For example, my university had a very extensive dining system, all within a block of town, which was full of restaurants. The dining halls were more expensive than the restaurants. But everyone had to buy a super-expensive meal plan, to prop up this ridiculous side-business of the university.
You'd probably be right up to about 80%. Then all of the edge cases come in and it falls down, having a real impact on students. If you want to operate a university on the Pareto Principle, then fair enough, but there's a whole lot that goes on that requires tweaks and human consideration which breaks fundamental systems, usually when students need them the most.
Or why did the university need a whole administrative department to cater to the “needs” of every kind of minority you can think of (one department per kind of minority)?
Which cohort foots the bill for the cost of the radical transformation projects? The severance packages, the business process mapping, the specification creation, the tender process, the development of hundreds of interlinking system integrations, training, maintenance, licensing costs etc...
Splitting wasteful salaries across thousands of students over a number of years and trying to improve piece by piece and not cause a local issue (whilst increasing people power temporarily in other areas where needed which aren't be focused on) is the only viable solution we have at the moment.
Believe me, I'd 100% love to hear solutions to this problem that take all the moving pieces into account to solve it, and it'd have been great not being saddled with £00,000s of student debt too!
But, us techies can sit on HN and bemoan the inefficiency of little pieces of people-powered work and how it should all be automated! efficient! etc... till we're blue in the face. But without building a university, healthcare system, multi-national conglomerate... from scratch, we can't just retrofit our napkin systems and processes without causing a shittonne of unthought impacts.