So, I'd like to draw your attention to this quite strange Washington Post... meme review... they made a few years ago. I'm not sure what the purpose of being a Paper Of Record is, but they sure are squandering it.
The Washington Post says, hey, not many people are getting degrees in gender studies, lay off – but doing that successfully responds to a critique I'm not entirely sure is being made. The point cranky people make isn't that colleges offer no degree programs that benefit people's lives – although it's very possible some people like Robin Hanson believe that to be the case – it's that people are making obviously foolish choices.
If we can have a national dialogue about the Tide Pod Challenge, then "it's not that common" does not address the critique.
The whole analysis is massively stupid anyway, which should be obvious if you think about it for a bit. Certainly these degree programs are more common than they used to be, so the critiques (which are also extremely weak) aren't even rebutted strongly, merely responded to. The concern isn't addressed but acknowledged.
A much better (but still flawed) analysis can be found here, made by Georgetown. But the numbers ought to surprise you – why would Net Present Value go up so much over time if (as we all experience) no one cares where you went to college if you've been in the field for more than a decade? If that's true, how can that be squared with an increasing NPV between decades 3 and 4?
Even the programs which list a negative change in income over time for people with the degree can list a near-7-figure NPV, so clearly they're messing up big time. These things should be considered at the margin, which means comparing not just against jobs you'd otherise have, and not just against debt you accumulate, but also against years of job experience lost while schooling and the actual loss of income in that time. It's an intensely costly affair when compared against the true alternative – and although it's a bit hard to figure out all the data needed to make the analysis, it's simple to say almost no liberal arts degrees make any sense at all. It'd simply be impossible. You're paying money to have access to an income stream identical to average, and it's ludicrously hard to find a good default to compare it to.
All of this being said, the distribution of credentials is patently absurd to discuss without comparing it to a distribution of jobs. This was my main problem with the Washinton Post piece. It's absolute madness to use any reference for the numbers aside from that. So here is the closest match I could find between their graphic and the job market.
Take it or leave it, but I think there's important info in the second chart, and it enhances the value of the first. Nothing you'd be surprised by, but as the old joke goes, if you're hoping to get a Gender Studies job, all the Gender Factories have shut down.