The goal of any pay system should be to pay someone for the value that he or she creates for the customer. In the case of a school, the teacher should be paid for the value created for the student (and therefore the taxpayers).
Now, there are two really difficult questions to answer. The first is how do we measure value for the taxpayer and the student. Unfortunately, we have chosen to measure this primarily by test scores. There have been many arguments made against the use of test scores as a valid measure of successful education. Some have argued that the measures are too narrow. Some have argued that it's easy to cheat. Some have argued that it leads to an education of memorization and not of real thinking. I believe it is also flawed because there are so many other characteristics of a successful person than just what is measured by academics. A school that develops academically accomplished criminals is not a good school. Likewise, a teacher who develops academically accomplished criminals is not a good teacher.
The second is what makes a good teacher. You can probably guess that I don't think that a good teacher is simply a teacher who develops students who attain good test scores. Moreover, I do not believe that a teacher has enough control over student test scores to be judged primarily (or even partially) on his or her students' test scores.
While largely agreeing with the new education reform package in Tennessee, union President Earl Wimon said, "Teachers do not control all the factors that go into a successful test score. We don't think it should play an overarching factor.”
Wiman also said "that the union is still considering whether to push for language to be inserted into the legislation capping the weight of test scores in evaluations at 35 percent. Others have suggested test scores should account for 50 percent of a teacher’s evaluation, a level Wiman called 'immoral.'"
I don't know that I'd call it immoral. I'd use the word foolish (that's being polite). Although it probably is immoral in the sense that it's unfair. Often morality and reasonable go together.
The reason that it's unfair is that teacher's don't control enough of the factors that lead to high test scores for them to be judged on the scores, especially not when (as some in Tennessee are after) the percent of evaluation is 50%.
If school leaders (and teachers themselves) do not understand enough about what makes a great or even a good teacher, then they ought to find someone who does. There are plenty of studies about what makes good teaching. There are good rubrics for evaluating teacher skill, knowledge, ability to engage students, and other characteristics that can be observed that teachers can be rewarded on those characteristics. If those teachers who are determined to be good according to these rubrics are not effective in bringing up test scores over time, then it becomes pretty obvious that the rubric isn't right or that there is something in their student population that is wrong.
I realize that this takes more thought than developing some matrix that says that if a teacher's students achieve a certain level of test scores, then he or she receives a specified bonus or raise. The point is that if we are going to commit to a strategic pay system, then let's make it really strategic. Let's use the factors that affect value created. Let's use characteristics such as student engagement, affect on overall school morale, affect on student morale, self-motivation for professional improvement, use of techniques to invite student participation, and, yes, even things such as comparative marketability of the teacher's subject. Until we abandon a system that pays based on things that don't improve teaching, such as longevity and advanced degrees, we'll continue to get what we pay for. Strategic pay systems don't pay for student performance. They pay for teacher performance. The two may be related, but they aren't synonymous.