Trigger warning: I am not fully going to answer the question posed in the headline above. But it is one needing answering, so we should do some serious thinking about it.

Let us consider a class of 100 freshmen at an average American high school in the fall of 2011, most of them probably born in 1997 (now 23 years old). If this group were typical of Americans, the overwhelming majority (say 90 of the 100) would graduate from high school, although some would not do so precisely in the expected four years—but for the sake of our example, let’s assume 90 students did graduate in the spring of 2015.

If the group were typical, about 60 of the remaining 90 students would enter college, some at community colleges, but more would attend four year universities pursuing a bachelor’s degree. About 40 % of kids attending four year universities actually graduate from college in the traditional four years—if we stretch things to six years, the proportion approximates 60%. Community college attrition typically is even higher. Therefore, it would be probably be true that at most 24 of the original 100 students would have bachelor’s or even associate degrees by the summer of 2019; if we stretch things to allow for fifth and even six year seniors, we might end up with as many of 36 (60 percent) of the 60 entering college actually graduating by this June.

But merely having a diploma doesn’t guarantee a person a job utilizing the skills associated with college level (as opposed to lesser) training. Data from the New York Federal Reserve shows that the “underemployment rate” among college grads (bachelor degree holders) approaches 40 %. These are many individuals who are doing jobs that are typically performed by high school graduates and require no higher level educational skills. Thus if we have 36 of the original 100 students that entered high school actually getting college degrees, only about 22 (a bit over 60%) of them will actually enter into jobs suited to their college training. In some sense, three out of four entering college fail to achieve expectations concerning their education.

That does not mean necessarily that the non-achievers were all “failures” in some sense—many, for example, probably are leading reasonably productive and happy lives with decent jobs and just a high school diploma. But it does also suggest that there are many young Americans having generally unpleasant experiences ending up with their vocational or professional aspirations unrealized, a small portion of those early on based on unsatisfactory high school experiences, but many more thereafter.

Part of the problem relates to poor information. High school and even college students are not terribly knowledgeable about labor markets, nor even about their probability of college success. Colleges lure kids into programs when they know the probability of academic success is low. High school guidance counselors have been brainwashed by business and political leaders into believing college works for nearly all, and sometimes push students beyond what they realistically can be expected to achieve.

One way to deal with this problem is to get better, more objective information to parents and students. The College Scoreboard of the U.S. Department of Education is a good tool, gaining greater utility each year, but still has some real imperfections.

There is also a perverse incentives problem. The salaries and job security of top college personnel are often dependent on enrollments—tuition fees and enrollment-based state subsidies help pay their salaries. Thus at many schools, high school students are induced to enroll based on blatantly misleading advertising.

One solution to perverse incentives is to make colleges have skin in the game. Specifically, if large numbers of students drop out of school and default or fail to make payments on student loans, the college should share some of the financial liability otherwise falling on taxpayers. Incentivize colleges to lower the attrition rate.

Many kids belong in different kinds of postsecondary training programs, things like coding academies or truck driving school. Pell Grants and state subsidies supporting traditional college education are often biased against other forms of vocational training that is often cheaper and more appropriate to some student needs.