Wednesday, June 18, 2008

How to get into graduate school in economics

As most of my friends already know (and as should be obvious from my network change on Facebook), I am going to the University of Chicago this September, to do my PhD in economics.

Since several of my friends are economics students, I wanted to dispense what wisdom I have about the application process in one sitting, for ease of reference.

My opinion (rather than the wisdom of internet message boards) is this: there are two reasons why you might do a masters degree in economics: to get a fancy job in London (or Johannesburg, or New York, whatever) and earn enough money to bathe in; or, so that you can get into a fancy PhD program in the U.S. and become the next Joseph Stiglitz or Gary Becker.

For private-sector job purposes, you would do well to go to LSE or Oxford (not too hard to get into from UCT). Business science economics will do fine for that, assuming you maintain a 70+ average. It's not necessary to do all that much math or statistics if this is your goal.

For PhD admission, you need a little more. There are two kinds of courses that count in your favour: math (calculus up to linear algebra is the minimum requirement, and real analysis very good to have), and postgraduate econ. Everything else on your transcript will be ignored. Remember, the admissions committee won't know which classes are which, and which ones are the hard ones, so you can mention that in your SoP, as long as it doesn't become a facsimile of your grades. Here, I'm focusing on American, or good Canadian or British PhDs (e.g. UBC, U Western Ontario, Queen's University, U Toronto, LSE, Oxford, or UCL), since the process is similar for most of them. (And because most economists worth reading have done their PhDs in such places.)

Although UCT has sent people from business science straight to PhD, you could do up to MAM300W to give yourself a bit of an edge. (The material becomes very useful in micro theory and econometrics later on.)

Roughly, you need the following to put together a PhD application:
  1. GRE scores. Only take the general test. This costs around $160, and you can book an appointment online at www.gre.org. There are three sections: analytical writing (two essays - one argumentative and one where you analyse a given argument), verbal (vocabulary, grammar, comprehension) and quantitative (high-school level algebra, geometry, numeracy). For economics, the only thing that counts is your score on the quantitative section, which is (bizzarely) scored on a scale from 200-800. If you want to get into a top-50 US programme, you need to crack 770 on this section, and 800 for any school ranked above 25.
  2. Three letters of recommendation. This is, by far, the most important part of your application.
  3. A personal statement of purpose. This should answer the questions "Why do I want a PhD?" "Why should School X choose me?" "What makes me qualified to study this?" etc.
  4. (optional) A diversity statement.
  5. (optional) A writing sample.
From reading endless discussions online, the consensus on how decisions are made is something like this:
  • The admissions committee gets the 800 - 1600 (in the case of MIT) applications sometime in January. They check to see if you have a GRE Q score of >780. If not, your application goes in the recycling bin.
  • Assuming you get through, they check your letters of reccomendation to see if they know and trust your letter writers, and to see if your letter-writers haven't said anything bad about you.
  • Remember, everyone's letter says that student X is hardworking and smart. That's essentially noise.
  • The way you can deal with this is to get your reccomender to write about specific research projects that you have done for her ("Under my direction, Jesse collected data from a field experiement in which subjects played three rounds of the dictator game. Jesse helped clean the data and performed some initial econometric analysis for me...")
  • If you get through (and by then, it's down to about 200 people) they read through your transcripts, see if you've taken real analysis and maybe even topology or measure theory, etc. Here is where your grades and math background can help you.
  • After that, they might read your personal statement, reread your letters, etc, trying to determine if you are a good fit for their program. For example, if you tell UChicago that you're a heterodox economist who wants to break the back of neoliberal capitalism, wave your $90 application fee goodbye.
What I've said here is not really a complete guide to the process, which does require quite a lot of preparation and planning. (and money!) So here are some places you should definitely go on your quest for further information.
  • The best place to look for advice is the TestMagic Forum. This is a discussion board dedicated to economics PhD or masters applicants, and many past applicants (current students) continue to post there. Most of the questions you have will already have been answered, so search the forum or check the permanently posted topics (near the top of the page).
  • A good (but slightly outdated) page is econphd.net. If you want to see what the material in (the first year of) econ grad school is like, check out the extensive collection of lecture notes. If you want a step-by-step guide through the admissions process, read the admissions guide. If you want to see where you should shoot for in your favourite subfield (beyond the obvious - yes, we all know Harvard and Berkeley are amazing at everything, but you can discover some interesting anomalies here: Caltech is rubbish at macro, Northwestern kills most of the Ivy league at IO and Oxford is not as good as you might think...), browse the rankings. (Slightly newer rankings can be found here [scroll down], though access is limited to the top 10).

Thursday, June 5, 2008

My (all-soy) beef with econometrics (plus: a rambling argument for fieldwork in economics)

Only my economist friends will understand this angst, but this has really been bugging me for a while.

How can econometrics adjudicate between different data-generating processes?

We have lots of techniques for estimating coefficients - OLS, 2SLS, GMM, systems estimation, but that only picks out the model that appears to fit the data best from a restricted class of possible DGPs. And the class of DGPs - to noneconomists, this is the "mechanism" that "generates the data," i.e. a probabilistic model that we suppose (hope?) characterises the real world - is simply assumed a priori, often chosen for convenience, not for plausibility on economic grounds.

I find it troubling that empirical work hardly pays any attention to the theoretical implications of the functional forms it uses. If we estimate a linear wage equation, what does that implicitly say about firms' marginal revenue product curves (if we want to believe that labour markets are competitive)? If we presume semilogarithmic household demand functions, what are we saying about intra-household bargaining and the utility function of the household (or, by extension, its constituents)?

I know there are lots of nonparametric estimation techniques out there, which I guess I feel are more honest. Obviously, I'm aware that they're way more data-intensive than parametric techniques - the curse of dimensionality and all that.

I wonder, though - and Rob, Sean, Alex, Dean (sorry about the technical terms) and Simon, help me out here, since you're really the intended audience for this post - if there shouldn't be more to the scientific process (as it relates to economics, at least) than the estimation of conditional mean functions? OK, so we get the partial derivatives of some function which, as far as we can tell, represents reality. Those are our "effects," our betas. We've dealt with selection bias, attrition, and endogeneity.

Now what? Interesting though it is, can you think of anything else that we as economists should do to explore the validity of our ideas? If we were molecular biologists or solid-state physicists, we could go screw around in a lab and hope to notice related things that might give us clues about our theories. But how can economists do anything of that nature? We come to the data with an idea in our heads that we want to test. But our ability to let the real world speak to us in return - to poke around and see something that we weren't quite looking for - is severely limited. This is the anxiety that makes me so enthusiastic about fieldwork - actually going out there to root around in the dirt to see what's going on out there.

My research interests

I dislike having to explain that despite the graduate degree in economics, I'm never going to be rich, I don't know which stocks are hot, and that no, I don't have a prediction for GDP growth (or inflation or interest rates). For example:
Me: Uh, (shuffles feet), I'm doing my PhD in economics.

New Acquaintence: Oh, wow! So you're going to be so rich one day.

Me: (uncomfortable laughter) Well, maybe. Probably not though.

NA: So, what's gonna happen with (some macroeconomic issue that nobody can predict)?

Me: I don't really know. I'm more interested in applied microeconomics - like, have you read "Freakanomics"? That's the kind...

NA (interrupting): You know, I have a friend who studied economics. But he thought it was too theoretical. Now he has this consulting business, and he's making millions. That's REAL economics.

Me: Uh-huh, I guess so.
I was indulging in a little metacognition (well, more like mental housekeeping) the other day, and I realised that there are basically two themes in the research questions that interest me: sex and violence. I'm really curious about crime: why don't more people steal things? I'm often surprised (and grateful) that I haven't been attacked in my house yet. After all, we suburbanites do have a lot of nice stuff, and there are a lot of people who are really poor. Why haven't the suburbs been swamped by ambitious thieves? Sure, the rich can invest in high walls and alarms, but those aren't too hard to get over or around.

On a similar topic: according to burglars and thieves, they frequently use the proceeds of their crimes to impress women. If it were sex they were after, why not simply rape the objects of their desire? It would probably be easier, both physically and logistically, and these men clearly don't have a problem with using violence to achieve their ends. Perhaps they fear retaliation and ostracism within their own communities. Then again, it's not hard to disguise yourself, and there's a reasonable chance that the victim would keep quiet about it.

Wednesday, June 4, 2008

Oh wow, this thing is still here.

I didn't really have the time or conviction to keep it up during math honours last year. But now I have all the time in the world, at least until September, when, as most likely everyone reading this knows, I will be starting my PhD in economics at the University of Chicago.

I think perhaps the reason I don't write is because I haven't really done any research, unlike the other members of the econblogosphere. I prefer to believe this than the alternative, which is that I don't have any original thoughts.