[Kinrany]

Administration has no clue which projects are good and which "scientists" won't spend the money on a sea of booze.

Two mechanisms should be used:

1. Administration decides which projects had real world effect, years or even decades after the work is done.

2. A directed social graph is built from citations and from working together on projects.

Money should be allocated to the network. Administration decides who did good work in the past; funding is allocated to everyone endorsed by those people; scientists vote with their grant money for projects that they believe in.

[bluenose69]

I've never known an administrator who had even the faintest clue what science was all about, let alone what previous was 'good' or what projects would have 'real world effect'. If they had insights on such things, they would likely be working in science, not administration.

Indeed the mess we are in is because of administrators who think organizing academics is like organizing work on an assembly line. They find something quantifiable and insist on more of that. Papers are quantifiable, so they go for that. To try to get at quality, they count citations. Always it's counting, and I'm not sure if that's by analogy to the assembly or because the administrators are challenged to do much more than count.

[mach1ne]

I think this would lead to a hellpit of sycophantic behavior. And, genius or not, I would not trust any group of people to be immune to the influence of the sycophants.

[Kinrany]

Would that be worse than the current disaster though?

[mach1ne]

Not necessarily. Still, a drastic change asks for strong evidence of improvement.

[derbOac]

Honestly your proposal isn't too different from what I've seen going on in many university administrations. My initial reaction is that it is the current disaster.

[derbOac]

I'll just say that whatever it is, it has to be corruption-proof. And in the current environment, I regularly see gaming the sorts of things you're suggesting.