#why does everything i upload decrease in quality help me

Baumol's Cost Disease

I should start by talking a little bit about what makes costs rise faster than inflation. This is the less important part of the post, but logically comes first; if this is old news to you, skip down to "But we're still missing something".

First we should remember that since inflation is an average, something has to increase in price faster than inflation. What sorts of goods should we expect to do this?

The traditionally answer is Baumol's Cost Disease, classically explained with the claim that "it still takes four experienced musicians to play a Mozart string quartet". And that's true, but it depends a lot on how you frame the good:

it’s true if you define the good narrowly (listening to a live performance of a 40 minute Mozart composition) but why should we define the good narrowly? If instead we define the good as “listen to music for 40 minutes” then it’s clear that costs have fallen dramatically.

Classically this is what "disruption" means: providing a different, and probably lower-quality good, but one so much cheaper to produce that you largely replace the original good.

What makes the cost disease bite?

The next obvious question is what factors can prevent this disruption. The two that occur to me right now (and I'm sure there are more) are regulation and status.

Regulation is an obvious culprit, and one that anyone who follows policy has seen some examples of. If, say, the government requires everyone providing a certain service to jump through a series of hoops, then there's no way for a disruptive challenger to cleverly avoid the need for some of those inputs.

E.g. it's hard to disrupt the medical industry because regulations require a licensed doctor before providing a number of services. (One could argue that the whole point of some regulations is to prevent certain types of "disruption"---maybe we actually don't want anyone to disrupt the medical industry with service that is 1/10 the price but five times more likely to kill you. Similarly with financial services).

Note that this behaves the same if the "regulations" are coming from private actors. One way in which the education sector is hard to disrupt is that many jobs want "a degree from an accredited four-year university", and there's a bunch of necessary-costs baked in there. (In this case, the regulations are partly imposed by potential employers who demand a degree, and partly by the accrediting agency which determines who counts).

Status is Robin Hanson's bailiwick. And while I find habitual Hansoning kind of annoying, that doesn't mean he isn't sometimes right.

I'm happy to admit that I get some of my clothes super cheap; when I say "this is an awesome suit and I got it at a thrift store for fifteen dollars" that's basically bragging. But it's hard to imagine a parent bragging to their friends about this great discount doctor they found for little Timmy.

Rather, people are inclined to compete over how many resources they devote to their children. It's much easier to imagine someone proudly talking about how much they've scrimped and saved to get the best for their child, than to imagine them boasting of how they managed to find a much cheaper school that's almost as good and then spent the rest on an awesome car.

But we're still missing something--Or are we?

So this all gives a pretty decent reason to expect college to rise in cost faster than average. So if prices had risen by 100% and college had risen by 150%, I'd be totally satisfied at this point. But that's not what happened; look at the graph up above. College costs have been rising five times as fast as inflation, and that seems to demand an explanation.

Except. We need to be very careful about how we define college costs here.

Need-based aid

One of the big trends in college financing over the past few decades is the rise of need-based aid. There are a number of federal and state grant programs that help support needy students, and those have been getting more generous over time. And a lot of colleges, especially the more elite colleges, provide a bunch of financial aid based on their evaluation of the students' perceived need. It's quite difficult to be too poor to afford Harvard, if you can get in.

So how does this affect college costs? Fortunately, we have those numbers. Unfortunately no one talks about them.

Here's a graph of sticker prices and "net prices" for non-profit and public four-year colleges, adjusted for inflation:

That doesn't look so bad! Running the actual numbers we see that for public four-year colleges, since 1991 real net Tuition, Fees, Room and Board has increased from \$7840 to \$14210--still a big increase, but not the absurd one posited by our initial graph.

In private colleges, the real net TFRB has increased from \$19,380 to \$26,100, which is still faster than inflation, but well within the "normal" range I posited earlier.

And what about community colleges? Well, we have this graph: and we see that real net TFRB has been flat, and if anything slightly decreasing since 1996.

Summarizing this, and averaging across students, we get something like the graph taken from this article:

(We should also notice how much of net TFRB is coming from the room-and-board portion, rather than the tuition-and-fees portion. Public four-year in-state colleges have average net tuition and fees under \$4000.)

This system is opaque and terrible

So things aren't anywhere near as bad as they looked when we started this post, at least in terms of actual college costs. But I don't want this to be taken quite as a defense of the way we structure college pricing, which is honestly quite awful.

First of all, the system is incredibly opaque. I had a bunch of friends in high school who said they weren't applying to any Ivies because they couldn't afford to go if they got in. Now it's true that they couldn't afford the sticker price, but neither can half the student body. But if you don't know that you can't respond to it.

And even if you know to look for these numbers, it is disturbingly hard to figure out how much you might have to pay. I'm sure some of you have been wondering why I keep pointing to Harvard, which is an outlier for basically everything. The answer is that they actually publish statistics, and an easy to use cost estimator. I tried to run a cost estimate for some made up numbers at a couple of other colleges (including my employer); they wanted me to spend ten minutes essentially replicating a full tax return, and then wouldn't give me numbers unless I gave them the name and mailing address of the prospective student.

So actual cost is much lower than sticker price, but it's really hard to figure out how much. And that's a huge stressor/disincentive for students applying to colleges, and makes it difficult to make sensible plans. (And like all opaque systems, benefits "insiders" who understand the system over "outsiders" who aren't aware of any of this shit).

Second, it leads to bizarre marginal incentives. If you've saved diligently for college, then your "need" is lower since you have more assets, so your actual cost is higher. Along some margins, money you save to pay for college is pretty much 100% wasted, since it reduces your need one-for-one.

Third, it totally fucks over people in some unusual situations. For instance, your "need" is assessed based on your parents' income and assets unless you're over 25 or fall into one of a few other weird cases. This is true regardless of whether your parents are helpign to pay for your college, or supporting to you in any way, or even on speaking terms with you.

(I have a friend who has trouble filling out the financial aid forms because you need your parents' financial information to fill them out, and their parents are just not responsive about telling them anything. I can only imagine it's even worse for people who are actually disowned/not on speaking terms with their parents. Can they fill out these forms at all?).

But we're handling it the wrong way

So the system is terrible and I don't want to be read as defending it. But the actual problems with the system are not the problems people complain about.

One big problem is that people don't know that the sticker price is kinda fake, and costs are in many cases lower than they appear. The absolute worst response to this problem is to highly publicize the sticker prices and complain loudly about how they're so high that no one can afford college. It seems like focusing on showing people clearly how to access all the grants, and keep costs down, would do a lot more good than yelling about the sticker price.

(Compare: buying a new car sucks. But information about how to negotiate with dealers, and how much other people have paid for similar cars, is really helpful. Rants about how scummy car dealers are, less so).