Before I read the article, my first reaction was "People. Research grants pay for people('s salaries)." (And other things as well, but that's usually the biggest category)
But apparently, in the author's university, a massive amount of goes to "overhead". That's quite surprising, in my opinion.
One thing that really annoys me as a grad student is the fact that the University charges tuition to my advisor even if I'm not taking any classes. Most PhD students finish their course work pretty quickly and don't take any classes after that and yet they get charged for tuition from the grant. The justification is that they're enrolled in thesis credits which is also a class but that's stupid imo, especially when the grad students don't get paid much. It's even more annoying when people say "oh your pay as grad student is low because technically you are getting free tuition.". Yeah no, I'm getting free tuition for like a year or two I take classes while I'm working there for 5 years.
I also wish the funding agencies allowed more flexibility is terms of spending excess money in different categories. I was surprised to know that left over money from a grant that's supposed to pay for an an undegrad cannot be used to buy a faster computer for research related to the same project or maybe even pay for data labeling services like scale.AI. They wanted us to hire a much more expensive undegrad instead of using external data labeling services at much cheaper prices.
Oh and also Fly America Act. Every federal grant requires you to use US airlines and I sometimes paid 2x what I would have paid otherwise (and only one of the author could do instead of 2 because of limited money).
Yea it's BS. I'm getting charged tuition (which is waived) while getting paid a 50% FTE contract as a graduate researcher yet I spend 100%+ time in the lab on research. So they're saying that they're charging me for that other 50%+ time I guess? Sure, they can call it dissertation work, some of it isn't directly on the grant I'm under but that's still messed up IMO. Some other students in my lab are there every weekend.
Waived tuition for PhD students is such BS and a clear tax loop hole that universities can increase their costs artificially (they get to pick tuition). I always told my peers that universities would treat PhD students very differently if they actually paid the full payment to the student via 1099 who could then claim the cost as a business expense.
The 50% FTE I think is heavily intertwined by the fact that international F1 students are not allowed to work more than 20 hours. You could have a 2 tier system for domestic vs. international students but that would cause a clear schism in the student population. Regardless, there is the very real problem that a huge part of PhD is indeed training. In my department, it was very real that your "non-training" hours looked something like 0/25/50/75/100 across the years. Basically, nothing in the first 2 years, taking courses and maybe helping a bit on the side then finishing the last year writing and publishing a lot of stuff and training the new PhDs. It averages to 50% and I think is generally fair from that regard that the university did not hire you as a 100% researcher out of the gate.
I really think that after the first two/three years there should be a system converting PhD students to an official researcher title, which is what they essentially are. Credit hours for thesis research, most of which is publishable and usually for some grant, is just BS, and even worse it helps justify the fact that one must work 20 hours per week while spending extra unpaid time for fulfilling the thesis research credit.
Both the university where I got my PhD and the university where I'm at right now have mechanisms to reduce your tuition to very, very near 0 (usually whatever is the minimum number of credits) once you pass your prelims.
From what I heard, Columbia will be even higher than Harvard. The cost of a PhD student is ~100k/yr (salary + tuition + overhead). Postdocs are more expensive.
Overhead is basically rent. You might think that a university rolling in with donations and names on buildings wouldn't need to charge rent but:
1. Nobody donates for upkeep. Not a huge market for the "jdiez17 memorial janitorial closet" if you know what i mean.
2. Endowment spending is typically limited to 4 percent, and often less. And often restricted in what it can be spent on.
Administrative overhead basically allows the university to convert its fixed assets into flexible spending on their basic needs that donations won't cover.
Also: overhead is limited by law, unless your university has a grandfathered exception.
Right, there are a lot of administrators, janitors, technical support staff etc being paid by the university. But I'm surprised they aren't paid by the students tuition fees, since are so ridiculously high already.
There a too many administrators imo, who are often paid quite a bit. I don't think a University needs so many associate deans, assistant deans, and directors. Unfortunately it's people in those roles that make future hiring decisions so minimizing spending there isn't usually their goal.
Student tuition is high but not a high part of the budget. 40 years ago tuition was mostly covered by the state and tuition charged to students was basically just a copay to ensure you weren't terrible. But since the end of the cold war there's basically been a one way ratchet: every time a recession hits the states cut uni funding to balance the budget, the unis raise tuition to compensate, and when the recession is over this now the new normal.
The immediate question that would come from that is why is a technician or grant admin paid off tuition, when a student will never benefit from that?
That also doesn't account for things that don't "feel" like your typical university but are major research centers, like medical schools, where a lot of faculty are purely research.
But apparently, in the author's university, a massive amount of goes to "overhead". That's quite surprising, in my opinion.