Exactly. In the natural sciences, PhD students are typically required to register for a 'thesis' course, and their lab is billed for the 'tuition' for this cour

Exactly. In the natural sciences, PhD students are typically required to register for a 'thesis' course, and their lab is billed for the 'tuition' for this course. The lab is already providing all the instruction and paying for facilities for research through overhead charges. The tuition is in fact a naked money grab by the universities, out of a lab's personnel budget, on top of what's already budgeted for overhead. If these tuition charges are merely overhead, then universities need to dispense with this accounting fiction and re-negotiate their overhead rates with the granting agency, who may decide they get better value at other institutions. If tuition is really a charge to cover some services a graduate student directly receives from the university, then the waiver is income to the graduate student and ought to be taxed the same way tuition remission is taxed for other university employees.