Going from Liberal Arts College to Econ PhD

[Unequalparts]

Hi Guys,

 

So I am a rising senior at a top 20 national Liberal Arts college this semester. I am interested in becoming a PhD-trained economist. I won't go too into depth about myself in this thread, since I want it to be a bit more general. For background, I have a solid GPA (3.9 in Econ including A's in intermediate Micro, intermediate Macro, and econometrics), faculty-mentored research that I am submitting for publication (in an undergrad journal obviously), and have a 3.6 GPA in the math department, having taken two proof-based courses so far. I am taking analysis next Fall.

 

With that said, I feel like even doing pretty well at a difficult liberal arts school, it is going to be difficult for me to get accepted to top-30 programs without doing a Master's program in between somewhere, which I would prefer not to do.

My research was not with a world-renowned economist, so I feel like even if I get a very nice letter of rec, it won't hold too much weight at the top programs. I don't have the opportunity to take graduate level courses to distinguish myself, and ultimately, I feel like I don't have the resources to distinguish myself in the way that students at large uni's with great Econ programs have.

 

But I guess my question is a more general one: What should a student like me at a small liberal arts school do? Are there many examples of economists who come from small schools and go straight to doctorate level programs? Or should I just think about aiming for a good European/Canadian masters program or finding a good research position after graduating?

 

General or specific advice is appreciated!

[Food4Thought]

It seems that students from liberal arts colleges who are gunning for top programs usually apply to prestigious research assistant programs instead of a doing Master's. Working as an RA will allow you to gain access to top researchers that will garner well-respected letters of rec. Try taking a look at career opportunities sections at each of the Fed banks and at NBER (Research Assistant Positions not at the NBER). Given you are at an elite liberal arts college, I'm sure your school is either well connected at a nearby Fed or a nearby major research university.

 

You'll want to check these websites around November; there will be a lot of postings, and you should be ready to submit an application.

[chateauheart]

Agreed with above. Economics is one of the disciplines where it's hard to go straight from LAC to PhD, essentially because LACs don't offer a lot of research opportunities for economics undergrads. This is in stark contrast with most other disciplines, where LACs have very good placement into PhDs.

 

The exceptions are the top 5 or so LACs.

[Zubrus]

Agreed with both above. This doesn't say much but most of the LAC students that were at top 10 visit days were from Swarthmore, Oberlin, Amherst, Williams, etc. (Only the big names). I agree that going the RA route is probably the best.

[Unequalparts]

Thanks for the responses guys.

A few followup questions:

Are those research positions extremely competitive?

Should I take the GRE this year anyways?

Should I bother applying to any feeder Master's programs?

 

Thanks!

[Food4Thought]

My thoughts:

 

1) They are extremely competitive but less competitive than a top ranked PhD program. But if you were to get one, the grad school ranking where you'll be competitive at will increase significantly. We're talking the difference between top 40-50 and top-20, or even top-10 (assuming grades are strong)

 

2) If you feel like you can get a good score if you take it now, then do it. It will feel a lot better to know that you don't need to worry about it when you finally apply. Plus, it gives you plenty of time to retake it if you don't do as well as you hope.

 

3) It depends. Since those RA positions are very competitive, you may want to apply to a few Master's programs. If I were in your positiom, however, I would spend most of my effort on the RA, but also send out a few Master's apps. More options are better, so long as you dont mind paying the application fees.

[Unequalparts]

My thoughts:

 

1) They are extremely competitive but less competitive than a top ranked PhD program. But if you were to get one, the grad school ranking where you'll be competitive at will increase significantly. We're talking the difference between top 40-50 and top-20, or even top-10 (assuming grades are strong)

 

2) If you feel like you can get a good score if you take it now, then do it. It will feel a lot better to know that you don't need to worry about it when you finally apply. Plus, it gives you plenty of time to retake it if you don't do as well as you hope.

 

3) It depends. Since those RA positions are very competitive, you may want to apply to a few Master's programs. If I were in your positiom, however, I would spend most of my effort on the RA, but also send out a few Master's apps. More options are better, so long as you dont mind paying the application fees.

 

Thanks!

A final couple of questions

On #1, do you think a 3.7 overall GPA is good enough to be competitive at these positions if my Econ grades are especially good and I have a pretty decent math background (I will have Calc sequence, Proof-based Linear Algebra, Diff Eq., Combinatorics, and RA by graduation)?

 

On #2, does it matter at all if you take the GRE a couple of years before you apply for PhD programs?

 

Thanks again for the feedback.

[Food4Thought]

I cant really give a yes or no regarding if you are competitive. This summer, I will begin an RA at a top-10 school. Send me a PM and I can share with you my experience with applying for them.

 

How much programming experience do you have? I would say that you should probably have a course in a some programming language like Java, C, or Python. It doesnt need to be super advanced, because most economic research isnt incredibly complicated from a programming standpoint. Your RA experience should give you familiarity with a statistics package like Stata. These professors want someone who will be able to quickly learn how to do things; this ability usually comes from prior programming courses. Do you know what kind of research you'll be doing with your professor. Data entry and other medial tasks wont help very much. On the other hand, if you are actually assisting in statistical analysis, etc., then this will be very appealing to professors.

 

As for the GRE concern, I would not worry about that. There is a little bit of a misconception on this forum that admissions committees go through your application with a fine tooth comb. That is not what happens. They look at what important classes you took and how well you did in them. They read your letters of recommendation and see what you got on the GRE. They have too many applications to look at each one with that much scrutiny.

Edited May 22, 2015 by Food4Thought

[Food4Thought]

To think of it in a different way, if you took real analysis during your freshman year, do you think admissions committees would view that as a bad sign? After all, taking RA earlier means you have more time to forget the material, and it might not reflect how you would do in RA if you took it later.

 

Sounds silly.

[Unequalparts]

Thanks for the response. I'll keep that stuff in mind. I might PM you later to hear your experience on getting an RA.

 

I have familiarity programming and doing data analysis in Stata, although I don't have any programming courses. I know a little python, and I am teaching myself Visual Basic this summer so I can do cool stuff in Excel 8D

Maybe I'll try and fit in a CS course.

[thagzone]

Thanks for the response. I'll keep that stuff in mind. I might PM you later to hear your experience on getting an RA.

 

I have familiarity programming and doing data analysis in Stata, although I don't have any programming courses. I know a little python, and I am teaching myself Visual Basic this summer so I can do cool stuff in Excel 8D

Maybe I'll try and fit in a CS course.

 

Just a suggestion: don't waste your time learning Visual Basic. It's not that VBA can't be useful, but you don't really want to do cool things in Excel--you want to get your data into a statistical programming environment as quickly as possible and do cool stuff with it there. Spend that time learning R (or more about statistical/data manipulation in Python) instead. Anything you learn in VBA you'd be much better off learning how to do in a statistical programming language.

[PhDPlease]

I don't think taking the GRE early is any sort of problem as long as your scores haven't expired (I think 5 years?). Personally my scores were several years old.

 

Assuming you've engaged in other quantitative pursuits (courses, RA work, anything else) since you took the GRE, your quantitative abilities would only have improved. I think adcoms would only be concerned that your skills might have gotten rusty if you spent a number of years doing things entirely unrelated to econ/math/quant.

[PhDPlease]

Just a suggestion: don't waste your time learning Visual Basic. It's not that VBA can't be useful, but you don't really want to do cool things in Excel--you want to get your data into a statistical programming environment as quickly as possible and do cool stuff with it there. Spend that time learning R (or more about statistical/data manipulation in Python) instead. Anything you learn in VBA you'd be much better off learning how to do in a statistical programming language.

 

I agree. I did a short course on VBA, but everything covered in the course seemed to be things I'd do in a statistical programming language so I haven't used VBA for anything. I mean it certainly isn't harmful if it is something you enjoy, but in terms of skills that will be most valuable, I would say VBA isn't the best use of time.

[thagzone]

I agree. I did a short course on VBA, but everything covered in the course seemed to be things I'd do in a statistical programming language so I haven't used VBA for anything. I mean it certainly isn't harmful if it is something you enjoy, but in terms of skills that will be most valuable, I would say VBA isn't the best use of time.

 

This is right on. I've only ever had to write a couple scripts in VBA, and they were expressly for the purpose of cleaning a boatload of disgusting individual spreadsheets to the point that I could load them into R to do the real work. Once you gain some competency in a powerful language like R, manipulating your data with VBA feels like trying to drive a car with square wheels. If you ever need to mess around in VBA, your knowledge of other programming languages will make it simple enough to code what you need to and get out of there (this is my experience, anyway).

[Unequalparts]

Alright, thanks for that advice guys.

I already program in Stata. Do you think R or Python would be a better next language to put my energy into? I've learned some Python in the past, but I wasn't really focused on data manipulation/statistics. If anyone knows any good tutorials that are specifically suited for that purpose, I would appreciate the suggestions.

[thagzone]

Alright, thanks for that advice guys.

I already program in Stata. Do you think R or Python would be a better next language to put my energy into? I've learned some Python in the past, but I wasn't really focused on data manipulation/statistics. If anyone knows any good tutorials that are specifically suited for that purpose, I would appreciate the suggestions.

 

My suggestion is that you learn R. I would recommend to someone wanting to learn programming in general that they deepen their knowledge of Python, but for your purposes, R is a superior tool. The reason for this is that R has far more robust offerings when it comes to statistical packages, it is unambiguously superior for generating data visualizations, and it has much better community support for the aforementioned features.

 

Python is certainly becoming a very viable tool for statistical programming and data science, but it is largely used for these purposes in industry applications. Python is a multi-purpose language, and so its flexibility allows developers to build statistical tools using Python and easily mesh them with a larger codebase on the back end of some polished product. However, in social sciences research, you will rarely find projects which are conducted from start to finish in Python--this is R's realm.

 

While your Stata skills are going to come in handy because many researchers and industry folk still like it, learning a "real" programming language will suit you very well. Stata is hand-holding for people that have data, are beyond point-and-click Excel manipulation, but still don't quite know what they're doing. In addition to its structural differences from a true programming language, because Stata is not open source, it is full of limitations. Try building a web application in Stata that allows other people to interact with a user interface to visualize and explore your data in real-time. Of course, you can't do that. In R, what you can do with your data is boundless (including doing just what I said). Stata is useful, but your options are quite bounded. It's a launching point, but it shouldn't be something to hang your hat on.

 

Working with data sets of significant size (I'm hesitant to call any economic data set I've seen "big" at this point) is going to be an increasingly important skill. For that reason, I highly suggest that you start learning how to manipulate relational database objects in R within the framework of data.table and NOT the base R object known as a data.frame. If you're interested in more detail about how I would recommend developing some of these programming skills (instead of whatever time you planned to spend on VBA), you can PM me and I'd be happy to elaborate.