The most controversial part of the plan is giving bonuses to teachers for 'raising children’s achievement'. Because the labor market in education is so highly regulated by union contracts school principals have little experience or expertise in evaluating the job performance of teachers. So teachers rightly distrust any measure which gives principals power over their pay or promotion; in the foreseeable future such a plan is doomed to failure.From what I gather from Denver (er, sort of the hometown these days), which is running a four-year pilot program of this sort, teachers set their own goals and standards, and get paid depending on whether or not they meet those goals. I don't really know how this gets regulated, but if teachers had incentive (or could be forced) to set high standards, this could very well work. Chattanooga, meanwhile, runs its pay-for-performance program by charting improvements in student scores over the course of a year. As the man says, the absurdity here should be apparent. On the other hand, both Denver and Chattanooga have seen decent improvements in their school systems thus far, so perhaps "fairness" isn't the only thing to consider.
But any scheme which bases merit pay on objective standards, like measured improvement in students’ scores, would be absurd. Children improve at uneven rates; their improvement is always due to the interaction of many factors of which any given teacher is only one; and no teacher should teach enough students to generate a sample size big enough to give statistically significant differences between teacher performances.