> The first is why the teachers that are bad are actually bad. Under what circumstance are you willing to buy a poorly made car? After all, maybe the workers

> The first is why the teachers that are bad are actually bad. Under what circumstance are you willing to buy a poorly made car? After all, maybe the workers could become better given additional retraining, etc. > The other is how they are measuring teacher performance. Feel free to provide a better measurement scheme. Until then, why should we prefer nothing? In practice, we find that the "need better measurement" folk are actually unwilling to accept any measurement scheme. Yes, even one of their own devising.