This might get interesting as grad students in science are there to be trained. This involves research, and in lieu of paying tuition a research or teaching as

This might get interesting as grad students in science are there to be trained. This involves research, and in lieu of paying tuition a research or teaching assistantship might be provided. A few students bring their own funds, but I never did anything with that good a school/prof. Yeah, I handled recitations, graded exams or proctored labs. But my major purpose was to complete the requirements of getting accredited in the PhD program (courses, cumulatives/proposals and then a prelim). That took, on average 2.5 years. From there, 1-2 years of research that (hopefully) generated the data/results to get out. Throw in some odds and ends that the prof felt should be added and that meant that writing up started in the 4th or 5th year. More if it involved lab work and less if it didn't. Each chapter of the dissertation was a paper, roughly, and if the papers were written before graduation then "stapling them together" made the dissertation easier to pull off. Otherwise, writing the 100-300 page document took 3-6 months and a fair bit of red ink. Once that was done, defend successfully and head off to the job or postdoc previously arranged. Getting the dissertation -> papers was critical in getting the cv started. Yeah, I did things in lieu of paying tuition which support my time 4 years and summers. This made me an employee that worked, say, 10 hours a week to justify this support. Way less than I spent doing my research, library and computer work. If I had heartburn with such support/work, I could have always supported myself paying 3 semester hours 3x yearly and feeding and housing myself. I was a student, after all, and the ability of a grad student truly helping anything but an assistant prof is quite limited. So, be careful what you ask for... If grad school is to be viewed as a job it's going to take longer than if it was an avocation. In synthetic chemistry, for example, running 1 or 2 reactions a day (including setup and workup) has a huge impact on the duration of one's research. Fortunately, laptops run unattended so theoretical/computational types have a staggering advantage over lab denizens. And if the departments feel too pressured, they can always divorce the TA's from the grad students and deal directly with adjuncts. Far cleaner process and no overhead. RA's is harder, but I'm sure any mid-senior prof would get more work done with post-docs rather than 2-3rd year grad students. See Dyson's comments on Youtube vs. his distaste for the PhD process and how the Institute at Princeton didn't muck about with them. Grad school, for my me and my peers, was the start of the process not significant in and of itself. I would be concerned that the more "complete" grad school becomes the shorter one's career might be (assuming a constant 40 year window of opportunity). If you have golden hands or are lucky/excellent, you will do well no matter what system is used to handle grad students. If, however, you're more typical or might have second thoughts... hope the process is as tolerant of such students as possible.