@SteveBellovin@infosec.exchange
Post #3629151
2026-07-06 14:01 UTC
@mattblaze@federate.social Yup. Let me put this in quantitative terms. As far as I know, in all US research universities every CS PhD student is fully funded. This might be by a student's own fellowship, from their employer or home country, but considerably more often it's via research grants. The funding has to cover all of tuition plus a stipend to cover (subsistence-level) living expenses, plus necessary research travel and equipment, e.g., laptops. (There's also "overhead", a separate complex topic I won't go into.) At Columbia, where I taught before I retired, that meant that every PhD student required $80-100K per year. A typical faculty member had a group of about 5 PhD students. This meant that we had to bring in an average of $400-500K per year in new research funding, every year, plus more if you needed specialized lab equipment or money for cloud services or the like. And I was not allowed to admit PhD students unless I could show that I had money in hand for their first several years.
Replies (1)
-
@mattblaze@federate.social 2026-07-06 14:17
@SteveBellovin@infosec.exchange Right. When I admit a PhD student, I'm basically committing to raising ~400K to fund them through completion, and some of that funding has to be obtained a few years AFTER they've arrived (because grants are typically ~3 years but a PhD generally takes 4-5 years to finish). And this isn't counting specific research costs. Heavily experimental PhDs often cost more. So our ability to admit students depends not just on current funding, but also on optimism about future funding.