N.B. Under a signalling model of education, higher education subsidies will lower the wages of lower skilled workers.
— Garett Jones (@GarettJones) November 19, 2014
Bryan Caplan, who is working on a new book about education that should be very interesting, has also commented on this issue a lot. I would take it even further that Garett.
1) As he mentions, students who don't attend college see lower wages because they lack the credentials that are more widely distributed.
2) For students who are incentivized to go to college, but don't graduate, the subsidy will be associated with immediate missed wages and work experience, and more debt from the portion of their education expenses that was not subsidized.
3) The only students truly subsidized are the students who do graduate. And, they will tend to have higher lifetime earnings than the students who are net losers from the subsidies.
See Caplan's post, linked above, regarding graduation rates.
Further, there is no market failure here. There are huge private resources devoted to distributing scholarships by need and merit. The students that Caplan would encourage to attend college have many private options. The effect of public funds on opportunities for students is likely harmful for many reasons.
1) It leads to increased tuitions over time.
2) It introduces student selection criteria into the marketplace of grants, loans, and subsidies that are likely to be less reliable than the existing private selection processes.
3) The additional students selected into college attendance because of public subsidies will end up taking some of the private funding sources that could have been used by students who would have graduated. According to Caplan's data in the link, among all but the highest quartile of high school graduates, as measured by math scores, about 40% of high school graduates attend but do not graduate from college.
There is a lot of concern, across the political spectrum, for the high level of publicly subsidized student loans. High default rates are definitive evidence of the lack of reliability in student selection among these subsidies and loans. This is a case where the effect of public spending on national welfare could very likely be negative. The spending hurts us, not least of which those of us who can least afford it. We should stop it.