In a recent post, I argued that Avery and Turner’s research, cited by The New York Times' David Leonhardt, ignores the experiences of students of color. For a variety of reasons, such as labor market discrimination, workers’ outcomes diverge significantly based on race. Research from CEPR, for example, showed that in 2013, recent black college graduates had more than double the unemployment rate (12.4 percent) of recent college graduates in general (5.6 percent), and more than half (55.9 percent) worked in jobs that do not require a college degree. Against this background, simply looking at the average return to college does not cut it; the payoffs are smaller and less certain for some groups of students than the overall average suggests.

Growing wage dispersion among college workers compounds this issue. While the rising difference between those with and without a college degree is well documented, fewer people discuss the substantial subset of college graduates who end up making less than high school graduates. John Schmitt and Heather Boushey found that for 25-34 year olds in 2009, one-in-five men with a college degree earned less than the average man with a high school degree. One-in-seven young college-educated women were in a similar position.

While the average person with a college degree outperforms the average high school graduate, it is likely that the marginal college enrollee (a person considering whether or not to attend college) will earn less than the current average college graduate, especially if we are talking about a large expansion in college enrollment. Furthermore, someone right at the margin between attending college or not would likely expect to have earnings as a high school graduate that are higher than the average high school grad even if they opted not to go to college. The potential student also cannot ignore the fairly high probability of leaving college before they earn a degree but after accumulating one or more years of debt, which can be especially high at for-profit colleges that also tend to target poor and minority students.

Considering these sobering facts, college doesn’t look like a “no-brainer” for everyone. As Schmitt and Boushey explain, individual students may rationally decide to avoid college even if this decision is socially unproductive.

Importantly, this blog post is not part of a financial advice column, and I would never discourage someone interested in applying to college from doing so. Instead, I and anyone else interested in increasing college enrollment needs to reconcile why it is that so many bright high school students decide to leave so much money on the table when they avoid college. The high price and weak economy are clear culprits. One way to increase the likelihood that students will decide to attend college is to lower the cost, thereby reducing the risk that they will end up with high debt and still have poor labor market prospects. Leonhardt ends his piece with the argument that 15 or 17 years of education should be the universal goal instead of 13. As the data clearly show, current policy is not likely to take us there.