“it seems to me that many people now talk about educational performance by students as simply and uncomplicatedly a product of school and teacher inputs. There’s a certain degree of variation between students in their academic performance, whether you’re looking at standardized testing or grades or whatever. What portion of that variation is controlled by teachers and schools is, I would argue, tremendously difficult to ascertain empirically, given the massive number of confounding variables and noise to deal with. But more and more, I see an attitude reflected in our popular discussion of education that presumes that teachers and schools control essentially all of that variation.”