While I agree with the common wisdom that giving student opinion weight in tenure and pay can lead to professorial pandering (in the form of better grades, lighter coursework, or easier tests), I also am of the opinion that student feedback can be useful to both the professor and future students.
Obviously, future students can benefit from a sort of database of reviews that indicate whether lecture and reading materials reflect examination questions (sadly not always the case), or whether the professor is an engaging lecturer. Over the course of three post-secondary degrees, I’ve had a number of courses that could be quite satisfactorily passed without opening the $150 book (written by the professor), just by attending the lectures. I’ve also had courses where the lecture adds nothing to the readings of a conscientious student. I feel that either the cost savings or time savings should be fairly passed on to the students.
Likewise, the professor might benefit from knowing that, to take one example, reading text directly from prepared slides projected during the class is a futile gesture (and a depressingly common one). Further, the tip that an accent is impenetrable might lead to more usefully writing upon the blackboard. There are any number of tidbits that may be useful for the conscientious lecturer.
Administrators can learn that a professor is habitually ten or 15 minutes tardy to his own sections, or has a habit of holding students long enough past the class (making them chronically tardy to other courses), or has other disturbing habits that might require institutional remedy.
My suggestion, then, is to add more work to the professors (or, more likely, their overworked TAs). I would propose that students be given a random number at the beginning of their studies at a given institution.
They would then submit their coursework and examinations under this number rather than under their names. This number would only be translatable to an individual student at the provost’s office, although perhaps student photo ID could have a picture/number match on one side of the card and a picture/name match on the other to verify the appropriate student is sitting for examinations. Then, when course evaluations are submitted, the evaluations could be compared against the grades assigned in an attempt to weed out legitimate criticism from complaining over marks.
I’m sure there are a number of holes in this method, but given the benefits that motivated students, faculty, and administrations can reap from evaluations, it might be worth our while to decouple them from grade inflation. I hope this is at least a useful start on the project.