Over at Crookedtimber, Harry Brighouse poses a very good question.
My own policy is to disclose very little, particularly with undergraduate students. The professor is not supposed to dictate positions to his or her students — I want my students to come to their own conclusions (of course they have to know the material for the exams but that’s a separate issue). In the graduate courses I disclose more. I think (hope) graduate students are less likely to passively accept the instructors views as the final word on a topic. Also, grad students are interested in what faculty think about the probable future direction of research.
For the record, when I discus behavioral economics, I usually say that I have something of a “bias” against behavioral though I emphasize that my views are in the minority. I’m not sure whether my students feel that my treatment of behavioral is fair (and balanced?) or not.