenes in the social world, that belief in the sacred will generate different predictions. And those hypotheses are bad ones.

At very least, it seems that there are much more important things to worry about than the underrepresentation of conservatives in social psychology. As Paul Krugman suggests, affirmative action for liberals in the military – given that the military has genuine power – might be much more important. It is worth the while of social scientists to investigate the causes and the effects of the imbalance in academia in general and in particular fields, but funding that work ought to be a much higher priority than the affirmative action proposal Haidt puts forward.

  • Facebook
  • Twitter
  • Reddit

6 Responses to Affirmative Action in Social Psychology?

  • Neil:
    Yes, these are exactly the two reasons that I called for affirmative action for conservatives. I don't care at all about underrepresentation of any group. I care about discrimination and about groupthink. I hope your readers will view the talk itself, and my supplementary materials, on this page: http://people.virginia.edu/~jdh6n/postpartisan.html

    Jon Haidt

  • Matt Sharp says:

    errr…affirmative action is discrimination, isn't it?

  • Simon Rippon says:

    Thanks for this great post, Neil, and thanks to Jonathan Haidt for stopping by with the link to materials on his website. My view is that if Haidt cares about "discrimination and groupthink," he should really offer much better evidence that they have had any significant impact in his field, including better examples. It's not "groupthink" if everyone agrees, so long as everyone came to agree by rational means rather than irrational pressures. Fundamentally, Haidt's conclusions from the statistics seem to rest on the implicit premise that it's no more rational to be, in 2011, a self-described (US) "liberal" than a self-described (US) "conservative". But this seems to me to be a subject for substantial dispute.
    Haidt's supposed examples of supression of research ideas in the talk are very weak gruel – he rests in particular on the cases of Larry Summers and Daniel Patrick Moynihan, but as Alison Gopnik rightly pointed out, neither of them were acting in their professional capacity as academic researchers when they made the claims that caused them social difficulties. Haidt thinks Summers did nothing wrong in hypothesizing that there are fewer very smart (and stupid) women than there are very smart (and stupid) men in order to explain their underrepresentation among the ranks of scientists at Harvard, but Summers was acting in his capacity as Harvard President at the time – a position that brings with it the (oft-exercised) power to unilaterally veto tenure decisions without explanation. It was Summers' role together with the fact that it was poorly-grounded speculation that made what he said so offensive and so complacent, not the substance of the idea as a legitimate subject for research.

  • Sebastian Hallward says:

    Where would we be as academics without the manic energy of liberal agonizing? where would we be without liberal "oversensitivity"? (in this case about whether "unrepresentativeness might reflect discrimination"?) My compliments to Dr. Levy for providing us with a concise example of the bias in question: of course the Democrats are "intellectuals" and the Republicans "anti-intellectuals" — as liberals we KNOW THIS. So too, Simon (alas) begs the question when he says "It’s not 'groupthink' if everyone agrees, so long as everyone came to agree by rational means rather than irrational pressures" — great, now define how we distinguish the "rational" from the "irrational"; whose beliefs fall beyond the scope of "rational" demonstration? conservative's? or liberals? obviously (as Simon knows), the invocation of "rational pressures" by themselves resolves nothing since what divides conservatives and liberals in the first place are differing substantive views of what '"rationality" really is.

  • Michelle Hutchinson says:

    In addition to the worries which the conference caused about there actually being so few conservatives there, might it have raised concern about people’s willingness to admit to being conservative? This is hinted at in the NYT article, with the student who says they avoid talking about politics because they do not identify as liberal. This might be thought to be more worrying than there actually being very few conservatives, because it would rule out the possibility of everyone having rationally reached a consensus. On the other hand, it might be considered better than the possibility of a genuinely homogeneous group of liberals in a situation of ‘groupthink’, since just raising the possibility of affirmative action may give confidence those who were tentative about acknowledging their conservatism.
    Sebastian – must liberals and conservatives differ on their view of what rationality is? Presumably neither group would want to call themselves irrational. But perhaps they might have the same conception of rationality, yet conflict in their views of how important rationality is. For example some people might want to down-play the role of rationality in morality and politics, in favour of something they see as more important, perhaps faith or tradition.

  • Peter Wicks says:

    A Sebastian implies, insisting on "rationality" seems unlikely to be an effective defence against groupthink. A better approach would be to ask themselves what assumptions seem to be shared by the whole group, and see what happens when someone questions them. Curiosity or anger?

    Of course, if students are unwilling to admit to being conservative then this would be clear evidence of groupthink (self-censorship being one of the well-known mechanisms).

Recent Comments

Authors

Affiliations