Once, at 3Com, Bob Metcalfe introduced a talk by one of his MIT professors with the little joke, "The reason academic politics is so vicious is that nothing's at stake."

The guy said, "That depends on whether you consider reputation 'nothing.' "

I guess what that shows is, you can always negotiate and compromise over money, but reputation is more of a binary. An academic can fake some work, and as long as he's never called on it, his reputation is set.

So yeah, a little more fear of having one's reputation ruined would go a long way towards fixing science.