Tag Archives: Why Does Tuition Increase?

More on whether state disinvestment is causing rising tuition

The Times Higher Education (UK) ran a short piece based off this recent post. From the Times Higher Education piece:

over the past 26 years, average real state funding per student has fallen by $780, while tuition revenue has increased by more than $3,500. This means that cuts in state funding can only explain about 22 per cent of the increase in tuition over the past quarter century.

But there is a second empirical problem… when you plot tuition rises against changes in state funding, for every year since 1992, a $1 change in state funding per student is rarely associated with a $1 increase in tuition…



When looking at the past quarter century, on average, a $1 change in state funding is associated with only an $0.08 change in tuition…

Check out the full piece here.

New Record High Revenue per Student for Public Colleges in Fiscal Year 2016

Reanalysis of SHEEO data reveal that public college revenue per student reached a new record high for the second year in a row. For the most recent year (fiscal year 2016), revenues increased by $140 to set the new record of $13,259.

(Previous posts describe why reanalysis of SHEEO data is necessary – to adjust for inflation rather than costs.)

Three points stand out from this reanalysis.

  1. There is a small upward trend in revenue per student at public colleges.

As the chart below shows, there is a slight upward trend in revenue per student at public colleges. Revenue in 2016, was $2,789 higher than in 1991, an average increase of $107 per year.

  1. There is no long-term trend of state disinvestment. State funding is cyclical while tuition revenue steadily increases.

The second chart unstacks the bars by revenue source which makes it easier to see two things.

First, despite numerous assertions to the contrary, there is no long-term trend of state disinvestment in higher education. While educational appropriations in 2016 were $780 less than in 1991, the pattern is driven by the business cycle rather than showing a consistent long run trend. Indeed, appropriations are up $885 from the recession low of 2012.

Second, in contrast to changes in state funding, which swing from positive to negative with the business cycle, tuition almost always increases. Only two years saw decreases in tuition revenue (2000 and 2008). In 2016, tuition revenue per student was $3,569 higher than in 1991, an average annual increase of $137.

  1. Increases in tuition are not driven by changes in state funding.

Declines in state funding are the most common explanation for why tuition keeps increasing. But as the chart below shows, there is little relationship between the two. Each year label illustrates the change in state funding and the change in tuition revenue for that year. For example, the 2012 in the upper left corner indicates that between 2011 and 2012, state funding per student fell by $645 and tuition revenue per student increased by $391.

If tuition increased to offset declines in state funding, then each of the years should fall along the red line, which simply plots a $1 for $1 relationship. But statistical analysis does not support the idea that changes in tuition are driven by changes in state funding. In fact, historically, tuition only changes by 8 cents when state funding changes by $1, as shown in the blue line, a relationship that is not even statistically significant at conventional levels (p = 0.27). A more promising avenue for exploration is hinted at by the intercept term ($140, p < 0.01), which indicates that even if there is no change in state funding, tuition revenue would still increase by $140. Figuring out what’s driving this result is more likely to yield solutions to the problem of rising tuition than continuing to believe in the myth that tuition increases are driven by cuts in state funding.

Rick Seltzer covers the HECA vs. CPI debate

Rick Seltzer’s recent article in Inside Higher Ed covered the HECA vs. CPI debate. His piece drew upon this post and quoted me:

Misreading the trends in revenue because of cost adjustments will lead to misdiagnosing the way costs are changing over time — and misdiagnosing, in turn, the reasons tuition has been increasing, Gillen said. The common narrative is that declines in state funding have led to higher tuition, he said. But because his analysis shows state funding has cycled over time while revenue has crept up and tuition steadily increased, Gillen believes there has been too much emphasis on state funding. Increasing state funding is actually more likely to feed the trend of higher costs and tuition, he said.

So other elements of higher education finance need to be considered, Gillen said.

“That’s really why I keep writing,” he said. “If I’m wrong, then the way to keep tuition low is to keep increasing state funding. But if I’m right and you keep increasing state funding, that’s not going to do anything to tuition. Tuition is going to keep going up.”

Colleges and universities will raise all the money they can, and they will spend all the money they raise, Gillen said. Under his logic, an increase in state funding does nothing but increase the cap on what institutions can raise and spend.

Following Gillen’s reasoning can lead to very different ideas for keeping tuition in check.

Public College Revenue per Student Reaches Record High

Revenue per student at public colleges in the United States has reached a record high.

Reanalysis of the new State Higher Education Executive Officer’s (SHEEO) annual finance report indicates that in Fiscal Year 2015, revenues per student reached $12,972, surpassing the previous high of $12,440 from 2007. This detail is easily missed however, since the SHEEO report adjusts annual figures for costs rather than inflation. Note that the price index they use, the HECA, stands for the Higher Education Cost Adjustment. While SHEEO is a model of transparency (they note their use of HECA clearly and post their data online), if the goal is to determine if funding for higher education has changed over time, you need to adjust for inflation rather than costs.

An op-ed in Inside Higher Ed has more details, but adjusting for costs rather than inflation will lead to flawed conclusions. Consider the following:

  1. Cost adjustment hides the (slight) upward trend in total revenue over time

The cost adjusted figures reported by SHEEO claim that total revenue per student in 2015 was about $170 different than in 1999, leading to the conclusion that total revenue is just about stagnant. The inflation adjusted figures reveal the truth about a slight upward trend in total revenue. The first figure shows inflation adjusted total revenue per student at public colleges over time, broken down between educational appropriations (“state and local support available for public higher education operating expenses”) and net tuition revenue (tuition revenue after accounting for state and college funded financial aid).


While recessions cause temporary dips, total revenue is growing over time by about $100 per student per year. The cost adjusted figures from SHEEO largely obscure this trend.

  1. Cost adjustment implies a long-term decline in state funding when in fact state funding is cyclical.

The SHEEO report is often cited as exhibit A by those arguing that state disinvestment is a plague that has been causing havoc in higher education for decades. Yet adjusting for inflation rather than costs reveals that state funding per student is down $900 since 1990, not the $1,700 that SHEEO claims.

The strongest argument in favor of the long-term state disinvestment story is the fact that inflation adjusted state funding per student (educational appropriations) was almost $900 lower in 2015 than in 1990. It is certainly true that this is a large gap, but it is rapidly decreasing – state funding increased by almost $950 per student in just the past four years. The second figure, which simply unstacks the revenue bars hints at a more accurate story of what is going on: rather than being in a long-term decline, state funding follows a cyclical pattern, falling during recessions and rising during the recoveries.


  1. Cost adjustment misdiagnoses the cause of tuition increases. Tuition does not rise because state funding falls.

Perhaps the most commonly cited justification for increases in tuition at public universities are declines in state funding. Yet this notion finds little support in the SHEEO data. The second figure hints at the main problem, which is that state funding is cyclical, but tuition consistently increases. It’s difficult to explain how steady increases in tuition are driven by state funding which is subject to volatile swings both up and down.

The third figure directly tests the argument that tuition increases because state funding falls. Each year in the chart plots that year’s change in state funding and change in tuition revenue per student. For example, the “2012” label in the upper left corner indicates that from 2011 to 2012, state funding decreased by $621 and tuition revenue increased by $372 per student.

If tuition increases because state funding goes down, then there should be an offsetting relationship between the two with tuition increasing by $1 for every $1 decrease in state funding, as indicated by the red line.


Very few years fall along the red line. The blue line shows the best estimate of the true historical relationship, and indicates that a $1 decrease in state funding is correlated with only a $0.07 increase in tuition. Not only is that much smaller than the presumed $1 increase, but the relationship isn’t even statistically significant (p-value = 0.3). The historical relationship also indicates that if there was no change in state funding per student, tuition revenue would still increase by $133. To put that figure in perspective, the actual average annual change in tuition revenue over the past 25 years was $135. In other words, there is little reason to believe that changes in tuition are driven by changes in state funding.

Main Implication for college affordability

The main implication of these findings is that the traditional remedy for addressing college affordability will not work. The conventional wisdom, in part fueled by references to the cost (rather than inflation) adjusted SHEEO data, holds that increasing state funding is the key to making college more affordable. The assumption is that there is some cost of providing an education, and that this cost needs to be covered by either state funding or tuition. In such a world, the way to keep college affordable for students is to increase state funding.

But the results here clearly show that the conventional wisdom is wrong. The upward trend in total revenue combined with resounding empirical evidence that tuition does not reliably decrease in response to increases in state funding imply that increases in state funding are more likely to “feed the trend” of higher revenue per student rather than result in lower tuition.

If increasing state funding is not the solution to the college affordability problem, then what is? Researchers have not been able to settle on a definitive answer yet, but a particularly promising avenue builds upon Howard R. Bowen’s revenue theory of costs. Under this theory, the key to college affordability is to change the incentives that college’s face.


New Paper and Op-ed

I have a new paper at SSRN – Why Does Tuition Keep Increasing? The abstract is:

Tuition at American colleges has increased significantly over the years. Three common explanations for these increases are 1) declines in state funding, 2) increases in faculty salaries (Baumol’s cost disease), and 3) increases in college funded financial aid. Analysis of recent data shows these explanations for rising tuition to be inadequate. Even imposing the assumption of a $1 for $1 effect on tuition, an average of 18-33% of the change in tuition at four-year colleges is left unexplained by these three factors. More importantly, regression analysis provides overwhelming evidence that these three factors do not have a $1 for $1 effect on tuition. Using their estimated impacts implies that 66-91% of the increase in tuition at four-year colleges is left unexplained after accounting for changes in state funding, faculty compensation, and college-funded financial aid. The most likely explanation for the statistical inadequacy of these three factors as well as the most plausible explanation for the larger question of why tuition increases is that colleges are engaged in an academic arms race (Bowen’s Laws). Simply put, tuition keeps increasing because colleges are rewarded for spending more and raising tuition allows them to spend more.

If the full paper is a bit long for your tastes, there is a summary op-ed at Minding the Campus. I was even able to squeeze in one of the key charts from the paper: All.pub4.bargraph

Paul Campos on why tuition is so high

Paul Campos in the NYT has a good piece on whether tuition is so high because state funding has been cut:

ONCE upon a time in America, baby boomers paid for college with the money they made from their summer jobs. Then, over the course of the next few decades, public funding for higher education was slashed. These radical cuts forced universities to raise tuition year after year, which in turn forced the millennial generation to take on crushing educational debt loads, and everyone lived unhappily ever after.

This is the story college administrators like to tell… It is a fairy tale in the worst sense, in that it is not merely false, but rather almost the inverse of the truth…

Read the rest here.

Higher Ed Revenue per Student only $118 Short of Record High

New data from the annual SHEEO finance report reveals that educational revenue per public college student is at the third highest level in the past quarter century. When adjusted for inflation (see below), total revenue per student in 2014 was $12,329. The only two years with higher figures were 2008 at $12,396, and 2007 at $12,447.

However, most people won’t notice this because the SHEEO report itself and virtually all of the commentary focuses on figures using the Higher Education Cost Adjustment (HECA) to adjust for inflation. But the HECA doesn’t adjust for inflation, it adjusts for costs. If revenue goes up by 20%, and costs go up by 20%, then using HECA will show no increase in “inflation” adjusted revenue, which would be incorrect.

To find out how much revenue-per-student changed over time, you should adjust for inflation using something like the Consumer Price Index (CPI) (see this subtly titled report for more discussion on this point). While it is no secret that I have issues with SHEEO using the HECA instead of the CPI, SHEEO does deserve credit for being a model of transparency – their report clearly indicates that they use the HECA, and they make their data available (this year’s raw data isn’t ready yet, but they did make the highlights available here, including calculations using the CPI instead of the HECA.)

How could the seemingly boring choice of what price index to use possibly make any difference?

I’m glad you asked.

The distinction between adjusting for costs or inflation is important because it gives a false impression of why tuition increases. Adjusting for costs leads one to conclude that the cost to educate a college student has been about the same—$12,000— for the last 25 years. It therefore stands to reason that cuts in state funding cause tuition to rise dollar for dollar.

But if you adjust for inflation, the data no longer tell that story. The chart below shows total revenue per student over time adjusted for inflation rather than costs.

sheeo blog 1 While total revenue per student bounces up and down with recessions, there is an upward trend over time. From 2008 to 2012, state funding fell by 24 percent and total revenue by 8 percent, yet total revenue in 2012 was still 8 percent higher than in 1989. It is unlikely that we could experience such brutal reductions in state funding and total revenue and still end up with greater total revenue than we had 23 years ago unless there was an underlying upward trend in total revenue.

Because of this overall upward trend in total revenue per student, we shouldn’t expect to see a perfectly offsetting relationship between state funding and tuition. And if we look at the data, we don’t see much of an offsetting relationship. The next figure simply unstacks state funding and tuition revenue per student to show this more clearly.

sheeo blog 2State funding per student follows a cyclical pattern, falling during recessions; rising during recoveries. But tuition generally rises, regardless of whether state funding is rising or falling. Some might argue that tuition revenue rises even when state funding is rising to “catch-up” to the old baseline (note that tuition revenue doesn’t increase by enough to fully offset falls in state funding during recessions). But this argument is not convincing because of the overall upward trend in total revenue per student over time.

The next figure looks deeper into the relationship between state funding and tuition by plotting the change in tuition revenue per student in response to the change in state funding per student for each year.

sheeo blog 3The red line shows the presumed $1 for -$1 relationship where tuition revenue per student increases by $1 for every $1 decrease in state funding per student. If tuition only rises to offset cuts to state funding, then each year should fall somewhere along the red line. They do not. Far from being 1 for -1, the historical relationship is 0.11 to -1 (p-value = .14). Every $1 decrease in appropriations revenue per student is only associated with $0.11 higher tuition revenue per student.

From a public policy perspective, the bottom line is that tuition at state colleges does not rise merely to offset cuts in state funding. This means that increasing state funding is not an effective method of preventing tuition from rising. Based on the historical data from the last quarter century, a $100 increase in state funding per student will save each student only $11.

Another great post from Lloyd Armstrong

Lloyd Armstrong has a characteristically terrific post on why it is so hard to keep cost under control in higher education. My excerpts won’t do it justice, and only serve to peak your interest enough to to read the whole thing:

Universities often report a number that appears to indicate how much the university spends on instruction. We might believe that this number accurately represents teaching expenses and even do some analysis based on that belief. We would be wrong to do so.

John V. Lombardi, in How Universities Work

We enable (encourage?) this indeterminacy through our accepted form of higher education accounting – fund accounting. The focus of fund accounting is on management of individual sources of income, rather than on the expenses related to particular activities. As Lombardi points out;

Although fund accounting does not prevent universities from understanding their finances, it does not require them to do so.

Cost accounting, on the other hand, focuses on understanding the actual costs of a product, so that management can manage those costs… there is an enormous pressure on higher education to lower educational costs, which have grown much faster than inflation for many decades. The current system, which demands large annual real cost increases, is broken. It is very difficult to do experiments to lower costs of education if administrators do not know the original cost and the contributions of the various components of education to that cost…Charles Schwartz, Professor Emeritus at the University of California (UC) Berkeley… concludes that the actual amount of spending per undergraduate student on the education function at the UC in 2013-2104 is $7,500. This number is considerably less than the 2013-2014 average expenditure per student (averaged over both graduate and undergraduate students) of $18,060 reported by the UC. Schwartz’s calculated expenditures for undergraduate education are, in fact, even less than the tuition and fees currently required of the students!… Schwartz’s calculation suggests that the UC calculation has shifted costs equal to about $10,000 per undergraduate from the graduate education and research functions to the undergraduate instruction function…