Who should skip college?

Thursday, February 20th, 2025

The central thesis of The Case Against Education, Bryan Caplan explains, is that education has a low (indeed, negative) social return, because signaling, not building human capital, is its main function — but the selfish return to education is negative, too, for many students, depending on ability:

First and foremost: know thyself.

  • Don’t base your life choices on what your immediate social circle finds “demeaning.” As Dirty Jobs repeatedly proves, people routinely get used to jobs that initially disgust them.
  • Don’t base your life choices on whether parents and teachers constantly tell you that you’re “smart.” They’re not trustworthy assessors of your intelligence.
  • Don’t rule out options because they require “declining status.” If your family’s initial status is above average, declining status is the mathematical norm. That’s what “regression to the mean” means.

What should you do instead? First and foremost: Get objective evidence on your own intelligence.

  • If your SAT score is at 1200 or greater, your odds of successfully finishing a “real” major are quite good.
  • If your SAT is in the 1100-1200 range, it’s a toss-up.
  • If you’re in the 1000-1100 range, only try college if your peers consider you an annoyingly hard worker.
  • Below 1000? Don’t go.

[…]

What will go wrong if you ignore my advice? The most likely scenario is that you spend years worth of time and tuition, then fail to finish your degree. Maybe you’ll keep failing crucial classes. Maybe you’ll keep switching majors. Maybe you’ll die of boredom. The precise mechanism makes little difference: Since about 70% of the college payoff comes from completion, non-completion implies a terrible return on investment.

Comments

  1. Dan Kurt says:

    Forget the SAT as your touchstone as the IQ content has been watered down since the 1960s.. Pay the money and have a psychologist give you an IQ test, even a truncated one. If your results shows an IQ lower than 116 forget college and learn a trade. A military career is a great way to learn a useful technical job. If your IQ is 105 or lower, join a union and pray that automation does not make you redundant.

    Dan Kurt

  2. Jim says:

    Low-T economic parasite, Bryan Caplan, 53, would be warranted to plan to do something productive, now that President Trump is bearing down, shivs bared, upon Uncle Scam.

  3. Jim says:

    If it was this bad 42 years ago, just imagine how much unbelievably worse it is today…

    https://i.ibb.co/gFbhv9F3/fussell-class-university.jpg

  4. Mike in Boston says:

    My massage therapist can tell where it hurts, what needs to be loosened, and when a muscle has been pulled and needs treatment other than massage. My tree guy has a good sense of where to attach ropes and cut so the tree falls in the one place it won’t damage anything. I suspect neither of those guys would hit a score above Dan Kurt’s 105 on an IQ test. But I am sure that their ability to be good at what they do is not impaired by that fact; there is a sort of tactile, kinetic competence that seems mostly orthogonal to whatever an IQ test measures.

    There’s also common sense: they both knew better than to join the military and risk being sent to their deaths for no good reason by some fool politician.

  5. T. Beholder says:

    But, but, optimism!

    Jim says:

    If it was this bad 42 years ago, just imagine how much unbelievably worse it is today…

    Well, now these mommy’s smart babies grew up, hence the results.

    Mike in Boston says:

    But I am sure that their ability to be good at what they do is not impaired by that fact; there is a sort of tactile, kinetic competence that seems mostly orthogonal to whatever an IQ test measures.

    It’s a test disconnected from the real world or non-arbitrary practical abstractions (such as math). What can such tests measure?

    On the low levels: ability to follow instructions and learn procedures. So it can strongly suggest disabilities, but due to single-value output cannot reveal any specifics (what exactly the subject can and cannot perform), so not going to be very useful other than as preliminary.

    On the average and higher levels: ability to quickly figure out what the test-makers wanted the subject to do. Which does align well with suitability for some jobs… specifically, of the “bureaucrat” and “lackey” types.

    Hypothetically, it’s possible to build a mandarinate on something like that, but the likely result is an hierarchy filled by hapless followers, thus either paralyzed or guided by random fashions. But then, USSR achieved this kind of pervasive management senility simply via iterative promotion of “safe” mediocrity due to perverse incentives.

    See also: https://web.archive.org/web/1/medium.com/incerto/iq-is-largely-a-pseudoscientific-swindle-f131c101ba39

Leave a Reply