Graduate school advice

A loyal MR reader asks:

I am now beginning the process of choosing classes for next year. I thought your advice might again be useful. I am in the unusual position of finding nearly all fields potentially interesting. I also feel relatively capable of pursuing most of them, with the exception of pure theory or pure econometrics…I am tempted to do economic history + something else. If I do history, perhaps I ought to do finance, micro, or metrics in order to signal technical capability?

If you were in my place, what fields would you choose? Are their particular people…at XXXX…whom you would absolutely want to take a course with/work with? Is it possible to be a macroeconomist without doing extremely technical work? Are some fields more tolerant of heterodox views than others? You told us [once] that you thought econ. grad. school should be Micro 1 – Micro 16. Does this mean I ought to take more micro next year? Given my limited ability to know where my interests will lie in the future, how should I think about this decision?
My advice here is simple: 

1. To potential academic employers you are defined by your job market paper, your letters of recommendation, and by your publications, if you have any.  Your formal "fields" aren’t that important, nor are your classes per se.

2. Pick classes to learn skills and choose your classes on the quality of the professor, not on the topic per se.

A quick classroom visit often reveals this quality within thirty seconds.

3. Pick a mentor that you, on a personal basis, relate to very well.  This is of extreme importance.  If he or she doesn’t like you, all is lost.

By the way, here is Ben Casnocha’s advice on how to be a good mentee.  What other advice would you all give this person?   


But how would this person in a comparably new setting, determine the quality of the professor? A well known, well published professor need not be a good teacher nor for that matter a good mentor. My perspective is that a professor looks at graduate students as potential coauthors who are willing to do tedious data processing and computer programming. This is not a bad thing since it is a first and crucial step in developing a thesis and potentially getting publish while in graduate school. But what if this person did not have the requisite set of skills that a professor is looking for? It is possible then that the student ends up casting about in a lot of fruitless directions.

Take courses that cover material that are important to your research agenda that you would find difficult to cover on your own. For most people this means more metrics, more metrics, more metrics.

Even if you do not want to do pure econometrics don't shy away from taking that as your major field. You will probably never be sorry at the dissertation stage that you have too much of this sort of knowledge.

This was the best advice I was given in grad school (by another grad student who had done the same). Neither of us does "econometrics" as our research agenda now, but I think it has paid off in terms of our applied research.

>3. Pick a mentor that you, on a personal basis, relate to very well. This is of extreme importance. If he or she doesn't like you, all is lost.

This is of fundamental importance, but isn't sufficient.

You also want an mentor who knows what they are doing. Some (well-meaning) mentors suggest terrible problems, or don't know that they have special talents.

Take the time to look up former graduate students of anyone you are considering for a mentor. If you see a pattern of long time to graduate/non-graduation, don't work for that person.

I strongly second Tom's point. The most important thing is to do what you are personally interested in. An incredible amount of work goes into writing just one paper. If you are not personally interested in the topic, you will not enjoy the work and the experience will be awful. You can always take more metrics classes in addition to your fields of interest to pick up more tools.

(I'm a 4th year econ grad student)

I'm done with my undergrad, but I've been taking a few math classes the last several semesters (to prepare me for a Masters or PhD program). This fall, I'll have the opportunity to take "Math Economics" or "Real Analysis".

Which one is going to help me more for 1) getting into a program, and 2) actually using what I learn from the class in a program?


I agree with the third point. I can tell you from hard, hard, HARD experience that an ill-suited adviser choice will kill any enjoyment you get from graduate school and also diminish your prospects on the job market; since their letter carries the most weight in your job market packet.

Also pick don't take too many classes. The goal in graduate school is to learn to do research not gather massive amounts of tangential knowledge. The general idea is that by the end you should not need to take a class to learn a new field but instead just go out and read the papers on your own.

Pick classes to learn skills and choose your classes on the quality of the professor, not on the topic per se

Very, very true. The most pertinent topic in the world will be a complete waste of time with a mediocre professor.

A brilliant professor will teach you a lot you can use on your preferred topics, even if the topic he's teaching seems irrelevant to your interests.

"Given my limited ability to know where my interests will lie in the future..."

I think I read in something from the Harvard Business Review that interests tend to stay the same throughout life. Of course, you have to draw a big fuzzy box around an interest, like technology. So, if you can figure out a good way to triangulate your interests and give them a name, you might be able to have some confidence that you can use them as a solid foundation.

FWIW (I'm a first year grad student), my experience has been that taking more economics or sociology courses will, generally speaking, be of no use to you in most PhD economics programs. If anything, they're of negative value (not that they couldn't be interesting or useful after grad school -- I'm just saying, they're not going to make grad school less painful). Just take more math. It is impossible to overstate how useful taking more math will be. I'm not sure how useful computer science is, though knowing Matlab and Stata backwards and forwards seems like it would save a few headaches.

I'm 2 years out of my Ph.D. and now with a tenure track job. I do applied micro development and went to an Ag Econ program.

1.) Take math econ, take advanced micro, take econometrics in undergrad. Yes more math will help, best undergrad class I took was math econ.

2.) I lucked into a great advisor. Advice find other profs with similar interest. Take their classes (if they are decent teachers), write papers in the class that you will later use in you dissertation. Getting a good second or third letter on the market will help a lot.

3.) Find an advisor who will help you carry out an idea. Step 1, find an idea. Work hard on your second year paper it will completely change, but try to make it your job market paper or try to get a first publication or revise and resubmit.
If you aren't going to try to get a job in a top R1, think of submiting to a good field journal.

4.) Present your work everywhere you can. Pay to go to confrences, do departmental brown bags, have friends at another school do a departmental brown bag there. By the time I was on the market I had presented my job market paper at least 5-6 times (confrences, home university, and development organizations). Also present to non-econ audiences. Use Tyler's theory of working hard to get an explaniation you could give to your gradmother.

Tyler has summarized the only advice a grad student needs to be successful. Number one and Number three I followed to a tee, and had a great placement. I think in reality, number three may be the most important of all these. This means looking for someone who may not even be the so-called best in their field. For one, if you have someone like this, it's almost certainly the case that they will write you a good letter of recommendation and recommend you to people they know. But secondly, completing the dissertation is already so grueling. Having someone who likes you as a person will serve to get you over numerous roadblocks. I know that from experience. It was the single most important decision I made to pick my chairman. Far more than picking my fields, or even picking my university.

Of course, ex ante you don't know who this will be, but that's why it's important to spend time talking with faculty regularly about projects. You'll get a feel then whether they are (a) supportive of what you're working on and (b) whether you work well together. I knew it when my adviser and I spent 3 hours talking about Sergio Leone movies that he was the right person for me.

The one thing I instinctively knew to do, but did not do, was #2. My graduate program seemed to put less of a premium on this, and so I took ultimately three fields, but only one course in which I learned anything lasting. And that was a labor class. I only took one labor class, but the class was so enriching and the professor so good that it left a permanent mark on me and what direction I took. In retrospect, I wish I had taken a money field because we had an excellent money professor, who was one of my favorites in the department. But I'm an applied microeconomist, and hated macroeconomics, and so couldn't figure out why I should do it. But had I replaced IO with money, I would be a much better person today for it. Lesson learned.

I think there are still ways to follow this general advice as a tenure-track assistant professor. It may mean, for instance, choosing co-authors with whom you get along, or like #2, choosing co-authors from whom you can learn a lot. I am still in a position where I feel like I only have things to learn from people, and so maybe I'd feel differently if that wasn't the case.

But Seth is right, too - ex ante, you need the math. Properly done, economics majors should probably start freshman year as majors the way engineering students do. Taking four years of packed math, stats and economics courses seems the only way to really prepare oneself for graduate school. It was, for me, like a ten foot crevice separating undergraduate from graduate training, and it really takes for most of us a long running start to make it successfully. I got lucky.

Be open to the idea that academic employment will/need not be in an economics department.

I had no idea that economists taught first year contracts and torts until I read this story from the current Forbes:

(link from Greg Mankiw)

"One more thing. I know they say there is a tradeoff between picking a young advisor versus selecting an established one. In my opinion, there's no tradeoff. I had one each for my MS and PhD program. I had just as much time with the established professor, plus money, and instant clout. You can't research the advisor decision enough. But, this advice I put in with the finding a mentor advice and advice to floss (your advisor is not your mentor, btw. A mentor should be interested in your success, not just how to use you further their success--another problem with young advisors). I've never really met anyone who does it to the extent they are told to."

One comment and one question:
1. You were very lucky that your older, experienced advisor had time for you. This often isn't the case. Of course an up and coming advisor may be extremely busy too but they may appreciate you more, this brings me to my question:

2. Why would it be bad that a young advisor wants to use you to further their success? Wouldn't that align their interests with yours well? My concern with the established professor is that you are nothing to them; you're just another plebe. If the young advisor needs you like you need him/her, that should be good. Of course there are drawbacks; the big name always looks good; but I don't quite understand your arguments.

Just to jump back into this...

As PJ said...older profs have a network and reputation. However, many times the field has passed them by. I'd suggest looking at (every prof in your department) the CV of the professor. They say an academic economist does her most important work by 40. There are of course exceptions. Look to see how often he publishes in the last 5 years, where, with who (the more co-authors for an older prof, the less likely they are doing the majority of the work). Look up some of these papers...are they well cited subsequently? Econ is getting so trend driven so figure out what those are and you'll know what sort of research your prof is doing and if what they are doing is relevant.

Fabio at has a bunch of advice (18 posts so far) for graduate school:

"Why would it be bad that a young advisor wants to use you to further their success? Wouldn't that align their interests with yours well? My concern with the established professor is that you are nothing to them; you're just another plebe. If the young advisor needs you like you need him/her, that should be good."

Well, if it did align your interests. And I suppose it does to some extent. Otherwise it would totally fall apart. Maybe in theory the established professor has less time and less concern. But I don't really see the correlation in practice. In my real life, I've learned that my older advisor was a mentor, and the newer advisor was someone to try to minimize contact with, while attempting to keep the contact high quality. They are more likely to make you do things that postpone rather than advance graduation. Even on my committee, the older, more established profs are more helpful, in direct correlation to their age and establishment as far as I can tell.

I've seen non-tenureds hold back papers from grad students. I figure they must think they have a reason for this, but it strikes me as insane. They have little regard for their personal lives because your personal life might cost him tenure (in their mind).

Back to the older professors. Over time, I figure they've learned to deal with people. The greatest success factor for a professor has to be the success of their students over time. I think it must be that the ones who make it are partially selected for how well they treat their students. So, good ones are sprinkled throughout the new profs, but on average, harder to find.

I think even what you say that you are "nothing" to established professors can be a positive. You are nothing but a person. With a new professor, you have to measure up to their (often unrealistic) comparison to the ideal student. They seem to be quicker to replace you if you don't meet their standard (whether or not they've learned that their standard is capable of being met by their talent pool). Oh, and when they compare you to successful students of other advisors (e.g. Dr. Soandso's student Johnny already had 3 papers), the advisor has nothing to do with it, just so you know.

I don't think this is all the new profs fault. I think it's the system. In my opinion (and this is sure to raise even more hackles) grad students are sacrificed to the system of weeding out assistant professors. It partly has to be this way. In order to weed out marginal grad students, they have to take you to the margin. But, they seem to make a virtue of necessity.

Now, my opinions come from engineering experience and other departments and schools will differ.

IMO, Seth has posted the best advice on the board, so re-read it.

Undergrads - Take as much math/math-econ/stats-metrics as possible if you want to go to grad school. Hell, be a math major, econ minor. This will help you get admitted to better programs, and much better prepare you to succeed in them, than taking econ field courses.

Grad students
- Talk to all of your professors after class. Find one you like personally and whose research you find interesting. Stick closely to research he/she knows, you'll learn more and are much more likely to write a good paper. You have the rest of your career to branch out, it's too risky now.
- On your committee you will need senior people with clout. Junior faculty are fine to work with, but strong letters/communications from senior people will generally carry more weight. I don't care who your official advisor/committee is, just make sure you have a couple of well-known researchers who know you and your abilities well who will come to bat for you on the market.
- If you're not graduating from a top-25 place, it's critical that you come out with a publication record. Each department to which you apply is sorting through 500-1000 application packages. Your application will probably get 1-2 minutes on the first pass-through, and maybe 25-30% of the applications will make it to a more thorough vetting. You might have written a seminal job market paper (probably not...), but if you're outside of the top-25 and haven't published, few prospective employers will spend enough time with your package to discover it (unless, perhaps, you have a well-known advisor pushing your paper as such).
- Conference visibility and networking is important. If you catch someone's eye, they might be willing to ask their department's search committee to take a serious look at your application. Again, anything to avoid getting dumped after only 1-2 minutes. You just want a fair hearing, but search committees don't have time to give 1000 applicants a fair hearing.

Every success is based on continuous efforts. It is not possible be done over nigh.

Comments for this post are closed