Social Psychology Looks Itself in the Mirror: Why So Few Conservatives?

What does it look like when the Cathedral examines itself? The work of the social psychologist Jonathan Haidt allows us to answer: in a 2011 talk at a convention of the Society for Personality and Social Psychology (SPSP), he pointed out the political leanings of the discipline:

We have a statistically impossible lack of diversity in social psychology. This graph shows Gallup data since 1992. Self-identified conservatives have long made up about 40% of the American public. Self identified liberals have made up about 20%. So the ratio in America is about two to one, conservative to liberal. What’s the ratio in social psychology?

To begin calculating our ratio, I first turned to Google. … Google failed to uncover a single instance of a conservative social psychologist who is currently active.

I next conducted a small survey by emailing 30 social psychologists I know, spanning all levels from very senior professors down to grad student. I simply asked:… “Can you reply to this message with the names of any social psychologists that you believe are politically conservative?” There were 4 names mentioned one time each, but each of them was hedged with doubt, such as “I don’t really know, but she did work with Phil Tetlock.” So I won’t print these 4 names. Peter Suedfeld got 2 votes, and he definitely worked with Tetlock. Rick McCauley got 3 votes. The next most common candidate was “I can’t think of any conservatives.” And finally, it turns out there is a fair amount of agreement as to who the conservative is in social psychology, and its Phil Tetlock. So there you have it, we do have a conservative. That conservative blogger was wrong. Right?

Well, not quite. I wrote to Phil to ask him whether it was true, as widely believed, that he is a conservative. Phil wrote back to me, in characteristically Tetlockian fashion, and said: “I hold a rather complex (value-pluralistic) bundle of preferences and labeling me liberal or conservative or libertarian or even moderate is just not very informative.”

But I pressed on in my search for the wild conservative social psychologist, and I found him, hiding in a bamboo grove outside of Philadelphia. Watch closely: there he is. Rick McCauley, at Bryn Mawr College. Rick is the only social psychologist I know of who publicly acknowledges that he is politically conservative. …

But McCauley can’t be the only conservative in social psychology. If we did a poll of the whole field, we’d surely find at least, what, five percent? Well, this room is just about the best sample of social psychologists we’re ever going to find, so let’s see. If there’s around a thousand people here, we should have about 50 conservatives. That would be 5%. So please tell me, by show of hands: How would you describe your political orientation? If you had to choose from one of these 4 labels, which would you pick? How many of you would describe yourself as liberal, or left of center. [At this point, a sea of hands went up. I estimated that it was between 80 and 90% of the audience, and I estimated the audience size to be about 1000 people.] How many of you would describe yourself as centrist or moderate? [approximately 20 hands went up]. How many of you would describe yourselves as libertarians? [Twelve hands went up] And when I asked how many would describe themselves as conservative, or right of center? [Exactly three hands went up.]

As you can see, we have nowhere near 50 conservatives in this room, we are nowhere near 5%. The actual number seems to be about 0.3%. In this room, the ratio of liberals to conservatives appears to be about 800 to 3, or 266 to 1. So the speaker in the earlier talk was correct when he said, from this stage: “I’m a good liberal democrat, just like every other social psychologist I know.” …

I don’t think we should ever strive for exact proportional representation. But a ratio of two or three hundred to one, in a nation where the underlying ratio is one to two? When we find any job in the nation in which women or minorities are underrepresented by a factor of three or four, we make the strong presumption that this constitutes evidence of discrimination. And if we can’t find evidence of overt discrimination, we presume that there must be a hostile climate that discourages underrepresented groups from entering.

Haidt is correct, as Yoel Inbar points out in a follow-up survey: “such large statistical disparities are legally considered prima facie evidence of discrimination.”

Inbar’s survey provides the statistical evidence to back up Haidt’s claim: he surveyed the members of the SPSP mailing list, first asking whether they identified as liberal, moderate, or conservative with regard to social, economic, and foreign policy issues:

We found an overwhelmingly liberal majority only on social issues: Here only a handful of respondents described themselves as moderate (5.5%) or conservative (3.9%). But in the two other domains, we found a considerable amount of diversity. On economic issues, 18.9% were moderates and 17.9% were conservative (i.e., right of Moderate). Similarly, on foreign policy, 21.1% were moderates and 10.3% were conservative.

Six months later, Inbar sent out another survey, this time asking about not only political affiliation, but also whether the respondent had experienced a hostile political climate and whether they thought there was a hostile political climate for conservatives. He found that conservatives experience a more hostile climate than moderates, who themselves experience a more hostile climate than liberals—and that perception of a hostile climate for conservatives correlates almost perfectly with political orientation. There is a hostile climate for conservatives within social psychology, but it’s invisible to liberals.

Why did Haidt find so few conservatives compared to even Inbar? Six percent—the percentage who describe themselves as conservative “overall”—isn’t much, but it’s much more than three out of eight hundred.

As it turns out, conservatives have good reason to hide their politics:

One in six respondents said that she or he would be somewhat (or more) inclined to discriminate against conservatives in inviting them for symposia or reviewing their work. One in four would discriminate in reviewing their grant applications. More than one in three would discriminate against them when making hiring decisions. Thus, willingness to discriminate is not limited to small decisions. In fact, it is strongest when it comes to the most important decisions, such as grant applications and hiring.

Liberals often say that reality has a liberal bias—but they don’t notice the liberal bias in the institutions that try to find reality.