The previous post on the apparent decline in the average IQ of college graduates in the US over the last fifty years used the GSS‘ 10-question Wordsum vocabulary test as a basis for those IQ estimates. As was pointed out, vocabulary tends to increase with age (through the late fifties before peaking and then beginning to decline). That incremental increase through adulthood is very modest, however, reducing the gap between those who graduated in the 1960s and those who graduated in the 2010s by less than 1 IQ point.
To illustrate the change over time more starkly, the following graph shows, among those who spent at least four years in college, the percentages by year of participation in the survey who scored either 9 out of 10 or a perfect 10 out of 10 on the Wordsum test. Responses are restricted to those aged 25-40 at the time of participation so potential age confounds are eliminated:
The questions have been the same since the survey’s inception (see them here).
Forty years ago, 1-in-2 graduates could ace the Wordsum test. Today, 1-in-6 can. That’s just about perfectly in line with what we’d expect to see if the top 6% or so of the population is capable of acing the test.
Four decades ago, 12% of the population had degrees. Today, 33% does. If, in the early seventies, that 12% roughly corresponded with the top 12% of the IQ distribution, then the 6% of the population that aced the Wordsum test would comprise 1 in 2 of those grads. If today that 33% roughly corresponds with the top 33% of the IQ distribution, then the 6% of the population acing the Wordsum test would be a bit more than 1 in 6 of today’s grads. QED*.
This is a devastating refutation of contemporary educational romanticism. College isn’t anything close to a panacea. Education doesn’t increase intelligence. It doesn’t even appear, for most people, to do much in the way of increasing knowledge. These ten vocabulary words are all common enough that it’s almost inconceivable that after four years of college a student who paid attention and engaged with the classroom material would not have come into contact with all of them on multiple occasions.
The incentive structure for higher education is extremely perverse (or more charitably, is setup on the genuine presumption that spending time in college reliably increases earning power).
Student loans are the most difficult types of loans to have discharged, so lenders who lend to students are making what amount to government-guaranteed loans. After a few years interest on those loans runs at 5% or 6%. Lenders are thus earning 5% on things that putatively carry zero risk. It’s easy money.
The universities are getting paid by lenders no matter what so they have no incentive to restrict student enrollment either (unless they’re top-tier universities who trade on restricted access).
The zeitgeist says if you don’t go to college then you’re a loser, so lots of people who have no business being there end up going.
Outstanding student loan debt in the US is now at $1 trillion and climbing, but a huge chunk of that value is illusory value. Many of those loans aren’t going to be paid back. The people holding them do not have the prospects or the ability to ever pay them off. Lenders are booking them as essentially no-risk assets, but they’re not.
I don’t purport to be a financial expert–let alone an Economist!–but I can’t shake this feeling that we’ve seen something eerily similar to this before.
As the Z-Man is fond of saying, this will not end well.
GSS variables used: YEAR, EDUC(16-20), AGE(25-40), WORDSUM(9-10), BORN(1–except for 1974 and 1976 as the variable wasn’t introduced until 1978)
* Okay, not quite. About 12% of the population aces the Wordsum test. There is a correlation between intelligence and educational attainment, but it’s far from perfect and is becoming more and more attenuated as time goes on.