U. of Ark. political scientist and education scholar Jay Greene has been blogging about the proper role of science in education policy, and his thoughts (continued here) are well worth reading. In particular, he warns that trying to scientifically find “the one best way” of evaluating teachers or of teaching reading and then attempting to impose that putatively best solution on all children is ultimately misguided and destructive.

I’d add that it is also unscientific. Science is humble. You have to be willing to rethink and potentially discard theories that repeatedly fail to coincide with reality. Well, the theory that governments can operate effective, efficient, innovative education systems from the top down was never supported by the evidence in the first place, and that theory is now buried beneath a vast pile of contrary findings. The system best supported by the empirical evidence is a parent-driven education marketplace such as the one Greene recommends.