"The Bright Future Of Post-Partisan Social Psychology" By Jonathan Haidt

Daniel Kahneman, Daniel Gilbert, Mark Leary, Paul Bloom, Steven Heine, Alison Gopnik, David Pizarro, Lee Jussim, Roy Baumeister, Jennifer Jacquet, John Jost,David Dunning

Jonathan Haidt responds

DANIEL KAHNEMAN
Professor of Psychology, Princeton University; Recipient, 2002 Nobel Prize in Economic Sciences

Great piece, perfect for Edge — a real service.

DANIEL GILBERT
Professor of Psychology, Harvard University; Author, Stumbling On Happiness

Jon Haidt is one of the smartest people I know, an exquisite writer and orator, and a good friend. What he says about social psychology is both interesting and true. Unfortunately, what's interesting isn't true and what's true isn't interesting.

It is deeply uninteresting to learn that most social psychologists are liberal. Jon's survey of a few friends and his afternoon with Google merely tell us what people who collect serious data have told us many times before: in every room of the academy, liberals outnumber conservatives by a whopping margin. The interesting question isn't whether, but why?

One well-chewed possibility is that liberals are more likely to want to become professors. For example, liberals may be more interested in new ideas, more willing to work for peanuts, or just more intelligent, all of which may push them to pursue the academic life while deterring their conservative peers. Jon tells us that that such factors undoubtedly contribute to the lopsided ratio of liberals to conservatives in social psychology, but that they cannot fully explain it because the ratio is just too lopsided, representing what he calls an impossible lack of diversity.

Impossible? Really? How does he know that? Exactly how lopsided must a ratio be before we are allowed to conclude that it could not possibly occur without bias? Ten to one? A thousand to one?

Jon doesn't say because Jon doesn't know. And yet, that doesn't stop him from concluding that "the underrepresentation of conservatives in social psychology, by a factor of several hundred, is evidence that we are a tribal moral community that actively discourages conservatives from entering." This is, of course, an entirely indefensible conclusion. Four decades ago, the economist Thomas Schelling showed that in a world completely free of bias, communities can become totally segregated through the voluntary actions of people who have a slight desire to be near just one person of their own kind and not an iota of antipathy toward others. Could the segregated neighborhood of social psychology be an instance of this phenomenon? Jon doesn't know that either. And yet, failing to know the cause of our segregation doesn't stop him from prescribing a remedy that would make conservatives seethe: an affirmative action program, complete with a target quota of ten percent. Why is ten the right number? Why not twenty, or two? Because…oh damn, there's one of those annoying questions about data again. Can't we all just get along?

The true and uninteresting fact is that social psychology has many more liberals than conservatives and this may or may not be the result of an anti-conservative bias. Jon sympathetically cites Larry Summers while missing Larry's point, which was that a lopsided ratio of men to women in the Physics Department, or of liberals to conservatives in the Psychology Department, or of whites to blacks in the Fire Department, does not in and of itself provide evidence of bias. Such evidence requires us to dig deeper into the data with the sophisticated tools that science provides. Perhaps Jon's beautifully crafted speech was merely meant to start a debate. But scientists start debates by raising questions, not by making up the answers.

MARK R. LEARY
Professor of Psychology and Neuroscience, Duke University; Author, The Curse of the Self

Undoubtedly, every scientist, indeed every person, sees the world through the filters of his or her personal values, life experiences, and personality. And, undoubtedly, the greater the variety of personal perspectives that a group of people bring to any enterprise, the broader, more creative, and potentially correct the ideas they will develop. Although I do not doubt that the liberal leanings of behavioral and social scientists have led to impoverished ways of looking at human behavior as well as to occasional biases, the big question for me is the degree to which liberalism has introduced bias into psychological science. 

First, only a portion of social psychologists, and even fewer psychological scientists in general, study topics that have anything whatsoever to do with liberal values. To those who are unfamiliar with social psychology, Haidt's message could be interpreted to suggest that the entire field is tainted by liberal biases, but it's difficult to imagine how moral or political values would contaminate most theorizing and research in social psychology. I am not questioning that certain topics—such as those involving race, gender, morality, and sexuality—are sometimes susceptible to the liberal biases of the researcher. But those topics constitute only a portion of the field, and even much of the work on those topics would not seem easily swayed by researchers' values. It would be an instructive exercise for a panel of liberal and conservative researchers to review all articles published in the leading journals in social psychology over the past five years to see what proportion of the articles could reasonably be assumed to be biased by the researcher's ideology.

Second, the argument assumes far more homogeneity among "liberals" than, in fact, exists, and a casual reader of Haidt's text might assume that every social psychologist is a flaming liberal across all political, moral, and social domains. But as I think of social psychologists with whom I have discussed personal values, I find that we are a very mixed group. Although most would choose the label liberal rather than conservative to describe themselves, I can think of many self-identified liberal colleagues who are nonetheless strongly conservative with respect to morality, fiscal matters, law-and-order issues, family issues, and even gender. Bifurcating social psychologists into liberals and conservatives obscures a great deal of diversity that is relevant to understanding the implications of Haidt's argument. 

Haidt is undoubtedly correct that the field has a liberal leaning, that these values sometimes influence our work in undesired ways, and that broader perspectives of all kinds would benefit psychological science. However, much of the discussion of this issue lacks the nuance that is needed to benefit from Haidt's message.

PAUL BLOOM
Professor of Psychology, Yale University; Author,How Pleasure Works

Imagine that you are a beginning graduate student accepted into a top-ranked psychology department. The first colloquium talk you go to is about deception, from a famous social psychologist. In the middle of her talk, she makes a remark about how some people are simply incapable of ever telling the truth, and then she puts up a large picture of Barack Obama. People roar with laughter, and there's a bit of applause. You are a teaching fellow in a large Introduction to Psychology course, and the professor talks a bit about popular delusions, giving the example of liberals who believe in global warming. Al Gore is mentioned in a lecture on clinical psychology, in the context of narcissistic personality disorder. Everyone you know is a conservative Republican and assumes that you are one too, making off-hand jokes to you about brain-dead liberals.

But suppose you are, in fact, a liberal yourself. How would you feel about this new life you have chosen?

Dan Gilbert is right that we don't know why there are so few conservatives in psychology departments. There are no doubt many factors at work. But nobody wants to be part of a community where their identity is the target of ridicule and malice. This is obvious enough when it comes to other sorts of bias, involving gender, race, and sexual orientation. I don't know why there are so few female graduate students in physics, but if it turns out that there are pervasive and overt sexist attitudes in physics departments, it would be perverse to insist that this isn't at least part of the explanation.

The analogy isn't perfect here. Political views can be right or wrong; they can change; one can and should argue about them. But there's a big difference between civil debate and the sorts of examples I gave above, which reflect what Jonathan Haidt correctly describes as a locker room mentality.

Even if this mentality turns out to have nothing to do with why there are so few conservatives in psychology, it's still ugly behavior, surprisingly so for a community that claims to value diversity. Jon is right that we should do better.

STEVEN J. HEINE
Professor of Psychology, University of British Columbia; Author, Cultural Psychology

Science benefits from diverse perspectives, and key advances often occur when ideas slip across disciplinary borders. But many invisible norms and practices in a field can discourage the mingling of diverse ideas. Jon highlights just how homogenous social psychology is in terms of its members' political leanings, and he's right in noting that the field bears a substantial cost by not considering more diverse political perspectives.

I'd like to highlight another way that psychology is narrow: both the people conducting the research and the people who are the targets of the research largely come from a select few cultural backgrounds. Here are some indicators of the narrowness of the field: A review of international scientific productivity found that American-based psychologists accounted for 70% of the citations in psychology, a proportion higher than any of the other sciences reviewed (and approximately twice the proportion of chemistry). The next biggest contributing nations are all English-speaking ones: the UK, Canada, and Australia, respectively.

Likewise, my colleagues, Joe Henrich, Ara Norenzayan, and I, have calculated that a randomly selected American college student is more than 4000 times more likely to end up as a participant in a psychology study than is a randomly selected person living outside of the West. These nonrepresentative samples wouldn't be such a problem if people everywhere thought in the same ways, but the available evidence shows that in many key ways they do not.

Unlike chemistry, where the object of study is independent of the researcher's political or cultural perspective, psychologists study people. They often get the inspiration for their ideas by their own introspections and by observing those around them. A narrow range of perspectives isn't a problem if one hopes to explain just those people who share those perspectives. But often psychologists purport to be studying human nature, and when the field only attracts those with a limited range of political and cultural perspectives, they may produce an incomplete and misleading caricature of that nature.

As a result, the field selects topics that are a common concern for this narrow slice of humanity, but neglects many others that are a concern elsewhere; researchers project an understanding of how the mind works on to people whose minds might not work in those same ways; and interventions based on psychological principles that are identified within one set of borders are applied to people living under different circumstances, often with disastrous results (see Ethan Watters's recent book, Crazy Like Us, on the human costs of exporting American psychiatric practices).

As in all topics of inquiry, great advances are often made when diverse ideas come together. Psychology would benefit by becoming as politically and culturally diverse as other branches of science.

ALISON GOPNIK
Psychologist, UC, Berkeley; Author, The Philosophical Baby

Jon Haidt makes two rather different claims about the liberal over-representation in psychology — first, that there is an implicit censorship that keeps us from considering ideas that violate some general "sacred" principles, and second, that a greater diversity of political views would lead to a richer psychological understanding.

However, the examples he gives of this implicit censorship undermine rather than support his case. Larry Summers and Daniel Patrick Moynihan were not psychologists and were not making claims in the context of psychological or scientific discussions. Larry Summers was the president of Harvard University speaking to a group considering the underrepresentation of women at that very University, Daniel Moynihan was a policy advisor to the government. Summers claim was an implicit policy recommendation, do nothing, as was Moynihan's. In Summer's case in particular, the claim not only ignored all the complexities of gene-environment interaction that we teach in first-year developmental psychology but was part of a very long history of such claims being used precisely to justify policies of exclusion and neglect. Of course, policy decisions ought to involve values and ethical judgments — how else would you make them?

In fact, of course, studies of biological bases of gender differences, far from being censored or even the subject of disapproval, are ubiquitous, both in psychology itself and even more so in popularizations of psychology. Similarly there are many studies of the relationship between family structure and disadvantage. There is no evidence at all that such studies are somehow censored or discouraged in the scientific community. In fact, given the value scientists place on novelty and contrarianism, making an obnoxious claim on slender evidence is often actually a pretty good recipe for fame.

As in the case of other kinds of diversity, however, often the real issue is not whether views are censored but whether they are considered at all. And here I think Jon may have a good point. The kinds of topics that strike us as worthy of study are often influenced by our background knowledge, interests and general assumptions. Less reductive discussions of religion and more thoughtful discussions of, say, the values of tradition or sacredness, might emerge with a wider range of participants and views.

I've argued in my own field that the moral value of family life, a topic that has traditionally been seen as "conservative", has been relatively neglected. Even here though, it isn't at all obvious that the actual political views of scientists are the best predictor of whether they will broaden the field of study. Richard Shweder has been perhaps the most forceful advocate of extending our picture of morality to include additional values that are often conservative, even though his field, cultural anthropology, is much more self-consciously leftist than psychology. My guess is that we suffer more from institutional and intellectual narrowness than the political variety.

DAVID PIZARRO
Psychologist, Cornell University; Researcher in Moral Judgment

Jon Haidt is largely right about the little tolerance we have for conservatives in our academic departments. I have witnessed the effects of this bias while presenting some of my own work.

When I report (in talks given at other psychology departments around the US) that in a certain set of studies my co-authors and I observe an inconsistency in the way liberal participants make moral judgments, the amount of effort expended by some members of the audience toward defending the liberal response as "rational" never fails to surprise me. Especially when compared to the minimal effort they exert toward defending a similar set of results showing inconsistency on the part of conservative participants (this draws another criticism altogether — that reporting irrationality on the part of conservatives is so obvious as to be uninteresting). I am open to the possibility that all of these studies are horribly flawed. What I find disconcerting is how quick my audience is to find flaws in studies that disagree with their political views.

That said, I think a key part of Jon's argument is presented in a way that will prevent people from seeing his larger point. For Jon it seems obvious that "moral diversity" is something that should be encouraged. Really? Surely it's not the case that we ought to fight for equal representation of bigots, rapists, or pedophiles in our academic institutions. But I don't think Jon means this. I think what he means is something far less controversial — that, among other things, we should be open-minded, be respectful of others, and understand that disagreements can be had between people with good intentions.

I know that it's difficult to know where to draw the line between respecting alternative viewpoints and sacrificing our belief in moral right and wrong. But I think it's probably somewhere between "the government shouldn't be in the job of providing health care" and "rape is okay." My fear is that calling for moral diversity implies that the line cannot be drawn at all.

As eloquent as I found Jon's talk to be, he leads even better by example. Right after reading Jon's 2000 piece ("The emotional dog and its rational tail"), where he argued that moral judgment was largely driven by gut, emotional responses, I sent him an email. I was motivated to send it because, until that point, I had never disagreed with anyone as strongly as I did with him. Though I was a 3rd year graduate student whom he had never heard of, he immediately responded, set up a time to chat on the phone, and had an hourlong discussion with me about moral psychology. From that interaction I learned exactly what Jon means when he says that we should respect the viewpoints of those with whom we disagree. And I hope that this point is not lost in all the media attention his talk will undoubtedly bring.

LEE JUSSIM
Professor and Chair, Rutgers Psychology; Co-Editor, Social Identity, Intergroup Conflict, and Conflict Reduction; Stereotype Accuracy: Toward Appreciating Group Differences

I cannot sufficiently express my gratitude to and enthusiasm for Jon Haidt's speech. As he so refreshingly pointed out, liberal bias infects, distorts, and undermines the quality of our science.

Dan Gilbert usually possesses an extraordinary depth and breadth of knowledge of social psychology. His lack of awareness of the abundance of data attesting to liberal hegemony and scientific dysfunction indicates exactly the type of blind spot one would predict on the basis of Haidt's claim that social psychology has become a "sacred community" whose values "bind and blind." Here is a merely scratched surface of data:

1. In 1985, Stephen Ceci, Douglas Peters, and Jonathan Plotkin submitted research proposals to over 150 Internal Review Boards, all of which were identical except for one difference. The stated research goal was either to demonstrate that, in employment situations, "reverse discrimination" or "discrimination" was a major problem. Reverse discrimination proposals were approved less frequently than the discrimination proposals. Moreover, "political implications" were given as "reasons" to reject the "reverse discrimination is a problem" proposal about twice as often as they were given for rejecting the "discrimination" proposal.

2. Gilbert's commentary, when juxtaposed with typical social psychological discourse on inequality, is more evidence. Many social psychologists write as if inequality reflects prejudice and discrimination (eg. almost anything on Social Dominance or System Justification Theory; or any of several papers in law journals linking broad scale racial inequality to experimental implicit association tests).

When the topic turns to politically motivated scientific distortion in social psychology, the "people might assort themselves into different groups without a shred of discrimination" argument emerges. Unless one also makes this argument when attempting to understand differences in the distributions of different demographic groups into different settings, this is an extraordinary double standard, and itself reflects the biasing effect of liberal politics. The argument that inequality even partially reflects merit, culture, preference, etc. is the most offensive brand of bigotry when it involves a demographic group and therefore offends the liberal community; but it is merely good scientific analysis when used to justify liberal hegemony within social psychology. You can't have it both ways, and even the attempt to do so reflects exactly the type of problem Haidt highlighted.

3. A case study. I have a paper originally written as if it tested the "conservatism as motivated social cognition" theory implying that conservatives are much more biased than liberals (in our research, this was tested by having them evaluate written materials supporting either a liberal or conservative position). We found the opposite: That liberals were much more biased than were conservatives. We could not get this published. So we reframed the paper to test a "dual process" theory of political ideology, removed all mention of liberals being more biased than conservatives (although the data is right there for anyone to see), and the paper is now in press.

Three cheers for Jon Haidt's speech. if it leads even one researcher to be more sensitive to the extraordinary double standards and blindness that sometimes taint our field, it will have been a rousing success.

ROY BAUMEISTER
Francis Eppes Eminent Scholar and head of the social psychology graduate program at Florida State University; Author, Evil: Inside Human Violence and Cruelty

So far this discussion has been conducted on liberal grounds: Does fairness of process entail we should strive to make way for conservatives among us? Let me instead try to think like a conservative and ask, what would be some pragmatic benefits to our field of having conservatives among us?

I care much more about science and truth than about politics. I'd rather know the truth even if disagreeable. My ideal is to have no political preferences at all, because they impede one's willingness to accept the facts. (Not that I'm there yet.)

Leary is right in saying that political bias likely only distorts a limited number, a minority even, of research topics. But the composite understanding of the human mind depends on the parts being correct. If we let some parts be wrong, the whole will be wrong.

A first major benefit of hearing conservative views in academia would be a pro-business view. I have a rather feeble understanding of business, but at least I know how much I don't know. My impression is that most of my colleagues do not. Marxism did much better in academic theories than in practice. Many professors detest business and the people who do it, even though most of their students will work in business. Business has made our society rich enough to afford universities, among many other benefits. Most professors, including myself, do not understand how business works and what is great about it and what it needs. Having some respected professors to espouse pro-business views would inform and elevate how we understand a huge part of life and of culture.

The second benefit would be a stronger advocacy of traditional family values. I know even some liberals are now coming round, but on things like whether parents with a so-so marriage (not liking though not hating each other) should stay together until their children are in college, it might be good to have the conservative pro-family view represented.

A third would be to have our academic culture include a minority viewpoint that would hold a generally positive view of America and its traditions. I find that liberals are generally critical of American culture and its history. I have certainly complained much about the United States myself. But as I read more, I see that compared to pretty much all other countries and historical periods ever, the US is the better place to live. It would be valuable us to figure out what the Founding Father (Parents) and others did right.

Last, while liberals evaluate organizations based mainly on fairness of process, conservatives tend to look at performance outcomes. Social psychology used to study task performance but these days is almost entirely devoted to how people think and feel as dependent variables. A conservative presence in the field might re-kindle attention to what makes people and organizations produce and perform well.

JENNIFER JACQUET
Postdoctoral Fellow, University of British Columbia

Haidt's main point is well articulated, very interesting, and captures what I believe is a true statement: that the university environment can be one-dimensional. However, as the lone liberal in an immediate and extended family of conservatives (who came to liberalism through education), I must disagree with several points.

Being a conservative is not similar to being a woman or a minority (the
latter you cannot change). In other words, it's possible that one enters the field of social psychology as a political conservative and comes out the other side politically liberal. There are very few conservatives in social psychology possibly not because it's a hostile environment but because the field of social psychology self-selects for liberals and might even create them (e.g., if you learn/accept through the work in social psychology that people are largely a product of their environment, it makes it hard to then support political strategies that further disadvantage the poor).

The military self-selects (or creates?) conservatives. Physics seems to self-select (or create?) believers, which is very interesting indeed.

JOHN JOST
Professor of Psychology, New York University; Co-Editor, Social and Psychological Bases of Ideology and System Justification

Social psychology is not a "tribal-moral community" governed by "sacred values." It is wide open to anyone who believes that we can use the scientific method to explain social behavior, regardless of their political beliefs. Nor is our "corner" of social science "broken" when it comes "race, gender, and class," as Jonathan Haidt claimed in response to Paul Krugman. Rather, social psychologists have made cutting edge advances in understanding the subtle, implicit, nonconscious biases that perpetuate inequalities concerning race, gender, and class.

Haidt's essay sows confusion; he misrepresents what we do, how we do it, and why we do it. By focusing on scientists' personal beliefs rather than the quality of their work, Haidt perpetuates the myth that social scientific research simply exemplifies the ideological biases of the researchers. No doubt this energizes those who are eager to dismiss our findings. But polling firms are paid by clients, including political campaigns, and this fact neither determines nor invalidates the poll's findings. Similarly, the personal beliefs of social scientists may (or may not) be one of many factors that affect the decision of what to study, but those beliefs are, at the end of the day, scientifically irrelevant.

This is because we, as a research community, take seriously the institutionalization of methodological safeguards against experimenter effects and other forms of bias. Any research program that is driven more by ideological axe-grinding than valid insight is doomed to obscurity, because it will not stand up to empirical replication and its flaws will be obvious to scientific peers — all of whom have been exposed to conservative perspectives even if they do not hold them.

If we do concern ourselves with the results of Haidt's armchair demography, we should ask honestly whether social scientists are too liberal or society is too conservative. After all, when experts and laypersons disagree, we do not usually rush to the conclusion that the experts are biased. Haidt fails to grapple meaningfully with the question of why nearly all of the best minds in science find liberal ideas to be closer to the mark with respect to evolution, human nature, mental health, close relationships, intergroup relations, ethics, social justice, conflict resolution, environmental sustainability, and so on. He does not even consider the possibility that research in social psychology (including research on implicit bias) bothers conservatives for the right reasons, namely that some of our conclusions are empirically demonstrable and yet at odds with certain conservative assumptions (e.g., that racial prejudice is a thing of the past). Surely in some cases raising cognitive dissonance is part of our professional mission.

We need science, now more than ever, to help us overcome ideological disputes rather than getting bogged down in them. We do not need conservatives to become conservative social psychologists any more than we need liberals to become liberal social psychologists. Our "community" still holds that policy preferences should follow from the data, not the other way around. Sadly, Haidt puts the ideological cart before the scientific horse. I simply cannot agree that — especially in this political era — it would be good for our science to reproduce the ideological stalemate and finger-pointing that has crippled our government and debased our journalism. 

DAVID DUNNING
Professor of Psychology, Cornell University; Author, Self-Insight

Jon Haidt presents his case about political imbalance in psychology with vigor, clarity, and admirable passion.  That said, I wish respectfully to dispute two aspects of his account.

First, I wish to dispute the idea that Jon has met the enemies of political balance and that they is us—specifically, that psychologists create a hostile environment for those who are politically different. That may be the case, but to my knowledge no systematic data exist to support such a claim even though the "discrimination hypothesis" has been tested.  And a few personal anecdotes illustrating hostility, which I do not dispute, are no replacement for systematic data.

The second dispute has to do with the demarcation of the phenomenon Jon describes.  Before entering into an extensive discussion about a problem, one first has to make sure that one has properly defined what the problem is.  Here, I believe Jon has erred.  Political imbalance is not constrained to social/personality psychology.  Liberals outnumber conservatives in academics across the board, as well as many fields outside the academy, such as authors, journalists, artists—even bartenders.  The real issue is not why the small neighborhood of social/personality is so politically blue, but why this much larger circle of professions is.

But that, too, is a mistaken demarcation.  Just as important, there are professions where the political imbalance is equally great but runs in the opposite direction.  Conservatives greatly outnumber liberals among law enforcement, physicians, dentists, and religious workers.  Among the military, the margin is five to one.  And that's not all.  In 1976, roughly 25% of Americans lived in "landslide" counties in which one Presidential candidate beat the other by over twenty percent.  In 2004, that percentage tucked into landslide counties had grown to nearly 50%.

Thus, the real problem is that people—both liberal and conservative—increasingly find themselves in "epistemic cultures" that can contain little exposure to diversity of opinion.  The enemy of imbalance is not us, it is ALL of us—inside and outside of psychology and academics alike.  Here, I do not think that Jon's proffered remedy of affirmative action would work.  Perhaps we could set up an exchange draft where people could volunteer to switch their professions to establish political balance—law enforcement to academics, artists to the military.  But something tells me that would not work. 

Instead, the best remedy may be to make sure to expose ourselves to viewpoints that differ from our own. Robert Heinlein once boldly claimed that he never learned from anyone who agreed with him, and I concur that it is difficult. This leads me to join Jon on perhaps his most important point: Seek out disagreement and consider it.  Discussion with people on the other side of the political divide may create some nervy moments, but it might make us a little wiser at the end of the day, and thus less disagreeable to those around us.  And, maybe in addition, it may just create a less disagreeable society.

JONATHAN HAIDT
Psychologist, University of Virginia; Author: The Happiness Hypothesis

Response to Reality Club Comments on "The Bright Future of Post-Partisan Social Psychology"

Thinking is for doing, as William James said, and there is mounting evidence that human reasoning was designed in large part to do social stuff. When we need to justify our actions, our views, or our teams, we are brilliant at finding evidence and weaving it into arguments. But when it comes to seeking out evidence on the other side, few of us can get past the confirmation bias. Science is a supremely successful institution because it institutionalizes the cure for confirmation bias: other scientists. We have difficulty finding flaws in our own theories, but we can rest assured that our colleagues will help us out.

I am therefore grateful to my colleagues in this Edge discussion for their critical as well as their supportive comments. I'll make 5 points in response to those comments.  

1) Adding Nuance

Several flaws in my initial argument can be fixed easily by adding qualifiers. Most importantly, Leary is quite right that few of us "study topics that have anything whatsoever to do with liberal values." Let me state clearly that social psychology is by and large a healthy, vibrant field of scientists, not activists. I would also like to acknowledge Gopnik's point that discussions of biologically based sex differences are not uncommon nowadays among psychologists. They can still be hazardous, but the topic is not a true taboo. Also, Dunning is correct that the ideological imbalance is not unique to social and personality psychology. It is indeed part of a larger problem of segregation into "epistemic cultures" that affects groups and institutions on the right as well as the left. I do not claim that the left is any more tribal than the right. In fact, the data I have collected with my colleagues at YourMorals.org shows just the opposite: Conservatives endorse tribal virtues such as group loyalty more than do liberals. (Which congressman has an easier job, the Democratic Whip or the Republican?)

2) When is Diversity Better Than Cohesiveness?

The famous Tajfel studies of "minimal groups" show that if you point out the most trivial possible difference among people, they'll divide themselves along that line and then prefer their pseudo-ingroup. For groups that must suppress self-interest and work together to achieve a common goal, it is usually better to stress their common ingroup identity, rather than their internal diversity. It does not surprise me that conservatives are overrepresented in work that requires cohesion in large groups that are hierarchically structured, such as the military, the police, and the corporate world. Ideological diversity might in fact hinder their effectiveness, except at the top levels, where there had better be some leadership or advisory group that values truth over cohesion. It would be a disaster for the United States if the Defense Intelligence Agency was uniformly conservative and uniformly committed to finding evidence consistent with conservative foreign policy preferences. 

Now look at scientific disciplines, and you can see why comparisons to the police or the military (made by Dunning and Jacquet) are not quite appropriate. We scientists are not trying to unite to get a job done. Cohesiveness would reduce our effectiveness, at least the kind of cohesion that emerges when we all hold the same ideological commitments, vote for the same political party, and share the same confirmation biases. We psychologists urge diversity on other professions, telling them that it will be good for them. But as Heine notes, that argument may apply to us more than to any other profession.

3) The Causes of Ideological Imbalance

Gilbert is right to object to my phrase "a statistically impossible lack of diversity." He correctly notes that my concern is not underrepresentation (for which there are many "innocent" explanations), but rather that the ratio is so lopsided that it begins to look suspicious. My "show of hands" demonstration found a lib-con ratio of 266 to 1, and my informal survey of 30 social psychologists revealed that very few of us can name even a single conservative social psychologist with a faculty appointment. I should not have said this was statistically "impossible." Jost and his colleagues have shown that there are indeed some real and relevant differences in openness to experience and other personality traits that should make liberals seek out and excel in the academy more than conservatives. But how many standard deviations apart would the liberal and conservative curves have to be to get just one or two conservatives falling above the threshold for becoming a successful social psychologist? I don't know, but it would imply an intergroup difference vastly larger than anything we ever see when comparing across races or genders. I share Jussim's concern about the double standard apparent in the comments by Gilbert and Jost. Nobody would dare to apply such lines of reasoning (including the citation of Schelling) to exonerate a profession or large corporation that was entirely composed of white males.

4) Is there a hostile climate for non-liberals?

Gilbert and Dunning doubt that there is bias or a hostile climate.  Jost asserts that there is none. Gilbert and Jost fault me for making claims without evidence. It's true I had few citations in my talk, but there is indeed experimental evidence of bias. Jussim refers to one such study; a more recent one is Munro, Lasane, & Leary (2010), whose title serves as its abstact: "Political partisan prejudice: Selective distortion and weighting of evaluative categories in college admissions applications."
And what kind of evidence would it take to convince a skeptic that conservatives often face a hostile climate? Dunning questions the value of "anecdotes," but as an intuitionist, I disagree. I think it's vital for people to get an intuitive sense of what a hostile political climate looks and feels like before they can decide whether or not one exists. This, I think, is the power of Bloom's thought experiment. One moment of empathy can go a long way toward taming the confirmation bias and improving our receptivity to evidence on the other side. I've received dozens of emails from non-liberal students since my talk was first posted on Edge, and many of them confirm that Bloom grasped their situation. Here's one example:

I can't begin to tell you how difficult it was for me in graduate school, because I am not a liberal Democrat. As one example, following Bush's defeat of Kerry, one of my professors would email me every time a soldier's death in Iraq made the headlines; he would call me out, publicly blaming me for not supporting Kerry in the election. I was a reasonably successful graduate student, but the political ecology became too uncomfortable for me. Instead of seeking the professorship that I once worked toward, I am now leaving academia for a job in industry.

Many of the emails speak of getting "ground down," "worn down," or "fighting an uphill battle." One thing I've learned from these emails is that the problem begins before grad school. Even as undergraduates, many students who are not secular liberals get the message that they won't be welcomed in the academy. Self-selection is not entirely an "innocent" process.
            
A second thing I've learned from these emails is that while there are hardly any social conservatives in our pipeline, there seem to be many grad students, like the self-described "fiscal conservative" I quoted above, who could help us if we could just retain them. I've been hearing mostly from centrists, libertarians, and moderate conservatives. Two were liberal Christians. This means that I might have erred in focusing on the absence of conservatives. The problem may be better described as the near invisibility of people who are not secular liberals. If we can make these students feel more welcome, we can gain some ideological diversity with very little effort.

5) So What?

Gilbert says that "The interesting question isn't whether [there is an imbalance], but why?" I disagree. In my talk I raised the "why" question but quickly put it aside and focused on the question of "so what?" I argued that we should seek more diversity not for moral reasons, but, as Baumeister put it, for its "pragmatic benefits." 
            
Jost's comparison to "polling firms" makes my point nicely. Jost is correct that if a highly professional polling firm is paid by a client on the far left or far right, the source of the funds neither "determines nor invalidates the poll's findings." But now imagine that all polling in the United States was commissioned by conservative causes and candidates. We'd learn a great deal about the electorate's dislike of welfare programs and burdensome regulations, but we'd be blind to its desires for protecting children and preventing corporations from fouling our drinking water. That, I argue, is precisely our situation. We have great work on the psychology of conservatism (much of it done by Jost), but we do not have a similar understanding of the psychology of liberalism, which is taken for granted as the normal, healthy perception of reality, endorsed by "nearly all of the best minds in science." We have great work on stereotyping, but few of our textbooks and handbooks report politically uncomfortable findings (such as Jussim's) about the accuracy of many stereotypes. I'm not accusing anyone of misconduct. I'm accusing my field (and a few other social sciences) of suffering from a kind of market failure because nearly all the polls are commissioned by one half of the political spectrum. 

In closing, I'd like to note the importance of personal relationships for good thinking. We are by nature tribal creatures who readily jump to "battle stations" and use our prodigious reasoning abilities to defend our team, not to find the truth. But when you cultivate personal relationships and come to like people, you're not as quick to demonize or dismiss them when they contradict your cherished values. Pizarro succinctly stated the broadest message of my talk: "we should be open-minded, be respectful of others, and understand that disagreements can be had between people with good intentions." If we can succeed in attracting greater ideological diversity in social psychology, we'll have the chance for many more cross-ideological friendships, which will be followed, I believe, by more productive disagreements. That is my (updated) vision for our bright post-partisan future.