Pedigree bias in economics

Coastal Elite, a loyal MR requester, requests:

Why is there more "pedigree bias" for hiring of Economics faculty than
in other disciplines? For example, at any top 30 school, 95% of the
faculty will have their PhD from a top 10 school. In other fields, this
will not be the case. Some other disciplines have a much stronger
tendency to co-author than Economics, which should decrease the signal
to noise ratio on a CV. This should imply that pedigree bias should be
even stronger in disciplines with more co-authoring, but anecdotally
this does not seem to be true.

And very often it is the top six rather than the top ten.  I don't myself see the factor of co-authoring as the path to an answer here.  The default answer is to invoke a relatively high importance for IQ, in economics, combined with the relative absence of prodigies, compared to say math.  When prodigies and autodidacts are possible, top performers can come from anywhere.  But smarts and training and networking are all required.  The combination of those barriers give a big advantage, justified or not, to the very top schools.  It is usually believed that a candidate from a lesser school is lacking in at least one of these yet all are required for a big career success.  So maybe a "multiplicative model" of achievement in economics plays a big role in pedigree bias.

On top of that, economics is a relatively unified field with relatively homogeneous metrics of quality.  A school ranked #23 can "play it safe" by hiring a lesser MIT grad, rather than the best student coming out of Emory, and more or less know what they are getting and agree on that.  This tendency in the market is probably inefficient, relative to a first-best with a higher degree of intellectual innovation and less concentration of rewards across the top graduate programs.

Also, the top programs are good at doing admissions — surprisingly good in my view — and pre-selecting the kind of talent the rest of the profession is looking for.

Which other fields have a comparable "pedigree bias"?

Comments

The same rule holds in any field where there is little objective evidence to distinguish quality and a particular approach finds itself threatened by contrary viewpoints. Comparison with String Theory, Philosophy, and English might be interesting.

My experience after 20 years at a top 50ish liberal arts college is that I would MUCH rather hire the top student from a second or even third tier PhD program in Economics than someone in the lower half of his/her job market class from a Top 20 program (we can't afford the good ones). Being blinded by a PhD program's pedigree is the most consistent mistake in the job market we've made over the years. I think that part of it is that many top 20 candidates simply cannot teach undergraduates (at least not non-top 20 undergraduates) with any success at all. Obviously, my world is a different one from that of PhD programs, but it remains true that we have done much better with the best of the second best strategy.

Being a prodigy in mathematics is little related to mathematics research, so that explanation doesn't hold water.

Getting back to your post from the other day, how will the internet change this?

cip--

Hmm, 706+504 = 1210 is among the highest fields there. So economists scored among the top among prospective graduate students. (Physics, Math, Philosophy are higher by a quick glance, but I don't see others.)

But cip has never been good at reading or understanding figures, judging from comments here and cip's blog.

http://www.jstor.org/stable/3593086
Burris, Val. 2004. "The Academic Caste System: Prestige Hierarchies in PhD Exchange Networks." ASR 69:239-64.

Abstract
The prestige of academic departments is commonly understood as rooted in the scholarly productivity of their faculty and graduates. I use the theories of Weber and Bourdieu to advance an alternative view of departmental prestige, which I show is an effect a department s position within networks of association and social exchange-that is, it is a form of social capital. The social network created by the exchange of PhDs among departments is the most important network of this kind. Using data on the exchange of PhDs among sociology departments, I apply network analysis to investigate this alternative conception of departmental prestige and to demonstrate its superiority over the conventional view. Within sociology, centrality within interdepartmental hiring networks explains 84 percent of the variance in departmental prestige. Similar findings are reported for history and political science. This alternative understanding of academic prestige helps clarify anomalies-e.g., the variance in prestige unconnected to scholarly productivity, the strong association between department size and prestige, and the long-term stability of prestige rankings-encountered in research that is based on the more conventional view.

Law, for academics and non-academics. In what other field are practitioners asked so often for their class ranking a decade or more after they complete their program?

The only way for a non-top 10 department to "Crack" the system is to develop a reputation for producing students who are better -- on some relevant margin -- at publication in the top journals than typical students in the top 10. Usually this came from developing narrow technical specialties. This is how Minnesota, Rochester, Caltech developed a rep that allowed them to outplace students from many "higher ranked" departments. Other schools also had a brief period of placement success (cf. Purdue in the 1960s with Sonnenschein, Ledyard, Kamien, Schwarz, etc.)

What is hard to do is develop a rep in some field that the top departments do not value and hope that it becomes important. E.g. Experimental is prob the best example

How many markets upon closer inspection look mostly efficient?

I'm a second year PhD student in Harvard Economics, did my undergrad at Berkeley. For readers unfamiliar with the inside-baseball, Berkeley is ranked somewhere in the 3-6 range, while Harvard and MIT duke it out over the 1 and 2 slots.

I was always impressed by the grad students in the Berkeley econ department, but as a group they're just not in the same league as my classmates here. The drop-off in input quality when you leave 02138/9 is pretty steep, and it must be even steeper when you look past the top 6-10. I mean this in the least arrogant way possible.

The consensus is that most of the value-added comes from interacting with your classmates, so the peer effects are dominant. In terms of teaching, I actually found the Berkeley faculty put in more effort, but the courses here are at a higher level. (Since Raj moved, there's a DiD estimator available...)

I can't speak to the quality of the advising elsewhere, but the dominance of the top schools self-perpetuates pretty easily. Because 2/3 of the faculty at the top 6 went to Harvard or MIT, any potential advisor I end up with has personal relationships with members of the faculty anywhere I'll be applying for a job. It's a comforting thought.

Given the comments by some of the posters, perhaps it is not "bias" but rather quality that differentiates. Perhaps you would have to first prove there is a bias by showing that there were two equally well trained economists, and then work to find a bias.

As Clarence Thomas would say, sometimes there isn't bias.

Actually, Caltech is the only of the above-named top eight that even lists its faculty with their degree institutions. So perhaps that says something about the importance of pedigree in physics.

The Supreme Court.

Is there a similar "pedigree bias" among economists at the Federal Reserve or at in the various branches and agencies of the U.S. Government?

What are the chances that hiring biases at the Government and at the Fed in large part control the hiring biases at the "top" econ grad schools?

It seems to me that given a lack of objective measure of quality, one relies on brand, and especially the brand that everyone else is buying, thinking it must be the best value because everyone else is buying it.

Clearly no school of economists has any more credibility than the others, nor is there any evidence that the standard subject matter everyone seems to use has that much to recommend it.

If one school, that was obscure and in some backwater, produced three economists who were in the Obama administration laying out policies, based on their unknown or rejected theories, that when implemented delivered employment growth as they predicted, every other graduate of that school, and that school's rejected faculty would suddenly be the ones in highest demand.

The one difference between today and eight years ago is the absence of calls to cut taxes to produce economic and employment growth - that dogma seems to have been shelved given the eight years of tax cutting failing to do either. However, I'm sure that an economist calling for hiking taxes to create jobs would be called an idiot no matter which school produced them, so we never see real solutions coming from economists, and the growth producing tax hikes will occur over the objections of all economists who say anything at all.

So, with no economists able to predict anything with their theories, the only other criteria is where they got the degree - even the Nobel is no longer evidence of anything it seems base on comments about Krugman and his Nobel.

For historical reasons the U.S. dominates the study of market economics and economics doctoral education to a greater degree than is the case for the physical sciences or engineering. Not surprising since until recently many countries were hostile to market economics.

It's easier for these top U.S. schools to boast that they are the best in the world when they don't have to compete with the Russians, etc etc...

Easier to dominate a relatively small pond.

As noted by another commenter, the legal field places huge emphasis on pedigree. Probably 50%-66% of professors come from Harvard and Yale. Probably 90% of the rest will come from the top 14 schools. And your GPA and class rank are relevant for many years after you graduate.

I think few professions can match the prestige-centricity of the law.

"Not surprising since until recently many countries were hostile to market economics."

I don't doubt the US dominates economics research, but this doesn't sound like a good explanation. All Europe runs on market economies -- even China has a market economy! Just because these states supplement their markets with redistributive social services doesn't mean they have any less interest in the behavior of markets (indeed in many cases those social services are provided in markets).

And even if they were antagonistic toward markets, that still doesn't explain why they wouldn't build economic departments around critical analysis of market dynamics, etc.

The current admission requirements at top econ schools would exclude some of the faculty at those schools. Still the academic skills needed to get admitted to a top program are the same academic skills needed to succeed in the field. Probably in part why top economist often do their best work when young.

Econ PhD's even from top schools may not earn more then MBA's from top schools, unless they are outstanding PhDs or write books.

Law certainly, look at the Supreme Court. Yale, Harvard and Northwestern perhaps because President Ford wanted a Big Ten guy.

You do seem some of the same in Medicine. Some medical schools dominate in some specialties especially with regard to research.

No one from outside the top 50-75 will EVER don the halls of Harvard, MIT, Chicago, or Princeton's economics department as a full professor. That seems like pedigree bias to me unless you think talent is that skewed and a meteor will never attend a non-top 50-75 place.

Law has a similar pedigree bias - 10 law schools provide about 45% of all law professors. See http://www.aals.org/statistics/2008dlt/schools.html.

"I absolutely buy that the average technical skill at Harvard/MIT/Stanford GSB is better than anywhere, but it's completely false to suggest that better researchers will come from these places."

Funny -- I'm at one of these places and my perception is the exact opposite. Even though of course you will have some students with amazing technical skills at Harvard/MIT/Stanford, you also find such students at top 20-50 places. However, the students from the top schools seem much more able to convince a seminar audience that the topic they are working on is important and that the way they tackle a question is sensible. People from less highly ranked places often give absolutely dreadful presentations, even though in principle their papers could do really well, if only they knew how to sell them.

Don't underestimate the importance of marketing, folks.

Are these ratings of responsibility for the current economic crisis? (Harvard most responsible, and so forth?)

John Thacker - On the cited lists economics grad students rank 10th, 12th, and 13th - a versatile and solid Lake Woebegone above average, but excelling in nothing by my reading.

Consulting. Investment management. Any field where it is very difficult to judge quality, so we go by prestige to be safe. The lack of postdocs and short research record as compared to the sciences or math (where a new hire might have ten+ publications) means reputation of the school plays a larger role in hiring. Math for example is flatter because it is ancient and people more or less agree on what at least some important problems are. In economics whether something is important is more ambiguous, and the mathematics are generally not so difficult--it is the mapping of problem to math and selling it well. Unfortunately I fear the concentration effect is a sign of ill intellectual health in a discipline and a sign that academic politcs is playing even more enormous a role than usual.

@DanC-

Yes, of course, you're entirely right. I was commenting from memory--and forgot that Ginsburg was the first woman to be on both Harvard and Columbia Law Review.

And I can't believe I mixed up the Northwestern grads, considering they have two Supreme Court Justices (Stevens and Goldberg), and my alma mater (U of C) has only two rejected nominees to its name. Galling.

>No one from outside the top 50-75 will EVER don the halls of Harvard, MIT, Chicago, or Princeton's economics department as a full professor. That seems like pedigree bias to me unless you think talent is that skewed and a meteor will never attend a non-top 50-75 place.

boyz2men - John List (Chicago) famously got his Ph.D from the University of Wyoming I'll bet there are others.

Speaking of pedigree I heard somewhere that the best predictor of success in Ph.D programs (probably economics) was the US News rank of the undergraduate institution you went to. Hilarious.

Do people see this situation as a good thing or a shortcoming?

If it is viewed as a negative, does anyone have the ability and concentrated interest to do anything about it?

I think it's obviously a negative. Why should economics be so different from the sciences it wants to emulate in the long run? I can think of one theory. There is (I assume) a smaller marginal cost to add one more economics student to a department. But, to bring in another student to our lab is a nightmare despite the happy face put on by the PR department. So, maybe the top schools really can absorb all the top students in economics.

Why do you say that the chess prodigy Magnus Carlsen came from "anywhere"?

He is from the outskirts of the capital city of a country that has the highest Human Development Index (HDI) in the world, from a part of the world where chess most definitely is played and valued.

Re: anonymous "I was always impressed by the grad students in the Berkeley econ department, but as a group they're just not in the same league as my classmates here. The drop-off in input quality when you leave 02138/9 is pretty steep, and it must be even steeper when you look past the top 6-10. I mean this in the least arrogant way possible."

Wow, I went to undergrad in 0213* and am doing a PhD outside of 0213*, and I haven't had the same impression in terms of drop-off in quality. But if you replace the word "quality" with "arrogance" then maybe you have a point. It's hard to distinguish between quality and arrogance in terms of explaining this pedigree bias.

And in response to the comment about peer effects being everything, that's more likely to be true at a place where the faculty won't give you the time of day.

And lastly, if you were really so smart, you'd have recognized that identifying yourself as a Harvard 2nd year from Berkeley doesn't make you all that anonymous.

Variance in graduate student ability is very high, without postdocs there is relatively little info to go on. As Steve Horwitz said, many schools would probably prefer the top student from a top-30 school to a mediocre top-6 school candidate, but it can be hard to identify that diamond in the rough. It is then considered less risky to go for a top-6 school candidate.

As a soon-to-be newly minted PhD in theoretical astrophysics, on the market for postdoctoral jobs, I would have to agree with the commenters who suggested that in physics, astronomy, and related fields, the quality of your postdoctoral research is vastly more important than the school which issued your degree when applying for tenure-track faculty jobs. In 5 years, no one will care that I am from a small school without much of a reputation, only what research I manage to accomplish. Actually, what is far more important, especially for me now looking for postdoctoral positions, is not where I went to school, but who my thesis adviser is - in such a small field, most of the people who might hire me have a personal relationship with my adviser and will rely on that relationship to judge how much weight to give his recommendation.

Is it safe to assume that Professors of Econ are more inbred than those of other fields :)

Regarding the relative absence of prodigies when compared to math--are there really a significant number of math prodigies out there who pursue careers in math but that don't attend the top grad schools for math?

For economics it seems that the most important part of your education for making your career is your undergraduate education, which will determine whether you get into a top grad school. Without coming from a top school, it doesn't matter what you've done after undergrad if you want a job in academia? That seems like it can't be good for the profession...

Given the nonexperimental nature of empirical papers, and the applied nature of theoretical papers, economics research still fundamentally boils down to writing persuasive essays. Therefore, there is an artistry to it that is very social and involves a careful placement of new ideas within the context of old ideas in a way that will persuade editors/referees/conference-organizers, who are themselves established researchers. Put another way, being familiar with the audience is key to persuasive essays -- and top schools put students in close contact with the influential members of the audience.

If empirical economics were able to experiment, t-statistics and simple averages would be enough evidence. It can't, and so time series methods and its host of related issues are important, and particular assumptions (SVARs, GARCH, etc.) must be made that are a matter of taste (and rhetorical argument). If theoretical economics were math, the results would need to be cutting edge mathematical proofs. But it is applied math, and so the particulars of the environment are important to the proofs that often are a particular translation of an existing theorem from elsewhere. These particulars are a matter of taste (and rhetorical argument).

In the past, 30 years ago, the inbreeding problems in economics were more indentifiable because the top general interest journals (AER, EJ, JPE, QJE) entertained critical commentary pieces. Today, critical commentary is shunned by these journals; they have become "hostile" toward critical commentary submissions. This is consistent with the wisdom of Adam Smith in warning against clubbishness in the ivory tower that Klein alluded to and the folly of Stigler for assertions to the contrary of Smith.

In the good old days, thirty years ago or so, Economic Inquiry published an article entitled "Life among the Econ" that talked about the lack of a pecking order amongst the econ. Those days are gone because, again, critical commentary has been crushed by the top journals in our field.

The comment from anonymous on page 1 on the "steep gap" between Harvard and Berkeley is getting a lot of discussion. But anonymous is not so anonymous, because there are only two Berkeley undergrads who are 2nd year PhD students at Harvard. Here's the best post from the thread (Page 7).

http://www.econjobrumors.com/topic.php?id=5330

I would respect a confession and a heartfelt, thoughtful apology. It's just wrong for the guilty party to suck the innocent person into this mess.

Asking for the guilty party to write an apology is too much. I'll make it easy. All this needs is a signature.

----------------------
Dear all,

I'd like to apologize for my thoughtless comment on MR. I have been very impressed with the academic environment in 02138/9, but I got overzealous to the point where I inappropriately put down graduate students from all other zipcodes.

Upon deep reflection, I realize that it was inappropriate for me to make sweeping statements about the difference in quality of graduate students at Berkeley in comparison to graduate students at Harvard. Being a second-year graduate student, I've barely spent any time watching students from either school present their research, so I really had no clue what I was talking about. I made comparisons based on performance in core courses, which I know means very little for future career success. As a naive second year, I've accomplished little in my life other than taking courses, so forgive me for deluding myself into thinking performance in these courses is the be-all end-all of life!

I do believe my theory of the causes of "pedigree bias" is correct. Top schools get top faculty and top students who all benefit each other, on average. Top faculty are also better networked and can better place their top students. However, I know there are plenty of high achievers who don't get into Harvard or MIT. The odds are that several of these students are going to produce better job market papers than me, so the gap between my program and theirs can't possibly be as large as I had claimed.

I am guilty of trying to perpetuate a myth that research output from Harvard/MIT is so fantastic that future employers need not bother to take a serious look at job market packets from outside my zip code. But I realize now that my future employers are now a lot smarter than I gave them credit for, and my Harvard name alone isn't going to guarantee me a top-flight job when I graduate. My colleagues are right that self-aggrandizing behavior does no one any good. I'm just going to work hard, take advantage of the opportunities I have at Harvard, and let the quality of my research determine my job market outcome -- no more deception.

Also, this experience has allowed me to gain a better perspective on life. I studied economics because I wanted to add to a stock of knowledge in the world, not because I valued competing for a top academic position. Even if students at other graduate programs didn't have quite the admissions profile that got me into Harvard, they still have unique interests and ideas to offer to our research community. Ultimately someone has to get the top jobs, but more broadly it makes little sense to focus too much on an implicit model of higher-ranked schools producing vertically differentiated researchers, when unique interests and creativity allow for such horizontal differentiation in our field. This insight makes my claim of a "steep quality gap" between schools ever more dubious. I apologize, and I look forward to seeing what each of us can accomplish with our unique talents.

Sincerely,

___________________

A key feature of Economics is that 2 of the top 4 journals are permanently related to (1) a Chicago ├ęditor (JPE) and (2) an MIT editor (QJE). To start a successful career, good Phd student of either Chicago or MIT university will publish their job market paper in these two journals. Then Merton's (the father) Matthew effect with increasing returns applies.
Solutions: consider that top journals are public goods for the profession, and allow editors of other universities for JPE and QJE, or downgrade the rating of publications in QJE by MIT members, and in JPE by Chicago members.

Astronomy.

Caltech, Harvard, and Cambridge are known as the "holy trinity" for a reason.

There is a reason for this: to succeed in Astronomy requires access to the "big glass." Historically, the big three had the best observatories (eg Mount Wilson and Mount Palomar were Caltech owned and operated).

Now that the big observatories are so expensive they requires multiple governments to pay the bills, the network effects are shifting. For example, Hawaii and Arizona now have top -notch Astronomy programs because they are operators of premium government funded observatories ( Mauna Kea and Kitt's Peak, respectively) where their faculty get preferential access.

http://www.econjobrumors.com/topic.php?id=5330

Dropout rates for medical school seem to be in the 10-12% range, which is fairly low. It's certainly much lower than PhD programs (~40-50% although statistics are hard to find).

And most PhD students in econ have published nothing by the time they graduate and often nothing for years after. I was just looking at some guy's CV who published nothing the whole time he was an assistant prof at a top 10 school and then he got hired by a 2nd tier school after the first one fired him and finally started publishing.

Comments for this post are closed