I've been a public school teacher for 17 years and, like Gov. Arnold Schwarzenegger, I've grappled with the idea of merit pay. It would be terrific if we could reward good teachers and weed out the bad ones. The trouble is there has not yet been devised an effective means for judging good or bad.
You certainly can't rely on the scores of standardized tests because those correlate more with ZIP Codes and parent involvement than teacher competence. Even the best teachers aren't going to raise scores in schools where students don't care and parents aren't around.
In addition, such a measurement standard would compel teachers and students to spend even more time engaged in mind-numbing test preparation.
What about giving the principal more power to reward and replace teachers?
Well, then you're putting teachers in a position where they have to please "the boss." What if the principal doesn't approve of a certain teacher's idiosyncratic teaching style? Or what if a teacher speaks out about the principal's management skills? Couldn't the decision to reward or replace a teacher become personal and/or political? A principal's opinion can certainly be part of an overall evaluation, but a principal should never become a personnel officer.
What about peer review? It would be as imperfect in the teaching profession as it is in the medical or legal fields, where incompetent doctors and lawyers continue to practice because their mistakes are not made public and their colleagues are unwilling to speak out against them.
Also, if the present collegial campus environment were replaced by a competitive, business-like model, then teachers would begin to view colleagues as competitors and would be less willing to collaborate or share innovative ideas.
The best judges of a teacher's competence are usually the students, but do we want to enter a realm in which a bunch of second-graders decide whether or not to give their teacher a bonus? I don't think so.
So where does that leave us? Is there any legitimate way to evaluate the effectiveness of a teacher? In my opinion there is.
Here's my method: You ask a teacher at the beginning of the school year what he plans to teach his students. For instance, I might say I plan to teach, among other things, long division, ratios, adjectives and adverbs, writing a persuasive essay, the preamble to the Constitution and the Lewis and Clark expedition. Then a specific test would be designed (by the teacher, the principal and the test makers) to cover those areas. Which means that tests would vary from class to class and school to school.
In addition, the same test would be given twice, once in September (so you have a baseline score) and again in June. If the scores go up, then the teacher has delivered on his promised curriculum and the students have absorbed and retained the information. A formula can be devised to turn point gains into dollars. Such a strategy would not only help teachers refine their skills, it wouldn't pit teachers against each other, it wouldn't compromise teacher creativity and it wouldn't lower morale in low-performing schools because expectations would be based on real-world variables.
To come up with a fair merit pay formula, you must begin by including teachers and by taking into account those real-world variables that big-city teachers face every day.