Tyler Cowen takes a premise of Arnold Kling’s:
My guess is that most of the younger (aged 30-34) highly-educated folks are children of two highly-educated parents. My guess is that many of the older (aged 55-59) highly-educated folks are children of a college-educated father but not a college-educated mother.
I don’t think we have a recipe that says, “Take a child of two non-college educated parents, add primary education ingredient X, bake, and out comes a college-capable high school graduate.” The mystery ingredient X has yet to be discovered.
and spins it as well as he can:
I’m not sure that college education per se is the key here (and probably Arnold would agree; read his phrasing carefully). If you can bake up some low rates of time preference, you’re coming pretty close to the real mystery ingredient.
Actually, I’ll stick here for the answer that Kling alludes to without mentioning it: The (original) G.I. Bill, which is the reason many of those 55-59-year-olds had a college-educated father in the first place.