my bizarre market ideas

Teacher evaluations in New York will finally be allowed to be based partially on results, which is a step in a positive direction, if not without any possible flaws. One fear I've heard expressed is that grading teachers based on student performance will make teachers all the more averse to taking on difficult students, and I certainly think that students' incoming performance should be taken into account as well. If it isn't, using exit test scores becomes more arbitrary than if it is, and if teachers can maneuver to select their students, that obviously creates an incentive problem. One of the ideas I've had floating in my head is not to directly take account of students' previous performance, and also to allow teachers (or schools) to select their students, but to do so by bidding on them. Each teacher/school sets, for each prospective student, a standard for success, above which the teacher/school is rewarded in some sense and below which the teacher/school is punished in some sense, and the student goes to the teacher/school willing to bid the highest level of performance for that student.

I could actually defend this against some of the first objections that occur to me, and I'm not averse to doing so, but one of the points of this particular blog is an emphasis on brainstorming, and I have other things I want to do with my time right now.