This all ignores a basic financial reality underlying 'hard-money' academic positions: we are paid 9-month salaries that come from core funds (ie: teaching, pai

This all ignores a basic financial reality underlying 'hard-money' academic positions: we are paid 9-month salaries that come from core funds (ie: teaching, paid for by tuition, government funds, etc.). Yes, bringing grants matters, but given that grant $ have not really changed in decades, it is a huge relief to not have to fund one's own salary (beyond 3 summer months). During dry spells, students can be supported on TAships or fellowships here and there, and if you relatively low-cost experiments, small internal grants can tide you by. This is the real reason you'd have trouble with this model. You'd then need grant $ to directly pay 100% of faculty salaries. Unless you can solve this problem I don't see scientists leaving universities. The big exception is of course medical schools/ medical research, where faculty positions are largely soft money (some minimal base salary is usually covered).