Overhead Costs -- A Twisted Tale of Graduate Student FundingBy Jonathan Richmond
The recent controversy about the way MIT bills graduate student research assistant tuition might be worthwhile if it opens up discussion about the way graduate student research is funded in general. There are some gross inequities which rarely enter the limelight, and it is time to see them eliminated.
First, for all those confused about what gets paid where, is an attempt to explain a formula which would be at home in a Kafka novel. In the next issue of The Tech, the implications of this system for MIT's least fortunate graduate students will be detailed.
Virtually all research funding which enters MIT is charged a 57.5 percent overhead. Salary-related expenditures are also charged 41.4 percent for employee benefits. When employee benefits are charged, the 57.5 percent overhead is also levied on top of the employee benefits charge. For those who already have a headache, here is the net effect on salaries in simple terms:
<*p(0,12,0,8,4,0,g,"U.S. English")*t(156,2," ",0,0," "):
z8f"FranklinGothic-Book">Salary payment $100.00
MIT overhead @ 57.5 % 81.31
Total Research Sponsor Cost $222.71
<*p(0,12,0,10,0,0,G,"U.S. English")*t(0,0," "):
z9f"TimesNewRomanPS">In short, to pay $100 in MIT salaries costs a research sponsor $222.71 -- an additional 123 percent over the actual salary -- whether the employee is a graduate student or another staff member. Certain research centers levy an additional overhead charge which stays in the center. MIT overhead is, in that case, charged on top of the center's local overhead, while employee benefits are charged on the salary-related component of the local overhead (while MIT overhead is additionally added on the employee benefits charge).
At a center which charges 5 percent salary local overhead and 1.25 percent non-salary local overhead, it costs a research sponsor $236.77 to pay $100 in salaries.
MIT overhead also applies to a surprisingly wide range of non-salary expenses. If you charge an air ticket to Washington, D.C., for example, to explain to Rep. John D. Dingell (R-Mich.) how reasonable MIT's overhead charges really are, MIT will add a 57.5 percent overhead charge to the fare. That's more than five times as much commission as the travel agent makes for actually issuing the ticket.
It's an expensive proposition to buy research at MIT whether an additional local overhead is charged on salaries or not. For the sake of the rest of this column, let's consider a graduate student research assistant working for a department without local overhead. Got some more aspirin handy?
While the same overheads are charged for both graduate students and other employees, the benefits paid are very different. While most MIT employees receive subsidized health and life insurance and other typical employment benefits, graduate students pay for the full cost of hospital insurance, but get their tuition paid out of the money collected for "employee benefits."This is what the numbers look like for a graduate student RA in the School of Engineering over a nine-month academic year.
> Masters Ph.D.
Employee benefits @ 41.4% 4117 4546
Sub-total 14,062 15,525
MIT overhead @ 57.5% 8086 8927
Total Sponsor Cost $22,148 $24,453
The total remuneration received by students is actually greater than the amount charged.
Tuition 16,900 16,900
Total remuneration 26,845 27,880
Here's how the full cost of tuition is paid. While approximately $4,100 or $4,500 is paid by the research sponsors into the employee benefits pool for masters and doctoral students, respectively, $16,900 in tuition payments is withdrawn from this same pool for each of them. The resulting deficit is taken from the employee benefit money paid into the pool for non-student employees. MIT deliberately set the employee benefit levy for all employees at a rate high enough to cover these tuition disbursements, because research clients had balked at the cost of paying the full costs of tuition for graduate students on projects they supported. By spreading the tuition charge across all employees as a general employee benefit charge, MIT could disguise the fact that tuition was being charged against all research project employees. The Institute could not only hide the high actual cost of tuition, but also the fact that much of the money supposedly taken in to provide health and other benefits for the bulk of MIT employees was not being used for their benefit, but instead being transferred to the tuition account.
At the same time, the graduate students who get their tuition paid in this way do not get their hospital insurance subsidized or receive the other benefits which research sponsors can rightly expect are being provided when they hand over payments for "employee benefits."
Let's face it: MIT has been following a highly dubious accounting practice. Investigation into this matter is highly appropriate. There are implications deeper than those yet explored, however, especially for students who do not have assistantship appointments. And, if you will kindly take enough aspirin in the interim, these will be probed in the next issue of The Tech.