Every institution can tell you how much federal money it received. Very few lead with which agency wrote the check, and that is the part that determines what happens to the institution when policy moves.
Across 2,886 institutions with federal award records, the government obligated $63.39B. The median school's share of that is $4.24M. But the median is the least interesting number here. The composition is the story.
One number hides the story
Three research universities, three completely different institutions, and you cannot tell them apart from a single funding total.
MIT's federal money is a Defense story, because Lincoln Laboratory is a federally funded research and development center. Stanford's is split between Energy and Health, because SLAC is a national accelerator laboratory sitting next to a medical school. Johns Hopkins is overwhelmingly Health and Human Services, with a defence tail from the Applied Physics Laboratory.
These are not three points on the same scale. They are three different businesses with three different political exposures, and a funding total of "about a billion and a half" describes none of it.
A funding total tells you the size of the check. The composition tells you who can cancel it.
What concentration looks like
The useful question is not how much federal money an institution takes. It is how many independent decisions that money depends on. An institution drawing from six agencies has six separate budget cycles to survive. One drawing from a single agency has one.
Only 347 institutions draw meaningfully on five or more agencies. That is the genuinely diversified set, and it is small.
Most schools are an Education Department story
For 2,646 of 2,886 institutions, the largest federal funder is the Department of Education. That is not research money. It is Pell, and it is loan volume, which means it is really enrollment money arriving with a federal logo on it.
This matters because the two kinds of federal exposure fail in opposite directions. Research money is vulnerable to appropriations and agency priorities, and it moves slowly. Student-aid money is vulnerable to enrollment, and it moves as fast as your admitted class does. A school whose federal line is 100% Education does not have a research funding risk. It has an enrollment risk that happens to be denominated in federal dollars.
The single-funder problem
1,604 institutions receive 100% of their federal obligations from a single agency. For almost all of them that agency is Education. There is no diversification to fall back on, and no second relationship to lean on when the first one changes.
None of this is hidden. It is filed, every year, in public. It is simply never assembled per institution, which is the only form in which it means anything.
How to use this
- Know your own composition before a board meeting asks. "We receive $40M in federal funding" invites a follow-up you should not have to improvise: from whom, and what happens if that agency's budget moves 10%?
- Stop benchmarking against institutions with a different funding shape. A regional public whose federal line is entirely student aid is not competing with an FFRDC host, whatever the totals look like side by side.
- If your federal line is mostly Education, treat enrollment strategy as federal-revenue strategy. They are the same line item.
A note on method: a university owns a set of registration IDs, not one. MIT's Lincoln Laboratory files under its own identifier, and dropping it because the school "looks ambiguous" would erase most of MIT's federal funding. We sum every identifier we can prove belongs to the institution, which is safe because no identifier can be claimed by two schools.
The bottom line
The size of the check is the least interesting fact about it. Who signs it, and how many other people could have, is the number that predicts what your next five years look like.