Is There a More Efficient Way to Get Into a Graduate Economics Program?
Well friends, it’s been awhile.
Since my most recent post here (a roundup of all the research I‘ve done about prerequisites for an economics degree), I decided to start knocking out some of those requirements.
True to form, I made a very aggressive plan to complete all the math I would need to go for a graduate degree in economics at some point in the future. Then, I enrolled in a Calculus 2 course in August (and paid non-degree seeking hourly tuition rates to take said course at a U.S. university… yikes). This weekend, I’m scheduled to take my Calculus 2 midterm.
And I’m rethinking this entire plan.
Because frankly, there has to be a better way.
In the interest of full disclosure, the first version of this post was titled: “What is Wrong with Economics?” While I had fun drafting that manifesto (and may come back to it at some point), I realized while writing it that I need to conduct more research to see if my gut feeling about the problems with the way the graduate schools are deciding who does and doesn’t get to become an economist can be supported by any actual data.
See, I do try to act like a real economist sometimes.
Although it may never be ready for this publication, one good thing did come from working on the draft manifesto: I realized what was bothering me so much about the pile of prerequisites to get into graduate school in economics.
In a word, the pile of prerequisite requirements is inefficient.
Because the top graduate programs in economics seem to prefer (or explicitly require) that PhD candidates prove their aptitude for a PhD in economics indirectly.
And doing things indirectly is inefficient.
Why Be Direct?
The best way to learn anything is to do that exact thing, and the best way to signal aptitude for anything is to do that exact thing successfully.
I’ll leave arguments about whether or not economics is in the process of becoming a branch of mathematical theory for another day (and another post). Regardless of how we may feel about the math in economics, the fact remains that the top graduate programs seem to have universally concluded that the math skills obtained by undergraduate economics majors are insufficient for graduate level economics. The top economics graduate programs have decided to address this math gap by adding so many math classes to their pile of prerequisite requirements that there is virtually no way to get through that much math without majoring in math.
But this begs an obvious question about efficiency.
If we’re worried about candidates having the math skills to make it through graduate level micro and econometrics courses, wouldn’t it be a lot more efficient to just make candidates take a graduate level micro course and a graduate level econometrics course before they apply?
If the idea of going to graduate school in order to prove you can go to graduate school sounds ridiculous, that’s because it is. Every other field of study I’ve researched since starting this project seems willing and able to take undergraduate majors from within their own discipline and fill whatever gaps they may need to fill after accepting students into their graduate programs. Overall, the field of economics doesn’t seem interested in that approach, so we won’t waste our time arguing about whether or not the economics graduate schools should take on the math gap they’ve decided their own undergrads have or if leaving the students responsible for it is justifiable. Ultimately, if the economics programs want potential graduate students to prove they can do graduate level economics before being accepted to graduate school, it’s a lot more efficient to make students directly prove that they can do graduate level economics than hope that they are indirectly proving that by taking a lot of other math and stats courses.
Let’s pause for a few arguments I know I’m setting myself up for at this point.
“The math isn’t just a requirement to prove students can succeed in graduate level economics courses.” “PhD candidates also need math aptitude to complete a dissertation and future papers for journals.” “I used something I learned in Topology or Real Analysis in my work as an economist at some point.”
Fair enough. But frankly, none of these arguments strike me as strong enough to justify the money or time investment required to take formal math courses on every math topic that every economics PhD might possibly find useful at some point in their careers. To be clear, I’m not arguing that economics PhDs won’t need to know any math. I’m arguing that this is 2020, and there are a lot of ways to learn math that are not taking 30 hours of formal math courses through a university (and paying undergraduate university tuition for the privilege of hoping you are learning at least some math you may need at some point).
Let’s Be Direct.
(You Can Have Your Math Cake and Share it Too)
While indirectness is inefficient and inefficiency is annoying to any aspiring economist, the indirect nature of the pile of prerequisites has a more sinister side effect: it deters applicants.
Currently, I’m 2 months into the first math course of 4 (minimum) that I would need to take to apply to graduate school in economics. I work full time, so that’s at least a year of math courses at non-degree seeking undergraduate tuition rates. To be clear, that timeline does not include any of the stats or data science/programming that I would also have to take in order to apply for a pre-doctoral research role. Let’s also throw in the time and money it would take to score well on the GRE, yet another signal of quantitative aptitude required in the pile of prerequisites for graduate school in economics. Without including the 1–2 year predoctoral job that I’m highly unlikely to get even if I am somehow able to figure out how to teach myself the correct type of programming for academic research without ever working in academic research, that is conservatively 2 years of non-economics coursework and $10,000+ in tuition & fees for the privilege of applying to a graduate program in economics.
Does that sound inclusive? Does it encourage applicants from non-traditional backgrounds, like anyone who didn’t go directly into a graduate program after undergrad, to apply? Do we think that pile of prerequisites is attracting many first generation applicants? How about applicants living below the poverty line? How about a wide variety of diverse applicants from groups historically underrepresented in the field of economics?
How to Make the Pile of Prerequisites More Direct
As stated above, taking a graduate micro or econometrics course is more efficient than taking an entire undergraduate math/stats/data science major in addition to your economics major. However, to be clear, forcing students to pay for ad-hoc graduate school courses before being accepted to graduate school isn’t a good option either. It’s difficult to find online graduate programs in economics, much less online programs that accept non-degree seeking students at the graduate level, much less online programs that accept non-degree seeking graduate students that would be recognized/considered adequate by the top graduate schools in economics.
Luckily, there is a better way.
In my opinion, the best way to make the pile of prerequisites more direct is for the graduate programs in economics to step in and take some responsibility for the gaps they want to fill. Instead of requiring hours of formal (read: expensive and slow) undergraduate math courses, the top economics programs could create their own free pre-doctoral online course. This course could cover all the math required to be successful in microeconomics and econometrics at the graduate level. Even if it doesn’t teach each math concept, it could easily tell students specifically what to study, so students could use other free online courses, MOOCs, and resources to teach themselves anything they may be missing. If there needs to be a formal test of this economics math aptitude, the pre-doctoral online course could include one. In this way, the top economics programs could make sure all potential applicants directly learn exactly what the programs want them to learn, and directly prove they know that specific material. Again, the best way to learn anything is to do that exact thing, and the best way to signal aptitude for anything is to do that exact thing successfully.
As a believer in the value of directness, I’m a big fan of pre-doctoral research roles in theory. Although I doubt they would use this term in academia, pre-doctoral research roles remind me of apprenticeships (and apprenticeships are a wonderfully direct way to learn). Unfortunately, some of the indirectness in the pile of prerequisites to apply to graduate school in economics seems to have seeped over into the pre-doc research world. So far, I haven’t found a single pre-doctoral research job description from any of the U.S. schools that does not require knowledge of a programming language. While bootcamps and online courses are readily available, finding programming courses specifically for academic research is significantly more difficult. At the risk of repeating myself, I think a little responsibility on the part of the pre-doc employers would go a long way toward directness and inclusivity. Even if the programs don’t want to create their own programming training, they could simply refer to the online course, MOOC, or bootcamp they want applicants to complete. At the very least, they could narrow down the programming language they use instead of listing 3 or 4 possible languages in each job description, so students would know what to study.
The Challenges of Signaling, Comparison and Incentives
We have so many direct and efficient options to learn the math, stats, and programming languages required to be successful in an economics graduate program in the year 2020 that spending the time and money required for a second (or third) undergraduate major in math is far from the only option. However, there is a challenge with the alternative learning options discussed in this post: how to signal aptitude.
Tomorrow, I could find a multivariable calculus course online and enroll. I could learn the subject completely and spend a fraction of the time and money I’d spend as a university student. However, I wouldn’t be able to signal or prove my aptitude to a graduate school admissions committee with an A on my transcript (their preferred signal of aptitude). I may not have a transcript, and if I did, they may not consider it since it wouldn’t be from a formal undergraduate course.
Whether or not we agree with standardized signaling in general, we can acknowledge that a standardized format for signaling aptitude (grades on a transcript in specific courses, scores on a specific standardized test, etc.) makes applicants easier to compare to each other. However, expanding the accepted signals of aptitude is possible. Within this post, I’ve purposely presented some ideas that would allow top economics programs to retain full control of aptitude signaling while loosening control on the methods of learning. Put simply, the graduate programs could easily create their own standardized pre-test for potential applicants which they would arguably have more control over than their current methods of aptitude signaling. With that standardized signal of direct aptitude, there is no longer any need to police how students learn the material to pass the test: if I can pass after taking a free online multivariable calculus course, that should be all that matters.
Unfortunately, should is the key word in my previous sentence. Because despite logical arguments in favor of efficiency in both learning methods and aptitude signaling, we have a potentially insurmountable challenge on the road to directness. The top graduate schools in economics have no incentive to do anything differently.
After all, if you have hundreds of applicants for each spot in your program, and they all comply with your existing pile of prerequisites — indirect and inefficient as that pile may be — why would you change the pile to potentially increase the size of the applicant pool?
Could we incentivize the top economics graduate programs to make their pile of prerequisites more direct and efficient?
If so, how?