On the Economics of Teaching with Graduate Students

Views differ on whether using graduate students as instructors is cost effective for undergraduate teaching. The case for is that research departments will have vigorous doctoral programs. That is their core business. Doctoral students require funding. That is a fixed cost associated with the program. So in essence the graduate student labor is free (though it is recognized that if they didn’t teach they could take more course work). The case against is that it is a way to subjugate the grad students to slave-like conditions, retard their progress toward degree and impede their growth as researchers, all in the name of providing surplus for the research faculty in the department who prefer to teach grad students and supervise the occasional dissertation.

Whether one view is right versus the other is not intrinsic to doctoral education but rather depends on how the doctoral program is run. So instead of resolving that issue here, I’m going to content myself with a simpler question. If there is doctoral program and the grad students serve as TAs in undergraduate instruction, overall does the program break even, generate a surplus, or require a subsidy to operate?

I’m going to work through the arithmetic to get a feel for how to answer that question. But before I do, let’s make the following observation. In the calculation yesterday I was implicitly assuming a three credit hour course taught solely by the instructor. When grad students serve as TAs they frequently do so by running a recitation section that is one hour a week. Frequently they will teach several recitation sections. Let’s say they do three of these. If the average size of these sections is 25 students, then in the language we used yesterday, because the contribution margins were per course we should think of the TAs as teaching 25 student in total, not 75 students. Yesterday credit hours were in the background. Today we need to keep track of them a little more carefully.

The basic question being asked is now straightforward to consider. Graduate students teaching undergrads, narrowly defined, is a surplus generating activity. Research faculty teaching grad students is a subsidy requiring activity (recall the tuition and fee waivers for grad students). What is the net overall?

I’m going to do the calculation in the context of a department where revenue is tuition financed. I’ll use convenient numbers so the arithmetic is simple but they won’t differ too much from reality.

Let’s suppose the grad students get half-time assistantships that pay $20,000 per year and that the teaching load in the department is that a TA gets three recitation sections per semester (as I suggested above). From the calculation yesterday, the break even number of students per year is 50. This means section size of 25 students per semester. If sections are actually of size 30, then under the same assumptions as in the post yesterday, the TA generates a surplus of $4,000. If section size is 35, the surplus is $8,000. (Graduate students who don’t pass the TOEFL are typically assigned as graders. The calculation for them is a little different but not in a fundamental sense, so I won’t do that arithmetic here.)

So if a graduate cohort has 20 students it will generate a surplus of between $80,000 and $160,000 depending on actual section size averaging between 30 student and 35 students.

These graduate students take courses, probably three per semester, maybe two if they are a little further on. If the standard faculty teaching load in the department is four courses per year, then under yesterday’s assumptions, the opportunity cost of teaching a grad course is $25,000. So if that cohort of 20 grad students takes their courses in lock step in a section of size 20, the overall program generates a surplus. If they take a disparate set of courses that vary in size, with some courses say having only 5 students, then it is possible that the program will require a subsidy to be flush.

Certainly in Economics and I suspect in many other graduate programs as well, there is a core set of courses that all students must take. By the above, the core is probably surplus generating and the surplus is dissipated as the students get into their field courses. Rules on how frequently graduate courses are offered and minimal class size for them to be offered are determined by considerations similar to the ones outlined in this post. Naturally, assistantships vary by discipline as does average section size. But the nature of these calculations is the same.