The economic case for college is simple: college graduates make more money. In fact, they make over $500,000 more over the course of their lifetimes, on average. That works out, as we've noted before, to among the best returns-on-investment around. It works out to an annual return of around 15 percent a year. The stock market, by contrast, averages 6.8 percent annual returns, and housing averages 0.4 percent a year:

But there's a catch. This is the return to graduating from college and not everyone graduates. Only 58 percent of bachelor's degree seekers who enrolled in 2004 had gotten their degree by 2010, two years after they were supposed to receive it. The other 42 percent spent a lot of money going to college without getting a degree and all the benefits that come with it. When considering the benefits of college, one must take that risk into account. Not doing so would be like measuring the riskiness of a bond without considering the possibility that the company issuing it will default.

So, what happens to people who go to college and don't graduate? Do they make any more than their peers who didn't go on to college at all?

Actually, yes. Michael Greenstone and Adam Looney, who've done some of the most comprehensive work on the returns to college, have estimated the returns on going to college without getting a degree. They're very high!

On average, college dropouts make $8,000 a year more than high school graduates without any college, or $100,000 more over their lifetimes. Taking into account the cost of going to college for a certain period (1.83 years on average, for these students), the return on investment is significantly lower than that for bachelor's degrees or professional degrees (a category which includes medical, law and dental degrees, but not PhDs or master's degrees), but still higher than stocks, bonds, or any other conventional investment.

This doesn't prove the effect is causal, of course. Maybe students who go to college and drop out are likely to make more than students who don't start college for other reasons. Maybe they're more self-motivating or have stronger social networks, for example. But a huge body of economic literature suggests that schooling actually causes these earnings gains.

For example, Joshua Angrist and Stacey Chen found that people randomly drafted to serve in Vietnam received more education than those who didn't, due to the G.I. bill, and earned significantly more as a result. The group drafted and the group not drafted weren't very different otherwise; the selection was random. That suggests that the earnings gains are attributable to their greater access to college, rather than other cofactors. Studies comparing identical twins have found the same.

And some of this literature even touches on the "some college" question. Princeton's Cecilia Rouse and Harvard's Thomas Kane estimated that two years spent in community college — without attaining a degree — causes an increase in annual income of about 10-16 percent. Dropping out of college is unquestionably a worse economic bet than finishing it. But the evidence suggests that starting and not finishing is much better than never starting at all.