Marginal returns to schooling

There has been some debate for a while about whether expanding college education helps people, on average, or whether it just pushes more people into college that will not really benefit from it.  In the American Economic Review I got today there is an article by Carneiro, Heckman and Vytlacil that argues that the marginal return to education for the average college attendee is six times higher than the marginal return for the marginal person, which is low enough to conclude that these college expansion policies were a bad deal for those who where induced to attend school that would not have otherwise.

The article is not a light read, as you might expect from these authors, but the methods and conclusions are interesting.