I learn from Ezra Klein that Texas A&M will be giving out $10,000 bonuses to faculty members who score high in student evaluations. This strikes me as an astonishingly bad idea. Robert Farley has a number of important criticisms of Klein's post, but the nub of the problem is that student evaluations are only modestly correlated with teaching ability. I don't think they're irrelevant -- evaluations have helped me spot problems in my courses and correct them rather than repeat them. And those who receive high evals are probably, on balance, better educators than those who receive low evals. But there are a huge number of things that affect evals -- including the professor's personality, looks, political leanings, sexual preference, gender -- that have nothing to do with teaching ability.
More importantly, the A&M proposal just creates an incentive to rig the class to provide better evals. This can be done quite easily. You can give out food on evaluation day. You can give higher (undeserved) grades. Hell, for a shot at $10,000, it might even make sense to pay students for high evals. Yes, as Ezra notes, professors might try doing things like teaching better, but generally speaking, it's far easier to do other, less beneficial things to juice the evals.
A&M might be better off using this money to train their professors in modern pedagogy, or to hire professional teaching coaches who observe professors in the classroom, which almost never happens.