Skip to Main Content
Speaking of the Economy
college tuition costs are high
Speaking of the Economy
Sept. 14, 2022

What's Behind the Rise in College Tuition?

Audiences: Economists, General Public, Policymakers

Grey Gordon discusses his research and other economists' work on the macroeconomic and policy factors that shape the price of higher education. Gordon is a senior economist at the Federal Reserve Bank of Richmond.



Tim Sablik: Hello, I'm Tim Sablik, senior economics writer at the Richmond Fed. Joining me today is Grey Gordon, senior economist in our Research Department here at the Richmond Fed. Today, we're going to be talking about the cost of college, specifically looking at tuition and also student loans, which is a topic that's been in the news recently with President Biden's announced plan to cancel some student debt.

Grey, welcome to the show.

Grey Gordon: Thanks for having me. It's great to be here.

Sablik: Understanding the factors that influence the cost of college is one of your research interests. We're going to be talking about a couple of recent papers that you've written on this topic. There is a variety of things that go into the cost of college, but I think today we're going to be focusing mainly on one of the biggest components, which is tuition.

First, I thought it would be helpful to describe how college tuition has changed over the last three or four decades. Can you provide us with a quick overview of just how much it is risen and where those increases have been the largest?

Gordon: Sure. There's two main types of tuition we think about. The first is sticker tuition. That's the amount that almost everyone nominally pays, and it's going to differ by in-state versus out-of-state at public schools. However, colleges engage in a lot of what economists call price discrimination, where they charge different people different amounts. The way that happens is giving more or less need-based aid or merit-based scholarships. It's also common to consider sticker tuition minus this institutional aid, which is something we call net tuition.

While both real sticker tuition and real net tuition have grown dramatically, real sticker tuition has grown faster. Real sticker tuition has averaged around 4.5 percent a year while real net tuition has averaged around 3 percent.

Part of the trouble with either of those measures is they're growing faster than income growth, so tuition relative to income has become considerably more expensive. The growth rates have been a bit larger for public schools than private schools, and that's somewhat natural because public schools started from a lower level. But the growth rates are fairly similar across all types of schools.

Sablik: You mentioned how the growth rate is outpacing income. It's also, until recently, outpacing generalized inflation rates. What are some of the different theories about what is driving these tuition and cost increases?

Gordon: Yeah, there's no shortage of theories here.

The first theory, which is quite popular, is that cuts in state support of public institutions have resulted in colleges raising tuition to make ends meet, so to speak. State support of public schools has fallen, relative to tuition especially.

Another popular theory is that the demand for college has gone up as a result of the returns to college education going up. When you look in the data at the college earnings premium, which is how much the college-educated worker earns relative to a non-college-educated worker, that has gone up dramatically since the early '90s. With that increase in the college earnings premium, it incentivizes youth to enroll in college. That increase in demand — in a classic supply-and demand framework — increases tuition as well as enrollment.

Another theory is that as incomes grow, the ability of parents to help their kids pay for school increases. So, if colleges are not cost minimizers but revenue maximizers, something called the Bowen rule asserts. Then, that can drive up college tuition as well. Essentially, colleges know that parents can afford more, and so they eke out a little bit of that extra in the form of higher tuition.

The last theory I'll give you is probably the most controversial one, which is called the Bennett hypothesis. That goes back to an op-ed in the New York Times by the then-Secretary of Education Bill Bennett. He said that colleges were being greedy and jacking up college tuition in response to increases in federal financial aid. That is also a demand-type story where student aid increases the demand for college, so that pushes up tuition. That demand increase is coming from a policy rather than general macroeconomic forces.

Sablik: Thanks for providing that rundown of those different theories. I think we'll zoom in a little bit on that last theory, the Bennett hypothesis, which is something you examined in a recent Economic Brief, which we'll put up a link to on our website. In that brief, you and your co-author attempted to measure how increasing the amount that students can borrow for college impacts tuition. What did you find?

Gordon: One of the ways to look at the question of how much student loans affect tuition is to look at a measure called a passthrough rate. Imagine that borrowing limits were increased by $1,000. How much would net tuition go up? If net tuition were to go up $500 in response to a $1,000 increase in the borrowing limit, that would be a 50 percent passthrough rate.

What we focused on in that Economic Brief was the time-varying nature of the passthrough rates that our research indicates. When you zoom in on the late 1980s and early 1990s, which is when Bill Bennett wrote his op-ed, the passthrough rate was quite high: around 15 percent for subsidized student loans. Before 1993, unsubsidized loans didn't exist. But if you think about a counterfactual where you create the unsubsidized loan program and you allow people to borrow $1,000 and then you compute how much net tuition changes in response to that, you will find that net tuition goes up by around $600, which implies a passthrough rate of 60 percent, which is just massive. What that is telling you is that the creation of the unsubsidized loan program in 1993 was really important for the increase in net tuition.

Another piece there is not just the creation of the unsubsidized loan program, but also how the loan limits have changed over time. After the big expansion in 1993, there was so much credit in the system [and] so much ability for students to take out student loans, expanding that borrowing capacity further would not change net tuition very much. Our model indicates that in 1994 and 1995, the passthrough rate went from 15 percent down to zero for subsidized loans. Over time, as tuition kept creeping up and the borrowing limits did not, that resulted in the passthrough rate increasing up until under President Obama, where they peaked at 12 percent for subsidized and unsubsidized loans. But then President Obama authorized an expansion in student loans, which then dropped the passthrough rate back down to zero.

Sablik: I just want to make sure I'm understanding correctly. It sounds like from what you found Bennett's hypothesis was sort of correct, but it's not a symmetric or gradual or constant increase or relationship. That seems to be kind of connected to the fact that the changes in student loan limits don't happen gradually over time. They seem to happen all at once in these big, sweeping legislative changes. Is that right?

Gordon: That's right. One of the things we point out in full research paper is some empirical research. For instance, by looking at Nadauld and Shen, where they look empirically at past rates and the response of tuition to loan limits. [They] find what period they're looking at impacts the response from net tuition, or the response of tuition to borrowing limits, which is consistent with our models.

Sablik: But would it be fair to say that if you look over the fullness of time — I think in your study, you are going back to the 1980s or 1970s, say 40 to 50 years — on average, the increase in student loan limits have been exerting some upward pressure on tuition?

Gordon: Yes, we do one exercise in our full research paper where we say, "Okay, imagine that student loan borrowing limits were like in 1987. That's a really radical change from today. But tuition in that case would be like 50 percent lower than today. So, the cumulative effect of all those loan limit increases has been large, according to our research.

Sablik: Another thing I was wondering about — I kind of alluded to [it] in the beginning — recently in the news there's talk about loan forgiveness. Have you looked into that at all? Do you think that loan forgiveness could have a similar impact on tuition as changing the lending limits, or do you think it would operate in a different way?

Gordon: Yeah, I haven't used the model to analyze it. But I've thought about the question a bunch.

Let me just point out something that should be kind of obvious, I think. If you forgive student loan debt today and then you never forgive it again in the future and everyone knows that, then essentially loan forgiveness should have basically no impact on tuition at all. Because what you're effectively doing is taxing one group of people [and] transferring those resources to another group of people, the ones who are having loans forgiven.

To a youth who is deciding what college do I go to and how much am I willing to pay, if they really don't expect their loan to be forgiven, it doesn't matter to them. It might matter through taxes and that sort of stuff, but it has no direct effect on tuition or demand for college. If you're a youth and you think there's some chance that in the future they're going to do this again, that could have a real impact on tuition.

Say the youth was thinking, okay, there's a 10 percent chance of [loan forgiveness] happening again in the next five years. Well, one back of the envelope calculation you can do is if there's a 10 percent chance of getting a $10,000 loan write-off, that's going to be similar to getting a Pell Grant of an extra $1,000 — 10 percent times $10,000. What we calculate is the Pell Grant passthrough rate is roughly 60 to 50 percent, much higher than for loans generally. Take 50 percent, multiply it by $1,000. That's a $500 net tuition increase you might expect. Now, if you knew that it was not a 10 percent chance of having your loan forgiven but 100 percent, that might be more like a $5,000 increase in net tuition. So really, it's about expectations of what's going to happen in the future. That's what matters for tuition levels.

Sablik: That brings up a good point. I'm not sure if we mentioned it in our previous discussion about loan limits. But I think it's important to emphasize the mechanism through which this loan limit increase passed through to tuition, in that it increases the demand for college. It allows students to afford more college or more expensive college and that's what pushes up the prices, right?

Gordon: Absolutely. It increases the ability of students to pay for school. As long as they have a commensurate desire to pay, then you get net tuition going up.

Sablik: In your working paper, which we'll also put a link up to, you also examined some of the other factors that could be contributing to rising college tuition. Which ones stood out as particularly important to you?

Gordon: I'd say the other key drivers were increases in parental income, which makes college effectively cheaper for students because their parents will help them pay. Obviously, that's not a universal truth — some parents won't, some parents will. But that increase in parental income — paired with increases in the returns to going to college, that rise in the college premium that I mentioned, and the flip side of that which is Baumol's cost disease — those things together explain something like 50 percent of the rise in net tuition. In our data, the Bennett hypothesis explains 50 percent as well. Together, we're at 100 percent if you include all those things.

One caveat is that in some of our experiments, state support really matters, depending on how you think about it. If you kept state support as a share of revenue fixed over time, that would have really decreased tuition at public schools.

Sablik: So, you've solved the mystery of college tuition increases. Nobody needs to go hunting for any more theories. [Laughs]

Gordon: Of course, yes, you hit the nail on the head there. There's lots more to learn and to explore in this space. But we think we have a reasonable answer to the question of why college tuition is going up over time.

Sablik: Are there any other questions related to college costs and student loans that you're hoping to explore in some future research that you're working on?

Gordon: Absolutely. One of the things we're thinking about is a policy that's been proposed for solving the student loan crisis, as it's sometimes called. Under the current system when a student defaults on their loan, the government effectively pays the tab. The government guarantees federal student loans under the current system. Under this proposed policy, it would be the individual schools who would be responsible for guaranteeing the loan.

If you had a school that was essentially charging outrageous tuition and just trying to milk students, trying to get as much money out of them as they could without increasing their earnings capacity, then that type of school would probably have high default rates. If they had to bear the cost of those higher default rates, they would have two options. One would be to run out of money. [Laughs] The other would be to lower the tuition such that the tuition matches the extra earnings capacity of the students, so the students can pay back and not default on their loans.

One of the nice things about this policy is it doesn't really tie the hands of colleges. If delivering a quality education is very expensive, it would allow a college — say Harvard — to charge extremely high tuition, have huge student loans, and then have the students make a ton of money and pay those back. And the default rates can be low. It's sort of screening what colleges are being productive and helping their students, and what colleges are really fleecing their students and not being productive.

Sablik: We'll definitely have to have you back on to talk about that research once it's progressed further. Thanks very much for coming on the show to talk about what you've been working on so far.

Gordon: Yeah, thanks for having me. It was great.

Sablik: As always, I'll remind our listeners interested in keeping up with the work that Grey and our other economists are doing to visit the research section of If you enjoyed this show, please consider leaving us a rating and review.

Phone Icon Contact Us

Research Department (804) 697-8000