Could Steve Levitt get into a top Ph.d. program today?

Read the debate.  Steve says U. Chicago would nix him for lack of undergraduate mathematics classes.  He believes that Harvard or MIT "might still take a chance on me today." 

I am a strong believer in having at least one top school — Chicago once played this role — which accepts virtually everybody and lets competition sort them out in brutal fashion.  I am also a strong believer in having more graduate students at top schools know economic history than real analysis.  I don’t expect either of these wishes to come true anytime soon.


Knowing that the top programs require more mathematics, he just would have taken more math classes as an undergrad.

It's a cruel world.

Or let's say you discover you want to be an economics student during your sophomore year and don't really think about graduate school until toward the end of your senior year, and realize you've spent too much of your time and money on working at a magazine and drinking. This makes your undergrad GPA middling and your math aptitude beyond statistics mediocre. Let's also pretend, just for the sake of argument, that you did terribly at both Linear Algebra and REal Analysis. And that you're more interested in economic history than in theory. What should one do then?

My solution is go to a local UTexas branch that offers some M.Sc. programs that seem interesting part time so I can keep my day job, figure it out from there.

One should also add that (at least in my impression) the placement record of the Harvard and MIT econ PhD programs over the last fifteen years or so has been better than the placement record of Chicago.

But his GRE (quant) score was around 800. Not bad.

I've always had a strong interest in econ, but unfortunetly I'm not smart enough to go to grad school. My solution: get a B.A. in econ, internet connection w/PC, visit MR every day, and be an economist vicariously.

I would be curious to know which programs professional economists feel are
under-valued at the moment. If we exclude the top ten, which programs seem
to be on the rise or attempting innovative approaches to the econ PhD?
How do 'applied' econ programs compare with traditional econ PhDs?

Why dont they make a math test or many tests just in a saturday morning?
There are many people that is more intelligent, talented, and capable than certified ones.

Sorry I correct myself.
Those people that show to be better than qualified ones do not need to go to University

Quant GRE scores are looked at very seriously by
econ grad programs. After all, quality of undergrad
math courses varies greatly.

I dissent.

If Econ wants to stay in the liberal arts, then a fuzzy undergrad trajectory is fine. But the leading lights of the field clearly want to make economics more like engineering: a valued tool for solving real-world problems. This "colder" direction makes sense given the way Engineers curretly outnumber humanities types in corporate managements and other venues were economic power is exercised.

Speaking as an Engineer Ph.D., one of the prices for making a curriculum more Engineering-like is a massive loss curriculum flexibility and the abolishment of "late entry" in particular. That is, if a student isn't fully onboard with their major (say, Mech. Eng) by their sophomore year, they aren't going to get the B.Sc. in the typical 4-5 year time frame. Moreover, a student who wants to get into a top-notch engineering grad program would be wise to seek lab work and research-rigorous electives during final two years. Notice the resemblance to what Tyler is griping about.

Perhaps the real solution is to split Econ into two distinct disciplines, the way the engineering disciplines organizationally split from the natural sciences at the close of the 19th century (eg Chemical Eng split off from Chemsitry, Electrical Eng split off from Physics, etc).
This could provide the best of both worlds. The humanities-econs would be historically literate and well-versed in a wide array of knowledge who would carry on the great tradition of intelectual pursuit exemplified by Adam Smith, Marx, Keynes, the Austrians, and many others, while the econ-gineers would be real-world problem solvers applying ideas like those found in "Freakanomics" or "The Undercover Economist."

By the way, I got a perfect score on the math section of the GRE, but since it basically tested high schol math, IMO it was (or should have been) basically meaningless. My opinion was that a guy with two degrees in chemical engineering should be able to score at least 58/60 in his sleep, and anything less would be an embarrasment. The GRE measures little more than one's ability to take tests, and I'm just lucky to be a guy who does well on tests. That ability has gotten me into lot of places I probably didn't deserve to be.

Nonetheless, my GRE scores knocked the socks off the admissions guys where I applied, and was able to get me a nice fellowship to go along with the TA stipend.

JN says discussing work experience signals cluelessness about academia to admissions committees. I believe that many academics believe that, but I wonder if they should. Maybe that attitude betrays a cluelessness about the private sector?

I worked a bunch of different jobs before applying to grad school. I can find something relevant to a doctoral program in all of them, including the big intellectual challenges that admissions committees seem to value. Am I wrong to think that committees are unreasonably snobby on this point?

I also wonder about labor economists (and others) who have never been interviewed or interviewed someone for a private-sector job.


I have been in econ for quite some time now, and this is literally the first time that I heard that econ was part of the liberal arts. Silly me, thinking I was a social scientist all these years :-).

I am a little surprised about all the complaints about the "idolization of math in econ" on this thread. That's a complaint that would have had much more bite in the late 1980s than now - most of the exciting action in economics since then has been in empirical work in applied microeconomics (obviously helped by the advent of personal computing). The most important theory developments over the same period have been in incomplete contract theory (Hart-Moore and onwards) which is quite low-tech and doesn’t require a Russian mathematician to understand.

I don’t mean to imply that going through an econ PhD doesn't require a fair amount of math - one needs it simply to understand the existing theory models. But thre majority of the people who have done well on the academic job market for rookies in the last ten years were empiricists.

jn, what about doing research for a think-tank such as Urban or Brookings? Being a research assistant at one of these places seems like great experience for someone wanting to get a PhD in economics. I have A's in multi-variate calc and differential equations but C's in statistics. Should I improve my statistics grades or take higher level math?

Commenterlein, I agree with Bartman. The rise of the empiricists is a testimony to the desire for a shift but the signalling game -- especially in grad school -- is still tied to the "I know more math than you do." I've seen one or two students who might have become interesting scholars if they could have held their breath long enough to make it past MasCollel and the prelims.

vtconomist: Don't bother retaking stats. Take real analysis and above and ace those classes. A good letter from someone credible (not necessarily famous) who says you're a great researcher who will thrive at a first tier program helps too.

"Why can't I finish an MBA or MA, work for a while and then go back in for a PhD at a later date? Or are we all washed-up by 32"

When I was 35 I asked a cardio surgeon if it was realistic to consider becoming one myself with zero life sciences prep. It would take 9-12 years to get there and he said it has been done and simply depends on one's willingness to make the sacrifices. You're never washed up but you might have to cramp your lifestyle significantly. I decided against it.

I am strong in economic theory and very very weak on econometrics. That is why I am pursuing a second degree in history. I would rather write about economic history than get bogged down by econometric minutia.

I believe that GMU has its share of distinguished faculty who could do a good job of helping you out. Some of them have been known to start blogs...

And Kyle, if you think that doing econ history will relieve you of having to worry about econometrics, you must not be reading the Journal of Econ History.

Regarding the Chicago approach to admissions, I was admitted to the U of C MBA program in 1979 with 0 years of business experience. I would not have had this opportunity had it not been for their sink-or-swim admissions policy (one benefit of which, I think, is that it allows a school to admit a very diverse set of applicants). On the flip side, at age 35 (after over a decade of full-time business experience) I was admitted as a Ph.D. candidate in UCLA's econ program. UCLA had a Chicago-style admissions policy and, again, I would not have had the opportunity were this not the case.

Regarding our emphasis on mathematics: before complaining about it, try making sense the literatures of fields in which theory generation is all based on verbal argumentation. Theories are ambiguous, incomplete and illogical and, as a result, empirical work is also a mess.


Regarding applied econ programs, the market looks pretty good coming out of one. I just finished my job search, and here's how it went: I had 22 first interviews in total (counting both the CEEE and the ASSA meetings), and I got five fly-outs for on-campus interviews. Of those, I received three offers: one from the econ department of a small research university in Quebec (my home province), one from an econ department in very decent university out west, and one from the Terry Sanford Institute of Public Policy at Duke University. In case you wondered, I accepted the latter, and they even went so far as to give me a secondary appointment in the Department of Economics.

The point is that if you have an empirical bent, the name of the program will not make a difference. You just need to make sure to take rigorous classes in grad school and to have solid methodological toolkits (in my case, applied econometrics and contract theory) on top of your major field (in my case, development) so as to signal that you can do the work. That and make sure that you have one or two single- or first-authored publications (ideally journal publications in top field journals) by the time you finish grad school.

The other things you don't want to overlook are grantsmanship and teaching. Most hiring departments want someone who will bring in money to the department, and if you can demonstrate that you can teach a wide array of classes, i.e., a good mixture of technical (math for economists, game theory, econometrics) and non-technical classes (principles, field courses) can get you a long way.

In the end, I think it boils down to the quality of advising. I was blessed with an advisor who devotes a lot of time to his students and who his well-known within development economics, and he has given me extensive comments on my papers and on how to approach the market. The problem with many top programs is that you will be lucky to spend 20 minutes a week with your advisor if you're not one of the two or three star students in the department.

The bottom line is that the market for applied economists is not as restricted as you might expect. Sure, you might not get offers from Chicago, Harvard, MIT, Princeton, Yale, Berkeley, or Stanford (although an applied econ student this year got fly-outs at a couple of those places and did quite well on the market), but you will find a very decent academic position in a research university.

bartman, feel free to correct anything in what I wrote if you disagree. The above should have been prefaced with a huge "in my humble opinion". Sorry for the long-winded comment!

One quick point I would like to make, one that nobody has touched on, is that economics is not math. While math may make economic concepts more digestable in the sense that you can mathematically set supply equal to demand or fashion some statistical equation based on lagged values of a given dependent variable. In the real world, these types of models do not make sense. To me, equilibrium theory is one of the greatest shams of all economics and there is never going to be a point in time where aggregating across N individuals or N firms, is ever going to make clear sense of human action. People are not robots, everybody makes their own choices subject to their own values, it is the subjective nature of human choice that enable economic systems, markets, etc. to work. So, while math courses and high quantative GRE scores can buy you admission into even the most mathematically averse and vehemently against mainstream economic programs, they cannot buy you the common sense it takes to recognize basic economic principles and theories. So while you are nose deep in your calculus and real analysis textbooks, lift your head up and ask if economics is headed in the right direction.

Johnny: nobody has mentioned that "econ is not math" because everybody knows that econ is not math. However, mathematics is the language of modern economics, and if you can't speak that language, you won't find a place inside the mainstream. But it seems from your comment that you're not really interested in such a place, anyway.

tylerh: as someone coming from the engineering world, I found certain econ courses to be far tougher than any engineering course I ever took. But maybe that's just me.

vtconomist: looks like the analysis sequence at GMU starts with MATH 290, 315, 316, 321 and 431.

jn: going back to school in my late 30s was no picnic, and I probably would not have been able to do it if I was married with kids. (One of my cohort was a 40-ish Indian guy with a wife and kid, and the experience almost destroyed his marrage, and it did give him a mild heart attack.) I often asked myself what I was doing sharing a crappy apartment, eating a lot of ramen and spending 70 hours a week in my office, coupled with the angst accompanying every piece of auction theory homework... But now it's over, I'm glad I did it.

Michael Ryall: about the math imposing rigor, structure and discipline: I agree. The only complaint is that some guys place more emphasis on the math than then underlying economics. But that tide appears to be turning.

Robert: Mas-Colell is the bane of every econ doctoral student. You can take a look in it at Amazon.

Bartman: I appreciate your comments. I do not deny the inherent usefulness of mathematics in economics, I am in fact a graduate student of economics, despite your comment about me not being interested in higher levels of economics. My point is just that even some economics departments that claim to focus more on less mathematically applied fields, still have the most rigid of GRE and Math entrance requirements. If in fact they want to stay consistent with their views, the entrance requirements should be changed to reflect a different aspect of their respective applicants, and their quantative skills should not be the most revered. You might say, well it is not their choice, that the university stipulates the requirements and the department has to follow suit, but I cannot agree with the argument that the admissions deparment has full blown autonomous reign over the economics department, who can certainly screen in a manner they see fit and make their claim against holding the GRE and (Insert Math Course here) as most important. I merely want to make the point, that maximizing utility functions and production functions, trying to explain the future based on the most recent past, etc. do not hold much weight in my view with respect to the essence of economics as a science, a social science if you will. Yes forecasting may be useful in telling us that every single time you will in fact be wrong in your forecast, but it is those of us that are less wrong than others who reap the benefits, I just do not believe that in order for a person to be a good or great economist that they need to rely on a huge sum of math. The only reason it appears that way is because as a herd, economists flock to the math, because they feel it can simplify the real world and explain the seemingly unexplainable. So the literature is statistics intensive, the programs love teaching real analysis and topology, and the economists who come out to speak against the 'mathematicization' of economics are declared ignorant and shoddy because they do not understand the math and how good these models are, etc. All the while, such economic theorists and applied economic historians, who try to get to the roots of the theory are denied their most useful comments and thoughts, because sometime in the past, a few mathematicians and statisticians decided that they could use their math to explain human action, market activity, government policy, etc. and so goes the field of economics. There is a great lack of economists entering into PhD programs who are interested in the history of thought, political economics of the old tradition, etc., instead you see them exhibiting herd mentality and gravitating to the econometrics 9000 course, but when you ask them to explain some basic economic principles, they throw numbers at you and fail, in my opinion, to grasp what economics is all about. Have you ever wondered why forecasters are always wrong, every single time? Probability you will say, the model is incorrect mathematically, well maybe the theory behind the models is wrong and that rational agent theory and equilibrium theory, while mathematically convenient, are 'real-world' usless unless you aim to explain what does not happen.


I suspect you are still in the first two years for your grad program, working on your courses. That at least would make your perspective understandable, since indeed much of the course work makes it feel as if modern economics was all about math and calculating equilibrium response functions and other such things.

The more pleasant truth is that modern economics is much more than that, which is an insight I personally didn't really internalize until about the fourth year of my PhD. Simply look at who has won the Clark Bates in recent years and realize that very few of these people do any work on GE or any other form of high-level theory. If you believe that rational agent theory and equilibrium theory is useless (which I personally don't believe even though I do mostly behavioral econ for a living) then you will find plenty of allies among the faculty of the top econ departments.

Bottom line: Economics has long moved far beyond the "useless" theory models many here complain about. That doesn't get anyone around working their way through introductory graduate level micro theory, and at least imho there are good reasons to force people through it. And I am saying this despite the fact that I am mathematically challenged and hence suffered quite a bit from it.

Hey marc and bartman, thanks for the great posts, I deeply appreciate the information! I am much more interested in applied econ but there is less info out there concerning outcomes so you advice was very helpful.

Oh, and btw congratulations marc and good luck at Duke!

I command italics to end!

I still don't know what applied (as opposed to?) economics is.

Can somebody tell us which departments do a good job in applied economics? non-mathematical non-applied economics?

"Indeed, anybody getting a Ph.D. is getting a degree in applied economics."

A Ph.D. in what?

Another option is to get a Ph.D. in public policy. I don't know how the programs and placement compare with applied economics, but there are very good economists and econometricians teaching at some of the top public policy schools.

Comments for this post are closed