Why college is so expensive

Ok, I can't really answer that fully, and definitely not in the first post. So let's start with: how have prices and expenditures at liberal arts colleges changed over the last 20 years?

The first thing most people think about when they think about college these days is how expensive it’s gotten. How much more expensive, and more to the point, why?

---

First off, it’s true that the sticker price at liberal arts colleges has gone up a LOT in the last 20 years.

This plot shows the listed price for tuition, fees, room, and board at the top 100 liberal arts colleges in the United States. I’ve split them out into categories based on their average US News and World Report ranking over this time period (you’ll see why in a minute). See notes on the data below.¹

You can see in the plot that the sticker price of college has increased roughly 55 percent from 2001 to 2022. And that’s AFTER adjusting for general inflation. Whoa!! That’s a huge increase!

Another thing to notice in this plot is that the highly ranked colleges are charging a higher sticker price than the lower ranked colleges. Generally speaking, the top 10 and next 11-25 liberal arts colleges are charging about the same sticker price, those ranked 26-50 charge a bit less, and those ranked 50-100 are charging a lot less.

---

You’ll also hear that financial aid has increased, and that’s true too. How much has the price the average student actually pays at liberal arts colleges gone up?

This plot shows tuition, room, and board revenue minus financial aid per student. That net price has increased too, but only by about 17 percent. Again, that’s over and above general inflation, so a 17 percent increase is still a big increase. But not as much as the sticker price has increased!

Here again we see differences between colleges that are ranked differently. The top 10 colleges actually charge a bit less nowadays after accounting for financial aid than do those ranked 11-25. The schools ranked 26-50 are not far behind, while the schools ranked 51-100 charge a lot less after accounting for the financial aid discount.

What’s up with schools ranked 11-25 charging the average student more than schools in the top 10? That’s weird… keep reading.

---

How much is that average financial aid discount, you might ask, and how does it vary across school rankings and over time?

This plot shows the average discount rate for these groups of schools. Here we see that the schools ranked 51-100 offer the most financial aid (as a percentage of their sticker price), while the schools ranked 1-50 have bunched closer together. The average discount rate has increased over the last 20 years, from 23-28% to 35-50%.

This plot again hints at something important in the market: the discount rate for schools ranked 11-25 is lower than for schools ranked 1-10, while the top 10 look more like the middle 26-50. This means that schools ranked 11-25 are offering less financial aid than are schools ranked both above and below them.

Huh? Why are they keeping that money to spend on other things, and not giving it to students who need help paying for school?

---

To  understand this, you’ve got to know how many liberal arts college students are “full-pays.” That’s a crass way of labeling students (and their families) that are wealthy enough to pay full price.

This plot shows that for schools in the top 10 and those ranked 11-25, almost 40% of students receive no financial aid whatsoever.² The fraction of full-pays has not changed much over the last 20 years at these schools. By contrast, the fraction of full-pays at schools ranked 26-50 and 51-100 is both lower and has been falling.

---

What this means is the top 25 liberal arts colleges are in hot competition for students who can pay full price. They engage in this competition by spending money. Here we come to expenditures per student.

Expenditures have increased by roughly 30 percent from 2001 to 2022, after adjusting for inflation. More than the average price increase net of financial aid (17%), but less than the price increase for full-pays (55%). Let’s see how the deal for full-pay students varies across the rankings by comparing expenditures per student to the sticker price.

And here is where we understand why the schools ranked 11-25 are offering less of a financial aid discount than the top 10, and maybe why prices have been increasing overall.

The top 10 liberal arts colleges are spending much more money per full-pay student than those full-pay students pay in tuition, fees, room, and board. The full-pays who go to top 10 schools are getting a huge subsidy, almost $20,000 a year. By contrast, full-pay students at schools ranked 11-25 used to be getting a subsidy (meaning the college spent more on them than they paid the college), but now they’re more or less breaking even. Full-pays at schools ranked 26-100 are now and have for a while paid more than their colleges spend on them.

The schools in the top 25 may be in hot competition for full-pay students, but the schools in the top 10 are clearly winning. It’s a race to top quality, and they’ve got money to burn.

The schools ranked 11-25 are trying to keep up their expenditures to compete for those full-pays, and they’re doing it by offering less financial aid to non-full-pay students. Furthermore, I wonder if this hot competition at the top is what’s pushing prices up across the industry. The top 25 schools have normalized supra-inflationary price increases.

---

Wow, you’re probably thinking. Just wow. How on earth are the top 10 schools spending $20,000 more per student than even the full-pays are paying in? Where does all that extra money come from? And what exactly are the top 10 schools buying that other schools can’t afford? Does that extra spending actually translate into better educational outcomes?

Subscribe now for more later.

1. The data in these plots are from IPEDS (the Integrated Postsecondary Education Data System, data here). Every college and university that participates in the federal student financial aid program is required to submit information to this database, so it’s pretty complete. I think it’s also pretty accurate, but I’d love to hear from others who work with this data about their experiences.

I’ve excluded Bryn Mawr from these plots because there’s something funky going on with their revenue numbers between 2004 and 2005. Maybe I’ll chase down exactly what’s happening with them later; for now I’ve just dropped them.

I’ve adjusted all of these numbers for inflation as measured by the Personal Consumption Expenditures Chain-type Price Index, provided by FRED here. As any Principles of Economics teacher can tell you, the PCE usually reports a slightly lower inflation rate than the CPI (i.e. what’s usually reported on the news) but a slightly higher inflation rate than the GDP deflator. I prefer the PCE because it’s the broadest measure of the prices consumers in the US pay.

2. This actually probably understates the fraction of students who should be considered “full pays” because tiny vanity grants (like, “here’s $100 for being our favorite pizza shop customer!”) are included as “some financial aid” so those students aren’t counted as “full pays” and probably really should.