It's true - the costs of college education are generally outweighed by their benefits. But this only means that the choice to pay for college is a rational one, not that it isn't really expensive. Future income doesn't fly back through time to pay for college when the bill comes due. College is a tremendous expense. For many, that expense means being saddled with enormous debt through their early adult years. For those who work to support families, no future benefit can outweigh the loss of income in the present. Coming up with hundreds of thousands of dollars for education is not easy.
Also, the argument that college subsidies hurt the poor is a little silly. The poor have a light tax burden, and the actual dollar amount taken from the poor and spent on college education I'm sure pales in comparison to home mortgage interest deductions or any number of policies that don't serve the interests of the poor. Also, for a libertarian to argue that college subsidies divert resources from programs for the poor is a little eyebrow-raising. I mean, is this guy going to argue that if we got rid of these pesky subsidies we could finally have a guaranteed minimum income provided by the government?
Imagine for a moment that the one piece of data cited in the article went the other way. What if the cost of a college education actually outweighed the potential lifetime income to graduates. Would the libertarian author then seriously argue that we need government subsidies to fund higher education? Or would the government "[undermine the] efficient allocation of labor and [make] us all worse off by channeling too many workers into the wrong fields", as the author argues in point 2.
This is the most frustrating aspect of libertarianism to me, the argument that if the market does it, it is ideal. It's the best allocation of labor, of resources, of whatever. Why? Well, the market did it. As if the market were an omniscient god or something. It's just faith.
Finally, I would say that education is good beyond the increase in earning potential. While some of the benefits are fiscal, many are not so easily broken into dollars and cents. Learning and wisdom and knowledge make for a more informed and productive population, with more refinement in disposition and character. Ultimately, if the government subsidizes college education, doesn't that lead to a more educated and productive workforce, as well as greater equality and social mobility?
If college is made easier to attend, more people will go to college, including those who are denied it for class reasons. And isn't that a good result? It strikes me that this is where the argument actually is, not this silly business of whether lifetime income outweighs the costs, which is just an apples to oranges comparison.
p.s. The funniest thing on Volokh today, though, was this post, where the same author memorializes William F. Buckley and, admirably, notes the racism of Buckley's mag the National Review in the 50's. What's funny is that the first commentator then defends NR's editorial which defended denying black people the vote:
The main point of the linked piece on black voting rights is that voting should not be allowed when it leads to unenlightened outcomes. Many people can agree with that. I take it that what you find problematic is rather not that principle, but the empirical assumption it was based on, that blacks would vote in an unenlightended manner.Well, now.
Not knowing much about the time period, I don't know if those fears were justified.