The economic recovery that began in June 2009 is now 70 months old, making it the fourth longest expansion since the presidency of Martin van Buren. It is also a very sluggish recovery, with GDP growth at the lowest post-recession rates in history. The extraordinarily tepid pace of the recovery is rightfully worrying many Americans, with recent GDP figures signaling the potential end to the expansion. More troubling than slow growth is the composition of job creation.

From June 2009 to March 2015, the economy created a net 7.1 million new jobs. Of these, 6.5 million were held by adults with a bachelor’s degree or higher, while a further 1.9 million were held by adults who had been to college, but not completed a four-year degree. Now, an astute reader will say that there must be something amiss in Dr. Hicks’ mathematics. How can it be that we’ve created 7.1 million jobs, but 8.4 million of them went to those with a degree or some college experience? Well, these are net jobs. Employment for workers with only a high school degree dropped by 1.1 million and for those without a high school degree by 257,000 over the same time period.

It is worth summarizing: Since the end of the recession in Summer 2009, all the net job growth has gone to folks who went on to college. That is a 15.3 percent increase for college grads, a 5.6 percent increase for college attendees, and a 3.2 percent and 2.5 percent decline for high school grads and non-grads respectively. I think there are several implications of this labor market outcome.

The first is the absolute need for post-high school education. I’m sure there are occasional puppetry majors who are not working or underemployed; otherwise college grads are at full employment today. We probably could have more folks working if we had more college grads available. I have been and will continue to be a critic of universities where it is warranted, but these data offer pretty good evidence that we are underinvesting in education at both the family and state level. Where that underinvestment is occurring isn’t clear because the failure to succeed after high school may have its roots in early childhood education.

Second, we ought to be careful about what we do about the very real problem of income inequality. These jobs numbers aren’t fueled by CEO greed or stock wealth or any of the other shibboleths of envy. This looks precisely like a long-term structural change in the economy that values education. Tax policy won’t fix that.

Third, we should expect a significant revival of populism. The 2016 campaigns will everywhere be about ways to help poorly educated workers advance, and how to get everyone else more education. We ought to expect that almost everything else from global warming to international affairs will take a back seat to middle-class job opportunities. I only hope that the discussion will be about creating jobs of the future, not reviving jobs of the past.