College Should Not Win by Default

Go to school, get the diploma, land a good job. A four-year degree should no longer win that argument by default. Young college graduates’ unemployment advantage over high school graduates has recently shrunk to its lowest level since the late 1970s.^[1] A particular program should have to justify its price and the years spent out of full-time work. An associate degree or apprenticeship should get a hearing before the first tuition bill, not after a disappointing job search.

The New York Fed’s national data^[2] show how unusual the shift is. People ages 22 to 27 without a bachelor’s degree had higher unemployment than all workers ages 16 to 65 in every month from January 1990 to September 2025. Recent graduates of the same ages, holding at least a bachelor’s degree, used to escape that pattern: their rate was lower than the all-worker rate in all but eight months from 1990 through 2018. Since January 2021, it has been higher every month. In January 1990, recent graduates were at 3.4%, against 5.1% for all workers; by September 2025, the rates were 5.8% and 4.1% in the December 17 release.

Unemployment rate of recent college graduates and all workers, 1990 to 2025

College graduates still have an unemployment advantage over high school graduates their own age, but it is shrinking. In a November 24 study, Alexander Cline and Barış Kaymak^[1] of the Cleveland Fed examined the unemployment gap between high school and college graduates ages 22 to 27. It was 2.5 percentage points on a 12-month average through July 2025. For decades it had fluctuated around 5 percentage points. They trace the decline in young graduates’ job-finding rate to around 2000. That is a longer-running problem than one unlucky graduating class. They suggest that rising college attainment, alongside demand growth that no longer favors graduates, may help explain the decline. If so, choosing a degree by habit means ignoring the competition for the jobs it is meant to secure.

New York Fed economists Jaison Abel and Richard Deitz^[3] make a strong case for college. Their April 16 analysis estimated a 12.5% internal rate of return for the median graduate in 2024, allowing for direct costs and forgone earnings. The return had held between 12% and 13% for three decades. Cline and Kaymak^[1] also found lower job-separation rates among young college graduates than high school graduates. But the median estimate assumes four years of study, average costs and median earnings. In a companion analysis, Abel and Deitz^[4] put the median return at about 9% for finishing in five years and 7% for six. Their estimate for a graduate at the 25th percentile of earnings was 2.6% in 2024 at baseline costs. They also found substantial differences across majors while holding net price constant. The program and its cost matter.

I want students to see the price of each route, not a ranking of respectability. Colleges should put net tuition and time away from paid work beside a program’s completion and employment record, then make the same comparison for an associate degree or apprenticeship. The alternative should have to demonstrate its merits, too; skipping college is not proof of having avoided a mistake. A college selling a degree should be able to name the jobs it prepares students for and show what happened to its recent graduates who sought them. A student should see both records before signing for the first loan or paying the first tuition bill.