The looming college-enrollment death spiral

> Just pointing out that nonprofits do not have to balance revenues and expenditures at all. This is not a mere technical detail. Profit cannot inure to owners with nonprofits as it can with for-profits, but this does not prevent the organization from building up a surplus over time, nor does it prevent them from paying employees very handsomely. Otherwise it wouldn't be possible to have almost one hundred universities with endowments over a billion dollars.

I'm not an expert in non-profit accounting, but my general understanding is that it's not generally allowed to divert non-profit revenue into reserves at scale. The billions of endowments all come from non-budgeted and one-time gifts to the university.

> That's fair, but I suppose there's more to it than that. Any number of datasets point to the cost of admin rising far above the cost of faculty or maintenance; and a lot of them actually show on an inflation-adjusted basis that schools are spending less now on instruction than they did a decade or two ago. So perhaps there's enough truth to the idea that you're getting substantially less education per dollar than you used to.

All the above statements are true, but they all rely on the same error. Money spent by a university on research administration, or healthcare administration has zero relevance to the amount of money spent on Instructional administration. In order to make the argument that the cost of admin has gone up to the detriment of instruction, you'd want to exclude the administrative staff dedicated to non-instructional tasks.

As an example, if UCLA opens a new hospital in SoCal (not near any current UCLA undergraduates), they must staff that hospital with thousands of medical professionals. Sure, you can say the doctors are faculty, but you are still hiring thousands of nurses, technicians, janitors, cafeteria workers, etc. That makes the administrative cost of the University go up, but the Instruction is exactly the same.

Similarly, if UCLA gets a $100 million dollar grant for graduate-level AI research, they're going to have to hire quite a few staff members to administer that grant, to pay the researchers and grad students who are funded by that grant, to purchase and manage the computers full of GPUs, to build the building for the people and the datacenters for the servers, and to file the mountains of audit paperwork that the federal government requires in order to get the money and show that there wasn't any Waste, Fraud, or Abuse. All of the staff (AKA administrators) that make that happen now show up as an administrative cost of the University, and end up in a naive calculation of "increases in administration at UCLA", despite the fact that all of their salaries are paid (legitimately) by the $100 million grant that UCLA received, and instruction is completely unaffected by this.

This is why I like the expense data from the Education Department. It at least tries to break out the expenses related to non-Instructional things (Research, Hospitals, etc.) There may be an interesting discussion to be had as to whether Universities are spending more on Instruction or Student Services per-head in real dollars, but I pretty much never see intelligent discussion rooted in those numbers; instead it's the common, "Harvard has 4x the administrative staff today that it did in 1980. Outrageous!"

> I cannot accurately base those decisions on a methodology that assumes a cost of zero ("no direct schooling costs").

The study measures benefit of education. If you wish to determine whether cost>benefit, join it with research on the cost of education.