Academic economics is more winner-take-all than you might think

by on June 21, 2013 at 7:06 pm in Economics, Education | Permalink

John P. Conley and Ali Sina Onder write (pdf):

We study the research productivity of new graduates of top Ph.D. programs in economics.  We find that class rank is as important as departmental rank as predictors of future research productivity.  For instance the best graduate from UIUC or Toronto in a given year will have roughly the same number of American Economic Review (AER) equivalent publications at year six after graduation as the number three graduate from Berkeley, U. Penn, or Yale.  We also find that research productivity of top graduates drops off very quickly with class rank at all departments.  For example, even at Harvard, the median graduate has only 0.04 AER papers at year six…

The indicating post is from Angus, thanks also to Stan T. for a pointer.

1 Jim Clay June 21, 2013 at 7:41 pm

I’m thinking they meant “average” Harvard graduate, not “median”, since I don’t think it’s possible for the median graduate to have a non-integer number of papers.

2 Tyler Cowen June 21, 2013 at 7:55 pm

Average of a series of medians, one each year, is how I read it…

3 Will June 21, 2013 at 8:22 pm

It is AER equivalents, so Journal of Economic Theory, e.g., is only .5 AER and a four authored paper is only worth .25 AER.

4 Jack PQ June 21, 2013 at 8:02 pm

Two proposed explanations are: (1) In top programs like MIT and Harvard, lower-ranked students get saddled with a lot of TA work that does not lead to better research productivity , and (2) Publishing in AER, JPE, QJE involves not only great ideas and flawless execution but also credibility, i.e. networks. While the best student out of Harvard will land a job at MIT or Princeton or Chicago (with the all-star networks of co-authors), those at the bottom of the cohort will end up merely in top-100 universities, which are fine, but where you don’t have the same network.

Credibility is important, hence the role of famous co-authors

5 Will June 21, 2013 at 8:24 pm

The point of this paper is that department rank does not matter as much as we think.

6 gogurt June 21, 2013 at 9:05 pm

+1. Networking has got to play a huge part in this.

7 ben June 21, 2013 at 9:45 pm

Definitely- also think about the political economy of journal refereeing. Suppose each paper is assigned n identical referees by the editor. The editor is higher status than the referees, and everyone is very busy- referees don’t want to waste time, and the editor is not going to read the paper carefully enough to form strong opinions about it. So, your objective function as a referee contains the arguments of time, and not getting embarassed in front fo the editor. The Nash Equilibrium is to write favorable reports for papers written by people the editor might have a good impression of, and negative reports for everyone else. The only way to break into the club is to co-author papers with someone who editors are likely to know.

Of course, it doesn’t always work that way- sometimes the referees are not so concerned with what the editor thinks and/or are willing to devote enough time to thinking about the paper that they judge it based on its merits. But, in my experience that’s more the exception than the rule.

8 Dismalist June 21, 2013 at 8:25 pm

Not surprised. Thank whoever that in America there is some chance for those outside the early detection cartel to succeed in the academy, even if small.

System is working!

9 Kevin June 21, 2013 at 8:32 pm

Interesting. Make me want to replicate it in my ow field.

10 zbicyclist June 21, 2013 at 8:37 pm

The field of pain management?

11 TMC June 22, 2013 at 10:32 am


12 Paul June 21, 2013 at 9:06 pm

They cover grads from 1986-2000. In 1993 the Mosaic browser was released and the internet leveled the playing field at least with respect to access to working papers. I’m guessing that grads from the top departments in the decade prior to 1986 would have had a greater advantage in that respect. I recall Harry Johnson writing a long time ago that getting early access to current research in progress was the main advantage of going to a top grad program.

13 Dismalist June 22, 2013 at 12:14 am

Anybody who knows the name Harry Johnson can’t be all bad, or all wrong. 🙂

14 Turkey Vulture June 21, 2013 at 11:08 pm

Eliminate tenure and make student loans discharageable, and this is just another issue that is interesting almost entirely for the sake of knowledge.

As it is, though, we just have a bunch of expensive rent-seeking which might do a better job at eventually identifying the meritorious than my priors would have suggested.

15 Wiki June 21, 2013 at 11:52 pm

I’m betting that eliminating tenure would make this problem ten times worse with more pressure on current pubs and would add more importance to the well connected and fashionable.

16 Turkey Vulture June 22, 2013 at 12:10 am

My sense is that eliminating tenure would be an important step in de-linking research and teaching, which I think it is essential to unbundle, for the cost efficiency of both. There is probably a better route to the same end though.

17 Jack PQ June 22, 2013 at 8:44 am

Tenure is just salary taken in the form of a perk. Contracts are voluntary, thus mutually advantageous.

I think here tenure is a straw man, not to say it is unimportant in other regards.

18 Turkey Vulture June 22, 2013 at 9:17 am

This just means tenure is a piece of the Rents being sought, a piece that happens to make future rent-collection much easier.

Also, because this rent is given as a fairly discrete event, it would seem to lead to significant over-investments by those competing to receive it.

19 Tom West June 22, 2013 at 10:27 am

I have to that tenure appears to be a much bigger deal than it actually is.

How many high-ranking professionals get fired (as opposed to laid off when business needs changed or they need to cut expenditures? Not many.

What makes professors safe is the fact that that they are key employees in a very stable institution. Statistically, I doubt tenure changes that much at all.

The one big exception – if there’s no mandatory retirement.

20 Bill June 21, 2013 at 11:36 pm

This must explain why my papers were not accepted by AER.

21 sam June 22, 2013 at 12:09 am

Not a great compliment to the admissions process.

22 Rafael June 22, 2013 at 12:43 am

The attempt at measurement of research productivity is reaching ludicrous levels among the economics academia. Considering the huge problem of measuring research output quality in terms of scientific papers in the field of social science I think that try to measure that stuff is extremely problematic. In physics, at least the research is pretty hard (though modern research in stuff like string theory is even making the research output of the hard sciences less credible) so it makes some sense, but in economics?

This only shows that getting into the PHD program of a top 5 department is just the first competitive barrier into entering the elite of the academia, that one needs to become one of the favored students in a top department to get co-authors. The profession of academic economist is now a new type of professional competitive sport – a game to be played by a small set of people gifted in math: 99% of the research is irrelevant to anybody outside of the profession and only serves for players inside the profession to try to outrank other players.

23 Enrique June 22, 2013 at 1:05 am

My guess is that the difference in quality between, say, a Justin Wolfers and one of his “median” peers is not that large, like the difference between Usain Bolt and the runner-up in the 100m dash … so it definitely looks like a “superstar” effect is at work here

24 Jacob Lyles June 22, 2013 at 1:20 am

It seems like Economics is going to be one of the last disciplines to adopt open publishing. It is commonplace in biological disciplines (PLOS) and computer science (JMLR), Physics and Mathematics (Arxiv). I wonder why that is?

25 Jack PQ June 22, 2013 at 8:45 am

Arxiv is a working paper archive. We have repec and ssrn for that.

26 Seth June 22, 2013 at 4:15 am

Do we know what jobs these new PhDs hold? Not everyone goes into academia, but the top students at top 30 schools almost always do. Publishing AER equivalent requires a lot of work. Outside of academia, the incentives for those types of publications are not nearly as strong — and are not required to maintain one’s job.

How many AER equivalents did Jason Furman or Peter Orszag have during their first 6 years after completing their PhDs? I suspect that they were near the top of their classes using conventional measures. I’m guessing they wouldn’t fare well under an AER equivalent measure because they pursued careers where it was more important to get work out fast and to appeal to a broader audience than AER readers.

27 Mario Rizzo June 22, 2013 at 12:13 pm

The criteria being used for student ranking in a top department and ranking within other departments and publishing in the AER are all the same. (That is, they are all highly correlated.) The more your working papers look like AER articles, the more likely you thrive on the other margins.

The puzzle is why do the lower ranked students at the top departments not do as well as the higher ranked students at the lesser departments. My guess is that faculty members do not pay much attention to or expend much effort on behalf of lesser ranked students, anywhere. Who wants to have the students who are not the best in a department? What will your colleagues think of you?

The progress of normal science is the polite way to speak of the psychological insecurity of academic economists.

28 Eric Rasmusen June 22, 2013 at 11:19 pm

Table 1 of the paper is fascinating. At the 90th percentile, the top schools are (in no particular order) MIT, Rochester, PRinceton, Harvard, with a distinct gap after that. At the 60th, they are MIT, Princeton, Rochester, British Columbia, CarnegieMellon, San Diego,and then nobody close, with Rochester-Princeton-SanDiego tied in a bunch at the top ahead of the rest.

Is this a matter of which schools produce the best dissertations? Maybe nothing else gets published in good journals in the first 6 years. If so, the picture might look very different in the next 6 years, since dissertation quality depends so much on the advisor adn we would not expect faculty quality to fall off with department rank as much as student quality.

Ah– also see Table 2, which has class sizes. Rochester, UBC, SanDiego have tiny doctoral programs, so they can select just the best applicants and fund them fully, I bet, and they can give them lots of attention. This makes me feel better about we at Indiana’s b-school losing a PhD candidate to Carnegie Mellon!

29 djk reversal review June 24, 2013 at 8:18 pm

It’s in fact very difficult in this busy life to listen news on Television, so I only use the web for that reason, and get the newest news.

30 revitol hair removal cream reviews June 25, 2013 at 3:55 am

Provide anything else significantly more distinctive??? For instance, critiques or resources with reference to homemade beauty tips and secrets.
How can I sign up for your articles?

31 Buzzluck June 25, 2013 at 9:14 am

What follows usually are recommendations pertaining to widespread video
games such as Craps, roulette, Baccarat, blackjack and many others.
Suggestions some information on WinHoldEm, the most well-liked

32 Classy Coin June 25, 2013 at 9:47 am

Unlike our prior morals, My spouse and i all of a sudden found which it could possibly
in fact possibly be good to become bad with this Real time Gaming
video slot leisure. Inside Las vegas, Classy Coin are well known because of their stay entertainment, free buffets and much more.

33 discount party powder June 25, 2013 at 3:12 pm

If you are going for best contents like I do, just pay a quick visit this web
page daily for the reason that it presents feature contents, thanks

34 miami Club casino online slots June 28, 2013 at 6:57 pm

With Badugi, a hand consists of some playing cards, certainly
where an gamer provides several probabilities
to attract or even transform this mix, to accomplish the best side doable and get the marijuana.

On the web bingo sites provide enjoyable attributes to help attract far more miami Club casino online slots people to be
able to engage in online games on their websites.

35 binary options broker Reviews July 2, 2013 at 12:26 pm

Exchanging binary options broker Reviews calls for a comprehension in the areas,
and knowing precisely what what you are doing, a
comfortable stream connected with regular income can certainly help
appear towards you. binary options broker Reviews is one of the greatest binary options broker Reviews
agents on the net these days. There’s no liquidity to be able to this sort of investment decision because you never ever really unique a great root asset.

36 Online Casinos July 10, 2013 at 12:42 pm

You need to consult the vendor to use the French craps bets when
they are generally around the tire and not the table, and while many
people commonly are not marketed with Online Casinos, the
particular supplier will recognize what direction to go.

gamblingRalf endorses Anti-Martingale Program (if any kind of process in any way, of course, if roulette whatsoever.
.. ), mainly because it isn’t really seeing that nerve-wrecking since the people Online Casinos you don’t have to place cosmic gamble to be able
to win back the bank account adjust.

37 lucky red July 11, 2013 at 4:53 am

We shall lose time waiting for brand new interesting options
that make the game even more easy in addition to enjoyable.
A number of people choose to engage in using this method
just for pleasure.

Comments on this entry are closed.

Previous post:

Next post: