I think that's simply hard to do. What is student success? Test scores? Teachers will teach to the test or even help students cheat on standardized tests if their pay is at stake.
If the evaluation is known ahead of time, we'll just see Goodhart's Law in action ("When a measure becomes a target, it ceases to be a good measure" https://en.wikipedia.org/wiki/Goodhart's_law).
Not only that, but a huge part of "student success" isn't based on the teacher's performance. For example, if you give me a class of wealthy students whose parents have graduate degrees, it'll be a lot easier to teach them than a classroom of low-income students without access to resources who aren't set up for success.
In fact, school administrators could tip the scales toward teachers they like and retaliate against those they don't like even with "objective" measures. Let's say that we have a perfect test of student advancement - a perfect measure, can't be taught to, and can't be cheated on. I dislike you as a teacher and load up your class with all the known disruptive students. You spend your time on classroom management (controlling student outbursts, keeping students in their seats, etc) rather than teaching. This perfect, objective measure comes around and it turns out you're a terrible teacher! Your students aren't doing better than before your class. Except the fault of that is mine - I manufactured a classroom that was doomed from the start.
That's not to say these things don't happen in industry as well. I've seen managers give people they didn't like projects they knew would fail or gave them a team of under-performers to drag them down. However, it's kinda worse when kid's futures are in the balance.
Yes, there are certainly better and worse teachers, but it can be hard to determine who is doing well.
> However, it's kinda worse when kid's futures are in the balance.
Considering the alternative is paying people simply based on how many years based on seniority alone, we should endeavor to use a system that at least attempts to take performance into account.
If the evaluation is known ahead of time, we'll just see Goodhart's Law in action ("When a measure becomes a target, it ceases to be a good measure" https://en.wikipedia.org/wiki/Goodhart's_law).
Not only that, but a huge part of "student success" isn't based on the teacher's performance. For example, if you give me a class of wealthy students whose parents have graduate degrees, it'll be a lot easier to teach them than a classroom of low-income students without access to resources who aren't set up for success.
In fact, school administrators could tip the scales toward teachers they like and retaliate against those they don't like even with "objective" measures. Let's say that we have a perfect test of student advancement - a perfect measure, can't be taught to, and can't be cheated on. I dislike you as a teacher and load up your class with all the known disruptive students. You spend your time on classroom management (controlling student outbursts, keeping students in their seats, etc) rather than teaching. This perfect, objective measure comes around and it turns out you're a terrible teacher! Your students aren't doing better than before your class. Except the fault of that is mine - I manufactured a classroom that was doomed from the start.
That's not to say these things don't happen in industry as well. I've seen managers give people they didn't like projects they knew would fail or gave them a team of under-performers to drag them down. However, it's kinda worse when kid's futures are in the balance.
Yes, there are certainly better and worse teachers, but it can be hard to determine who is doing well.