Does the 90-10 Rule Cause Tuition Increases? (Part 1)

Concerns about the quality of for-profit schools in the early 1990s prompted Congress to pass a rule limiting the share of revenue that proprietary institutions could receive from specific federal student aid programs (Pell Grants, loans, and work-study) to 85 percent. The idea was that these institutions would have to show that private demand for their product existed and that students who attended would have some of their own money at stake, rather than relying only on federal assistance.

In the years since, the 85/15 rule became 90/10—allowing schools to take in a greater share of their revenue from federal student aid programs. Today, no one is happy with the rule as is. Some people who generally think the rule is a good idea (myself included), question why veterans benefits and other federal programs are counted in the 10 percent, instead of excluding it from the calculation or including it in the 90 percent. Opponents of the 90/10 rule—mostly  leaders of for-profit colleges and associated trade associations—say that the revenue cap drives tuition increases so that they stay above rising student aid limits.

It’s tough to prove the connection between 90/10 and tuition. Pricing is a complex process that takes into account multiple issues related to expenses like course materials, leased classroom space, and faculty; debt payments; profit margins; and student demand. Plus, not every student borrows, and of those who do take on debt, not all of them take the maximum. That’s why tuition at for-profit schools is not a few percentage points higher than the average loan debt and Pell award. Nor are tuition increases matched dollar-for-dollar with changes in Pell awards and student loan limits.

Fortunately, the U.S. Department of Education’s Data Center presents one way to start measuring the potential connections between a school’s 90/10 percentage and its cost. On that site is a document (PDF) listing each school’s 90/10 percentage for the fiscal year that ended between July 1, 2007 and June 30, 2008.

To test the connection between 90/10 and cost, I extracted these numbers from the PDF and put them in a database with a number of other variables related to size, sector, and demographics. In this post, I’ll talk about what the numbers show as a one-year snapshot. In a subsequent post later this week, I’ll look at how schools with a high 90/10 figure in this year have changed their prices over time. I’ll put a quick note on methodology at the bottom of he post.

First, the overall picture. The scatterplot below shows all schools’ 90/10 percentage on the x-axis and their tuition and fees and books and supplies for the 2007-08 academic year on the y-axis.

Perhaps the most interesting thing this chart shows is how clustered the cost of for-profit colleges are around certain levels, but there’s no clear story here about cost and 90/10.

Just for some context, here’s the graph a school’s percentage of Pell students compared to 90/10. As expected, there’s a clear connection–the more students that get a Pell Grant,  the higher a school’s 90/10 percentage.

But looking only at the overall picture of 90/10 isn’t a fair analysis—there’s a lot of variety in the for-profit sector in terms of size, sector, and demographics. So I decided to break the database down by a number of different ways. Below are several examples, but if there is another combination you would like to see, let me know in the comments and I will see what I can do.

The chart below shows cost versus 90/10 by a school’s sector:

Only four-year schools appear to show any pattern here. And the one displayed is somewhat negative—schools with lower prices appear to have somewhat higher 90/10 percentages. This could be a connection between 90/10 and prices if the low tuition levels meant that Pell and student loan dollars could cover almost the whole cost, leading to a high percentage of revenue coming from this source. But it’s curious that a similar pattern doesn’t show up in the other sectors.

The next thing I tested was size. I grouped schools into the following categories based upon the number of students: 0 to 99; 100 to 499; 500 to 999; 1,000 to 4,999; 5,000 to 9,999; and 10,000 and over. The idea is that bigger schools might be enrolling vast numbers of lower-income students, putting them at risk of 90/10 problems, while smaller ones might be in a position where a single student’s borrowing could tip the balance one way or the other. (Please note that I combined some of the largest size categories because the top two only had a handful of schools and to make the number of schools more comparable.)

Small schools don’t appear to show any pattern, while the larger ones again display a slightly negative correlation. The difference here is that none of the cost figures with high 90/10 percentages is particularly low—they all appear to cluster around a narrow range of prices. That suggests that prices may instead be a greater reflection of competition within the sector and amongst similar schools, rather than a response to the revenue cap.

As a final test, I also went through and selected any school that is part of a publicly traded corporation and compared them to non-publicly traded schools.

What’s striking about the publicly traded graph is how many schools have high 90/10 percentages and how many are clustered around a cost band of about $5,000. That gives the appearance that schools with somewhat lower tuition are on the higher end of 90/10, but that also might reflect the large publicly traded companies competing amongst each other to have prices in a pretty close range. By contrast, the non-publicly traded companies really are all over the place, though they are mostly concentrated at or above 50 percent.

So what does all this show us? Only certain groups of colleges appear to show any connection between 90/10 and cost. These tend to be bigger, publicly traded institutions. But isolating how much of the pricing there is determined by 90/10 versus other factors (such as competition) is hard to say given how tightly clustered their reported costs are. Also keep in mind that larger colleges are going to be more sophisticated in carefully managing their 90/10 percentage to stay under the cap, so the picture presented can be somewhat distorted.

Of course, one of the difficulties with the above charts is they only present a snapshot in time. We don’t have 90/10 data over multiple years to see how that percentage changed over time compared to tuition. But since these figures are a few years old, it is possible to see how the tuition rates changed among colleges. This can be useful in seeing whether the high 90/10 rates did actually result in colleges raising prices. In the next post, I’ll look at how the change over time in tuition compares depending on a school’s 90/10 rate from these data.

Methodology: All 90/10 figures came from the department’s data center; all other information came from the Integrated Postsecondary Education Data System (IPEDS). Because the 90/10 figures are organized by OPEID, in many cases a single line might represent multiple campuses (the University of Phoenix, for example, has one line for all of its campuses). Where possible, I included all the students at any campus covered by that OPEID. So the percentage of Pell students reflects the total for all campuses, as do the size variables, and racial breakdown. For tuition and fees, I took the figure reported for just the OPEID of that campus. I did not include cost of attendance because those figures are less under the control of the institution since they are an estimate of cost of living in that area and can often be almost as high as the tuition and fee figure. I also removed all price figures that were two or more standard deviations above or below the average in order to eliminate outliers.