so what would the new incentives be? don't publish... and get paid anyway?

At least that's how traditional Western universities used to work. The profs sometimes published something, and they taught students. Looking at science back then, it seemed to work out rather well.

If possible, this would be ideal.