Which city makes a PhD stipend go furthest? The answer is not on any department website, and that is not an accident. Departments love the nominal number. The nominal number is their number. The adjusted number is yours.
Here is the ranking nobody publishes but everybody needs: PhD stipends adjusted for what life actually costs in each city. Take the stipend, subtract realistic local housing costs, and look at what is left. The nominal ranking and the adjusted ranking look nothing alike. That gap is where bad decisions live.
A top-tier stipend in a coastal metro can leave less disposable income than a mid-tier stipend in a college town. Nobody mentions this at the visit weekend, when they are showing you the nice part of town and the free lunch. A ten percent stipend bump does not cover a housing market that costs twice as much. It just does not. Do the subtraction yourself and trust the arithmetic over the assurances.
The expensive-city premium is real but rarely complete. Some programs in pricey cities do pay meaningfully more, and credit where it is due. But many pay a little more and call it handled. Handled is doing a lot of work in that sentence.
College towns are the arbitrage play of PhD admissions. Lower stipends, dramatically lower costs, often the same quality of training. If your field does not require a coastal location, the financial case for a strong program in an affordable city is overwhelming. I mean genuinely hard to argue with.
Run any offer through a cost of living adjusted stipend calculator, whether it is a SaaS tool or a spreadsheet you built yourself. But remember it is only as honest as its inputs. You want a cost of living adjusted PhD stipend figure that starts from what your actual program pays, not a national average wearing your city's name. The adjusted figures across 798 programs at phdstipend.fyi will surprise you. They might change where you apply, which is exactly what honest numbers are for.