Under the project studied, there were random assignments of teachers and students to classes. The striking thing in the findings is the identification of "Good" and "Bad" kindergarten classes, as shown by more variation in the Kindergarten test scores than would occur with random variation. The "Goodness" of the classes then have significant effects on their members for all those later life outcomes...
Actually, Professor Chetty was very careful in the seminar to say that there was no decisive evidence that it was the teacher who was the cause of the "Good" classes
2. John Kay on regulatory capture. Read the whole thing. Thanks to Alex Tabarrok for the pointer. When you read the piece, think of banks and financial regulation. I continue to predict that the blessed Elizabeth Warren will prove to be a friend to incumbent banks, as financial innovations that threaten them are likely to be deemed unsafe.
3. Karl Smith points to demographic data that support my last-stand-of-the-WORST pre-mortem of the election.
Incidentally, I asked Bryan in an email why, if the median voter theorem ensures competitive parties, my county of Montgomery County, Maryland is a one-party state. His response is that in "small" polities (under 10 million), you can sustain groupthink, but not in larger polities.
Hmmm....I don't see our local politics as sustained groupthink, although there is that. I see it as a vicious cycle. Everyone knows that Republicans cannot win here, so no interest groups support Republicans, so Republicans cannot win. I do not see why that same phenomenon could not emerge at the national level. Maybe at a national level there is more richness, diversity, and conflict among interest groups, so that parties can compete for their allegiance. Here at a local level, once you've kissed up to the teachers' unions and solidified your relationships with real estate developers, you can lock up the county.