Kudos to Illinois.
To follow onto Larry's post, everyone knows who the good and bad teachers are at a given school. The kids know, the parents know, the faculty know, the admin knows. Everyone knows. The problem is trying to come up with a measuring tool to objectively show what everyone already knows.
I think test scores are important, but certainly not the most important factor (if the second graders were at 25% and then the third graders moved up to 45% there was some good teaching going on but the percentage still sucked). I think there needs to be a combination of things.
Each year I would have a report given to the school board on each teacher. The report would include test scores, anonymous ratings (scale of 1-10 how would you rate this teacher on attutude, work ethic, dependability, craetiviyt ect) from school admin, fellow teachers and parents. Larry's input would be interesting, but I would guess this practice would weed out the bad and identify the good pretty quickly.
I am not for mass firings either. I would take the above sampling and tell each teacher that is in the bottom 10% they have a year to show improvement or face firing. I'd take the top 10% and either have a bonus structure or some other type of perk (these things don't have to be huge- give them preferred parking spots or an extra $1,000 for their classroom etc).
Larry- Would you agree that at your school you pretty much know who the good ones are and who the bad ones are? My guess is that most are good and there are a few bad ones that are either burned out or simply dont have the skills necessary to be competent.