The same caveat still applies, though. Sure, when you go to college, you are more likely to marry a college grad. But how much of that is causal? Perhaps you are already “the sort of person who marries college graduates”, and even if you don’t go to college, you’re already likely to marry college graduates.

Expand full comment

I think the $30k is possibly much less valuable than the impact of on the risk of divorce? It's something like a 30% decrease in divorce rate comparing a person with a bachelor's degree to a person with no college. And then if you do get divorced, it's likely to be much more financially painful if your spouse is a low earner (whether you want to look at it as the spouse paying alimony/child support or receiving it).

But definitely the impact on marriage prospects is not only a big deal, but a long recognized big deal. Even when I was in school, there were plenty of girls basically known to be looking for their "MRS" degree. We have talked about helping our kids get started in a business versus college and by far the biggest hesitation we have with the no college option is the impact on their friend group if they don't attend. The second is just that college was a fun experience for us and we'd like for them to have that also if it's not ruinous for them.

Expand full comment

That's another negative in my view. It has that word there, "marriage". Scary stuff.

Expand full comment

Yes, and don't forget the benefits of forming friends there too.

Effectively well-off friends serve as a kind of insurance against really bad outcomes so there is a huge benefit there not captured in the pure income data[1].

Also, and perhaps more importantly, college is a particularly good place to find friends and if you are the sort of person who could go to college you benefit from meeting people like you.

Analysis of returns to education often miss the fact that there are benefits just from sorting people on intellectual affinity. Sure, you may have friends either way but both intellectually inclined and non-inclined individuals benefit (individually) from having friends like them. Even if the country as a whole may suffer politically because of effective sorting at scale.


1: how much would you pay for the insurance of having someone with extra rooms in a nice school district take you and the kids in for a year or so if shit goes really bad ... versus friends who would also try to help but on who you would be a huge burden on and you and your family would be on a futon in the living room of a small apartment. Also friends help find you jobs when shit goes bad and while mean effect is captured in income data again it decreases variance which, given sublinear returns to income, is quite valuable.

Expand full comment