Arnold Kling  

Not Groupthink

Julian Simon Moment of the Day... 5 Miles of Economic Illiteracy...

Daniel B. Klein and Charlotta Stern discuss how academic disciplines become dominated by particular ideologies.

At the very top departments, more than 90 percent come from the worldwide top-35 departments; the top is almost entirely self-regenerating.

They call the result "groupthink," despite some misgivings about the term. I think they should listen to their misgivings and change the term. I agree with them about what causes the ideological concentration. I wrote,

There are more Ph. D's produced each year at top schools than there are job openings at top schools. On average, then, students get placed in lower-ranked economics departments than the ones in which they are trained...

Suppose that you think of yourself as the "child" of the professor who signed your dissertation. Thus, I am Solow's "child." But I have not given Solow any grandchildren.

In order to have lots of grandchildren, you need to have lots of highly productive children. In order to have lots of highly productive children, it is necessary--and far from sufficient--to teach where there are lots of top-flight graduate students. Basically, if you are teaching anywhere other than Harvard, MIT, or Chicago, your chances of having lots of grandchildren are pretty low.

In some sense, most professors in any discipline are "descended from" just a handful of senior eminences. Unless those eminences happen to hold diverse ideological viewpoints, ideology will become highly concentrated.

The problem is not one of social psychology. It is institutional/ecological. Academia is set up so that within each field a small group out-breeds everyone else. Economics happens to have a relatively diverse ecosystem, but it's all thanks to the University of Chicago's eminence. If Milton Friedman had been a conventional left-winger, what would have kept economics from turning out like sociology?

Comments and Sharing

COMMENTS (11 to date)
eric writes:

I remember reading Stigler and he said something about how a small number of people must necessarily dominate a 'field' of economics, because you need someone to be the gatekeeper, and you basically get fragmentation, but always a gatekeeper. That gives you a handful of top guys. How could that not be in the top 35?

dearieme writes:

What about subjects where one experiment can prove you wrong? Do they show the same pattern? Or does it arise only in those subjects that are religion-like?

shayne writes:


I understand (and applaud) Harvard, MIT and Chicago striving to have the "best", both in terms of faculty and students. But isn't there at least a potential "gate-keeping" problem (as Eric noted) with that? Those schools get to define what "best" is and their definition might not be the only one. Very bright minds from outside the Harvard, MIT and Chicago system are, by that system's definition only, "not the best", and may not get the recognition deserved or even get exposure for their ideas/contributions due to that artifact. I agree that "groupthink" may not be an adequate term for this phenomena, but are there inherent and robust mechanisms in the "best" schools to minimize their being, or becoming, exclusionary?

Daniel Klein writes:

You don't say why you don't like calling domination by a bad set of ideological sensibilities "groupthink," which, I think, we agree is the case in all humanities and social science fields except perhaps Econ.

What would you call it?

What is it you don't like about calling it "groupthink"?

William Newman writes:

Dan Klein: I've referred to it loosely as groupthink myself, but I don't particularly like it. What I chiefly don't like about it is that it seems unnecessarily vague. If I were to spend much time writing about stuff like that, I'd be looking for phrases like "litmus tests, cronyism, and groupthink," and making other distinctions like "doctrinaire," not just referring to a single undifferentiated "groupthink" lump. In my experience, litmus tests, cronyism, and apolitical groupthink can vary somewhat independently.

I've seen behavior that I think of as groupthink that I believe to be independent of politics, simply due to apolitical blindered habits of thought. For example, in computer software, often a group will talk itself into thinking a project is much easier than it is, but occasionally it will instead talk itself into thinking a project is considerably harder than it is. The first case typically involves at least some can-do political loyalty signalling, but the second case doesn't necessarily --- especially when the consensus group crosses multiple organizations in such a way that there'd be no obvious motive to signal loyalty. If I'm not deluding myself about the apolitical nature of this behavior, I'd prefer to reserve "groupthink" (and some other neutral phrases like "intellectual fashion," "fad," and "inbreeding") for apolitical behaviors like this, and use other words when the issue is complicated by things like ideology and mutual backscratching.

Arnold Kling writes:

I prefer to reserve the term "groupthink" for a phenomenon of social psychology, akin to Asch conformity.

I would borrow a term from evolution. Something like inbreeding, or genetic bottleneck. I don't know the terms in evolution that well.

From Wikipedia: "Inbreeding is breeding between close relatives, whether plant or animal. If practiced repeatedly, it often leads to a reduction in genetic diversity."

The Real Bill writes:

Let's reshape that Wiki quote.

Academic Inbreeding is breeding between close colleagues, whether professor or student. If practiced repeatedly, it often leads to a reduction in intellectual and ideological diversity.

How's that?

William Newman writes:

Arnold Kling wrote "If Milton Friedman had been a conventional left-winger, what would have kept economics from turning out like sociology?" How often have you seen a sociologist pressed for an expert judgment on what the future would hold or on the benefits, costs, and net effect of some policy proposal?

Economics has enough credibility that people actively seek and listen to its opinions for real-world questions. Then when someone can get some things more correct than other people, this stands out. If Keynesianism had been a better match to 1970s inflation/employment, and Hong Kong had tanked while Sweden and East Germany boomed, and the Great Society public housing projects were a continuing success while the all-volunteer army had been a chronic public policy nightmare for decades before being phased out, Friedman might be almost entirely without influence, at least unless he had hedged his bets by making quotable quips like "the modern liberal is engaged in one of mankind's oldest pursuits; finding moral justification for pursuit of power" which have so much literary merit that people who honor them today would have nothing to apologize for.

Even when you don't ask economists for their policy opinions or their predictions, you may find them pressing them on you: Milton Friedman with vouchers and spectrum auctions, the public choice theory people with the pattern of agricultural protectionism in agricultural and nonagricultural countries, Julian Simon with his commodities bets and his auctioning-off scheme for overbooking.

I don't know of many things like this for sociologists, and that's probably partly because of my ignorance, but I doubt it's entirely because of my ignorance. If the sociologists for whatever reason aren't engaged in policy and economic reality in such reputation-altering ways, then it could be easier for them to base their reputations purely on their mutual admiration for each other.

Troy Camplin writes:

It's not really "groupthink" -- something which occurs when you put a group together in a room to come up with ideas. In that case, you end up with a few very uninteresting proposals. What we see is more ideological gatekeeping. I have a Ph.D. in the Humanities, a M.A. in English, and a B.S. in Recombinant gene Technology. So tell me, what department is going to hire me? The title of my dissertation was "Evolutionary Aesthetics." This only lessens my being hired by a humanities department. Further, I invite everyone to come to my blog and read it and tell me, again, what department is going to hire me? What kind of dept., and what specific dept. in what university? I don't fit into the ideologies of most humanities departments. And those departments are deeply ideology- and politics-driven.

Daniel Klein writes:

All the comments are appreciated. Lotta Stern and I are very explicit about the differences between groupthink as traditionally studied and the case of academia. We suggest, however, that the differences in some ways compensate for one another in a way that sustains the main conclusions of a groupthink diagnosis. We call our groupthink take on academia an adaptation of the theory.

But I guess I still object to declaring that "it is not groupthink". Until we have a better way of talking about it, no reason to be so against the groupthink take.

There is much more to the process than inbreeding. There are other sorting/expulsion effects, as well as self-censorship and preference falsification effects. Maybe groupthink, which may be adapted to include all these, is the best handle we've got.

William Newman writes:

I actually agree that much of the observed behavior today can be called groupthink.

I was arguing against the idea, which I thought I picked up but seems to have been at least in part a misreading, that we got here by groupthink, instead of by processes like intellectual inbreeding and intellectual assortative mating and informal backscratching and all sorts of other things. If you took 50 random strangers, sealed them in a room 50 years ago, and investigated their views today, there'd be some expected baseline of what I'd call groupthink. If instead you take a research field with about 50 movers and shakers at any given time and let things like funding and promotion processes go for 50 years, I think that baseline expected groupthinkingness is likely to be less important than the other effects allowed by things like funding and promotion, and I was trying to argue against including those in the term "groupthink".

Comments for this entry have been closed
Return to top