The American Enterprise Institute is a think tank that generally favours free markets. Its president, Arthur Brooks, recently took to the pages of the New York Times to warn that the marketplace for intellectual ideas is badly failing and needs to be corrected from the outside.

Conformity and ideological groupthink, Brooks warns, are stifling intellectual inquiry throughout much of American academia: A lack of ideological diversity “is taking a toll on the quality and accuracy of scholarly work. This year, a team of scholars from six universities studying ideological diversity in the behavioural sciences published a paper in the journal Behavioral and Brain Sciences that details a shocking level of political groupthink in academia. The authors show that for every politically conservative social psychologist in academia, there are about 14 liberal social psychologists; 79 per cent of social psychologists admitted they would be less likely to support hiring a conservative colleague than a liberal scholar with equivalent qualifications”.

Brooks’ warning is certainly an important one. Groupthink is a real phenomenon, as many years of psychology experiments have repeatedly confirmed. If allowed to run unchecked, it can certainly limit intellectual exploration. But I believe that Brooks’ case has several important holes in it.

First, Brooks’ measure of groupthink — political ideology — isn’t always the best one. In physics, for example, politics of the left-right variety doesn’t matter at all. The kind of groupthink we need to watch out for in physics is associated with academic politics — theorists investing too much in one unproven theoretical paradigm, for instance. Simply taking polls that show that physicists tend to be liberal wouldn’t tell us anything about open-mindedness in physics itself. Even in social psychology, the field Brooks discusses, it isn’t clear how many issues are relevant to national politics.

Second, Brooks makes statements about “academia” in general, as if it is a unified, homogeneous object. It isn’t. Surveys show a healthy degree of political diversity in most academic fields. In addition, professors tend to be centrist, with only a slight leftward lean overall. Some fields, like anthropology, sociology, social psychology, and the humanities, are chock full of lefties. But economics, engineering, and business fields tend to be evenly split.

Third, Brooks doesn’t consider how much groupthink in a given discipline actually matters. Economics is hugely influential, for example, and even a slight political bias in economics might result in policy changes that hurt millions of people. But anthropology research has — sorry, anthropologists! — basically zero effect on anything. So if anthro research is ideologically compromised, very little harm will come of it.

Fourth, and most crucially, Brooks discounts the possibility that groupthink in academia is self-correcting. Academia is all about dissent and argument. If flawed ideas prevail, you can expect them to be challenged. This certainly happened in economics, where groupthink in the late 1970s and early 1980s almost certainly caused an over-reliance on models of purely self-interested, infinitely rational actors and smoothly functioning markets. In the late 1980s and 1990s, economists in almost every field rose up to challenge this orthodoxy, and dramatically broadened the profession’s outlook to include things like behaviourism, asymmetric information, incomplete markets, altruism and social norms.

So while Brooks’ worry is legitimate, his evidence for really harmful, persistent groupthink is thin. Not only that, but the most commonly suggested cure — affirmative action for ideological dissenters — would create more problems than it solved.

For one thing, making a special effort to recruit conservatives for social psychology departments would create a system that could be easily gamed. Whereas race and gender are very difficult things to fake, political ideology can’t be proved or disproved. Job applicants could basically just say “Hey, I believe in politically incorrect stereotypes, give me a tenure-track job in a social psych department!”

That would get real dumb, real fast.

Also, ideological affirmative action could damage scientific consensus. Suppose that all the available evidence says that giving poor children money improves their mental health. An ideological conservative, hired for his political beliefs, might be tempted to dispute this finding without good cause. In fact, he might feel pressured to do so, because to agree with the weight of evidence would be to give up his special position as the department iconoclast. Thus, evidence and data would lose out to ideology.

Ideological affirmative action is a silly idea. The best solution to groupthink is just better science — more focus on data and evidence. Humans are smart — we have our biases and our fads and fashions, but eventually hard evidence will conquer these things. The marketplace of ideas may not be perfectly efficient, but it works in the long run.

— Washington Post

Noah Smith is an assistant professor of Finance at Stony Brook University and a freelance writer for finance and business publications.