VAMs are an issue for the reasons mentioned in the article, but just because a teacher reliably produces the highest performing students does not mean s/he is a good teacher.
There was a teacher who was the only teacher of highest track Algebra II class at a local HS who had such a terrible reputation among the students that some would drop down one track in math to avoid having her. Numerous students (and their parents) complained to the administration and the official reply was: "Our top math students all came out of her class" which was a rather specious argument since all of the top math students also went into her class.
Unofficially she was close to retirement age, had seniority in the department, and nobody wanted to poke the beehive of forcefully reassigning her classes.
Shouldn't, on the other hand, the fact that "all the top math students also went into her class" be counted as positive for her? I mean, based upon your description, those who attended her classes excelled. That is good.
So I guess parents didn't want to put their children into her classes because that wouldn't have worked out; but maybe the reason for that is that their children were simply unfit for the level?
> the only teacher of highest track Algebra II class at a local HS
Being the only teacher of the highest track Algebra II class in the school means that anyone who wanted to take the highest track Algebra II class--which plausibly contains many of the best math students--would have to take it from this teacher.
The whole point of the kind of value added modelling that is the centre of this case is that it attempts to factor out things like the quality of the student to estimate the quality of the teacher, precisely so that bad teachers who by dint of circumstances are associated with high-performing students don't get high ratings.
The problem is that if student background counts for the greater part of performance even a good teacher may have difficulty scoring highly if they happen to get a "good" class (one that scores highly on student quality.)
On a larger scale, the anecdote we are discussing here suggests that teachers as individuals may not make that much difference to student's performance, since this bad teacher was still able to turn out the best performing students thanks (one is supposed to presume) selection effects alone.
In our school district, it appeared that teaching assignments were largely based on seniority and intra-district politics. It seemed as though the senior teachers most established in the hierarchy would try to get the classes and programs where the best students would be told to go, so as to seem more effective and have more funding.
Thus, in a program that was supposed to cater to the best students in the school district, we had an English teacher who was senile to the point that she could not keep track of assignments or grading or have a coherent curriculum, a History/Social Studies teacher who was primarily interested in pursuing mandatory, irrelevant projects, like expensive theatrical productions, and a Math teacher who didn't teach at all, and just had us go over homework each day.
We were top students not because of our teachers, but because we were all motivated, and were hand-picked for the classes based on test scores and prior performance. We were the children of involved parents, and a high percentage of them were professors. Students who didn't perform well could be easily thrown out by the teachers, as well. With these advantages, it was a given that their students would excel beyond other students in the district; what was not clear was whether we were actually learning as effectively as similar students elsewhere. We went there because it's where the school district told us to go, and told us we'd be given the best opportunities, but the result was primarily to make ineffective teachers look very effective.
In the end, a number of us left the entire district en masse when we realized that the system was entirely ineffective, and not to our benefit. I think that the aftermath showed just how ineffective they were. In three years, I went from being a middle school student who was not being taught at a significantly higher level than other students, and was struggling, to being a junior at a first-tier university who was excelling and had significantly better grades. At the same time the students from our former middle school were graduating high school, I was a first-year grad student. Two of my friends went similar routes, with similar experiences.
If you grab all the top students, and can throw out any student who doesn't perform well, you're obviously going to look very effective, even if you don't teach well and your students could be doing far, far better. Top students will continue to come to your classes, because you'll appear to be the most effective, while better teachers will not have a chance to succeed, because they won't have your numbers to draw interest from parents, or the ability to game them by hand-selecting students.
There was a teacher who was the only teacher of highest track Algebra II class at a local HS who had such a terrible reputation among the students that some would drop down one track in math to avoid having her. Numerous students (and their parents) complained to the administration and the official reply was: "Our top math students all came out of her class" which was a rather specious argument since all of the top math students also went into her class.
Unofficially she was close to retirement age, had seniority in the department, and nobody wanted to poke the beehive of forcefully reassigning her classes.