So Matt Yglesias has a post up about a new report on the college wage premium (the amount of extra money a worker makes over his or her lifetime with a college degree vs. high school only) which finds it increasing. Yglesias takes the opportunity to remind us that this is why college is expensive:

[T]he basic shape of things is that the college wage premium is quite high and appears to be growing, so it shouldn’t be all that surprising that college is also very expensive. This is particularly true when you consider that most people seem to enjoy college quite apart from the pecuniary benefits, in part because it’s often possible to get someone else (parents, the government, etc.) to pick up a share of the tab.

Leaving aside that last part alone for cordiality’s sake, I want to look at the conclusion he’s drawing based on a correlation between education and income. As anyone who’s finished sixth-grade general science will tell you, a two-variable correlation is vulnerable to confounding variables. Here’s a graph of the wage premium from the Economic Policy Institute: