Comment by hedora

I don't think these details are particularly hard to work out:

- You can shrink the deficient school to zero by reducing teacher count starting in the lower grades and moving up, and by allowing parents to opt for transfers in higher grades.

- The building still exists, so you could reuse it. Or, investors could build a new school. Obviously, there's some lag in the measurement, since it requires a few years of student data. I'd say look at the first and second derivative of the test scores. Note that the claw-back model deeply screws over investors that fund substandard schools. This is likely to create stranded real-estate for the next round of investors to buy at a discount.

- The metrics are produced downstream, so there shouldn't be measurement bias. There probably will be self-selection bias. There are existing funding mechanisms to deal with challenging student bodies. If those are working, then the per-student funding of the old school with skyrocket. If the old school still fails, then that produces a high-revenue group of students for some other new school to take on. If those funding mechanisms are not working, then it creates an externally detectable signal to the outside world that the problem is one level up (no schools in certain areas), making it easy for voters / courts to intervene (currently, those funding mechanisms are failing, and no one is held accountable).