One institution, measured the same way
971 matched computing doctorates from a single university, every figure carrying a national column measured the same way. The destinations are unremarkable. The dissertations are not, and the state keeps six times its share.
Everything on the previous four pages is national. This one asks the narrower question that a department actually wants answered: where do our own graduates go, and is it different from everywhere else? The short answer for Illinois is that the destinations are unremarkable and the dissertations are not.
1,804 computing doctoral dissertations were filed at Illinois between 2000 and 2022, and 971 of them linked one-to-one to a career profile. Every table below carries a national column measured the same way, because a single institution’s percentage on its own means nothing.
The window stops at 2022 because the national dissertation database stops receiving theses from this university after that. Part 08 rebuilds the last decade from the university’s own repository and carries the recent cohorts.
The academic-placement ranking is not a prestige ranking
The chart below is the argument against reading it as one. The three highest academic shares belong to Louisiana State University (31.6%), Indiana University (30.0%) and the University of Pennsylvania (29.9%). Illinois is at 23.9%, and the departments a reader would expect to be far above it are not: Princeton 25.5%, Washington 24.3%, Cornell 22.6%, Berkeley 22.5%. A high academic share mostly means graduates had fewer competing industry offers, or that the programme feeds teaching institutions. It describes the market a cohort met, not how well it was trained.
Which department the degree came from
The largest source of computing doctorates at Illinois is not the computer science department. Electrical & Computer Engineering accounts for 291 matched doctorates since 2011, Computer Science for 241, and they place differently: 23% versus 20% into academia. Any statement about “computing PhDs at Illinois” is a statement about at least two departments facing different labour markets.
From 2011 onward, which is when the dissertation records for this institution start carrying a department field.
| Degree-granting department | Matched doctorates | Into academia | Into industry |
|---|---|---|---|
| Electrical & Computer Engineering | 291 | 23% | 54% |
| Computer Science | 241 | 20% | 56% |
| Mechanical Sci & Engineering | 14 | — | — |
| Industrial&Enterprise Sys Eng | 13 | — | — |
| Civil & Environmental Eng | 12 | — | — |
| Library & Information Science | 12 | — | — |
| Information Sciences | 11 | — | — |
What Illinois doctorates write about
Where Illinois differs is upstream of placement, in what its doctorates write about. Against the national mix it runs heavy in languages and software engineering (11.4% against 6.4%), systems and distributed computing (8.8% against 6.0%), and scientific computing (5.5% against 3.1%), and light in machine learning as a subject, 5.5% against 8.6%, a little over half the national share.
That last one invites the obvious objection: perhaps Illinois machine-learning work is simply filed under vision or NLP instead. Partly, and not mostly. Taking the three together Illinois is at 20.7% against 22.6% nationally, so the gap narrows from 3.2 points to 1.9 and does not close.
Ten years later
Ten years out the two curves are still on top of each other: 21.1% of Illinois doctorates are in academia against 21.6% nationally. The industry side drifts apart slightly, 53.8% against 50.4%, which is the destination table read one decade later rather than one year.
The state keeps about one in six
33% of Illinois computing doctorates are in California a year out, against 32% nationally, which is no different from everyone else. The home state is where it diverges: 18.1% stay in Illinois, against 3.0% of US computing doctorates overall, roughly six times the national rate.
Measured the same way as everywhere else, Illinois destinations sit in the middle of the national distribution, at one year and at ten. What is distinctive is upstream: a dissertation mix tilted toward systems, languages and scientific computing and away from machine learning as a subject, and a home state that keeps six times the national share of its computing doctorates.