Applying to economics Ph.D. programs

So you want to be a grad student? Well, think twice. It’s not easy to be a grad student, and it’s not easy to be a professor, either. I tell my students that it’s not for everyone. There are already plenty of advices on applying to economics Ph.D. programs out there (see here or here), so I will limit this post to my opinion. Since part of my research involves dynamic programming, let’s discuss using backward induction.

After getting a Ph.D.

Since a Ph.D. is a big-time commitment, make sure you know what happens after getting a Ph.D. Typical jobs are

1. Professor at research-oriented universities,
2. Professor at teaching-oriented universities,
3. Industry (e.g., finance, tech, consulting, etc.),
4. Government (e.g., central bank, IMF, etc.)

I have little insight into the trade-offs of these jobs, as I have spent my entire professional career as a research professor. The advantage of being a research professor is that some of us get paid reasonably well (see my analysis here), and we have plenty of free time to do research, which for me is leisure. I am assuming that teaching professors are paid less well but have plenty of free time to do whatever they like with their lives. Industry jobs pay the most; I’ve heard of former students getting $300k starting salaries, but you will have to work like a bee.

The vast majority of Ph.D. students end up not being a research professor, just like most college tennis players don’t turn pro. This is obvious by doing simple accounting. Assume each cohort of the Ph.D. program consists of 20 students. Assume each department hires two people per year. Then only 10% of graduates can get research-oriented academic jobs. The truth is that it is even harder because research-oriented academic positions are competitive, and you need to compete against students from top schools, as well as advanced assistant professors who are not going to get tenure from top schools, and many of them are very smart. Hence, unless you think you are very smart, and even if you are, keep the expectations low. But honestly, academia has its own downside, such as internal politics and bureaucracy, so industry jobs are very attractive. No matter where you end up, you will have a good life.

Getting a Ph.D.

The first year consists of core courses, including mathematics for economics, microeconomics, macroeconomics, and econometrics. The second year consists of taking electives based on your interests. The rest is spent on working on the dissertation. Fumio Hayashi, who taught me macro and econometrics at the University of Tokyo, told us: “Study 10 hours a day, every day. When I was at Harvard, I studied 80 hours a week.” I think this is typical. Successful Ph.D. students are driven by strong internal motivation (whether it is wealth, fame, or scientific discovery) and have a strong work ethic. If you don’t like hard work, a Ph.D. is not for you.

Getting into a Ph.D. program

If you would like to be admitted to a good Ph.D. program, consider how admissions decisions are made. I have served on the admissions committee for many years. In a typical year, I evaluate about 200 files. Due to time constraints, I spend only 5–10 minutes on each file and cannot review the entire package. Here is the process I (not necessarily others) follow:

1. Transcript, especially mathematics. I browse the transcript and check whether the applicant has taken some advanced mathematics courses (real analysis is the bare minimum) and whether they are all A or close to it. Modern economics is highly mathematical, and experience tells us that students with insufficient mathematical preparation struggle. I don’t pay much attention to GPA, as it may have been inflated with ping-pong, dance, social justice, etc. At this point, all I care about are good grades in hard sciences like mathematics, physics, statistics, and (quantitative) economics.
2. Writing sample. I check whether a writing sample is attached. If a writing sample is not attached, I take it as a signal of a lack of ambition or creativity. If a writing sample is attached, I skim it. A research paper reveals a lot of information. What I mean is that attaching a research paper as a writing sample to your application can increase or decrease the likelihood of acceptance. When I see a sloppy term paper that just runs regressions, I would downgrade my evaluation. On the other hand, if I see something solid, I would upgrade my evaluation.
3. Letter of recommendation. I feel that letters of recommendation are becoming less informative because the content is inflated. What I would like to see are short but concrete letters that mention, for instance,
   • this student took such-and-such class (some course description) in 20XX and ranked \(n\) among \(N\) students,
   • among all students I have interacted with, this student compares to so-and-so who graduated in 20XX and went to such-and-such program,
   • This student came to my office hour, corrected an error in the lecture, generalized a model, etc.

If I don’t see this kind of concrete information in the letter, I will just ignore it.

Purpose of Ph.D. program

If you want to apply to a Ph.D. program, you should understand its purpose from the perspective of the provider. The purpose of Ph.D. programs, as with any other educational programs, such as undergraduate, M.A., medical, or law schools, is to train students. But there is a big difference between Ph.D. programs and other programs. With other programs like undergraduate, M.A., medical, or law schools, they charge tuition. Nobody says so (at least in public), but the real purpose of these programs is to make money by offering academic training. This is just common sense, as well as how capitalism works. However, Ph.D. programs usually offer tuition waivers and stipends. Stipends are often around $40k per year, and tuition is around $80k, compared to private undergraduate institutions. Therefore, Ph.D. programs lose more than $100k per year for each Ph.D. student enrolled. Why on earth do Ph.D. programs do such stupid things? It’s because the purpose of Ph.D. programs is to place Ph.D. graduates as researchers at prestigious research institutions, thereby increase their own prestige, hire better researchers, recruit better students, and ultimately make more profit through tuition and donation. Based on this logic, Ph.D. programs are interested only in prospective students who are likely to become creative researchers. Applicants should keep this in mind.

Some red flags and other considerations

Here are a few red flags that you would like to avoid when applying to Ph.D. programs.

1. Low undergraduate GPA. This signals that you are not serious. That said, there are exceptions. For instance, my undergraduate GPA was horrible, but I was still admitted to Yale. This is because my B’s and C’s were limited to medicine, and I had perfect grades in mathematics, physics, and statistics, and I took lots of these classes.
2. Too many years have passed since graduation. This signals that you are not committed. That said, we sometimes admit applicants who graduated long ago, but usually only when the subsequent experience is highly relevant.
3. Too many master’s degrees. One M.A. is okay. But multiple M.A. signals that you have been rejected by Ph.D. programs before.
4. Absence of writing sample. If you don’t submit a writing sample, I will take as evidence that you are not creative or passionate enough about research.
5. Spending too much effort on statement of purpose. We know people lie. Their actions, not words, reveal the truth. If you have time to polish your statement of purpose, use it instead to work on the writing sample.

Here are some other considerations.

• Letters need to be (mostly) written by economics professors (to be credible). This means that you need to take some economics courses. If you are not an economics major, one option is to pursue a master’s program first.
• GRE is used only for screening. A near-perfect score in the GRE quantitative is necessary but otherwise irrelevant once you pass screening.
• Statement of Purpose and Diversity Statement are mostly irrelevant. If you have time writing these, study more mathematics instead.
• Programming skills become important once you get into grad school. Learn programming languages at an early stage.