Q&A: How Do You Measure Value in Higher Education?

P. Morrissey / Flickr

The Rotunda at the University of Virginia.

In the State of the Union, President Obama gave a brief mention to higher-education reform, asking “Congress to change the Higher Education Act, so that affordability and value are included in determining which colleges receive certain types of federal aid.” Across the country, researchers are working hard on just this question: What “value” do colleges and universities add, and how do we measure it?

The Center for American Progress recently released a paper on performance-based funding for higher education, and last year, the Gates Foundation—in collaboration with HCM Strategists—began a project called “Context for Success.” Its goal? To give policymakers and colleges tools to better judge what works in higher education. To find more about what this entails, I had a brief conversation with Charles T. Clotfelter, a professor of public policy, economics and law at Duke University, who worked on the project.

Last night, President Obama mentioned that he wanted affordability and value to play a part in determining federal funding for higher education. But there seems to be a big challenge with regards to the question of value and how to measure it. What have you found in your work?

Traditionally, universities and people that study universities haven’t had terrific measures of the outcomes of what they do. In the same sense that a plant that makes donuts can count the donuts and somebody can evaluate if they’re good or not. In the case of higher education, we can count graduates, we can count the number of hours of people that sit in class, we can look at their grades—but it’s very difficult to measure what has been contributed by the institution. You don’t want to only look at things like how much graduates earn—the graduates might have been very talented applicants when they came in.

An economist would want to say, “Let’s look at the value-added. Let’s look at what’s been contributed by the university. And that should be our measure of performance.” Well, nobody has ever measured things like that. What we do is we have these rankings based on reputation. But that tells you as much about the endowments of the universities and the social class of the students as is does about the value added. People in universities think they are doing good things, but we don’t really have wonderful measures of that.

What kind of measures are researchers using to determine value?

Here’s a couple measures people have proposed. The first thought is to look at income or wages. In properly functioning labor markets, people are paid a wage that is proportional to their marginal product—if you are a more productive worker, you will be paid more. Even if that were entirely true, that still doesn’t take into account the fact that students who started at Williams and Princeton are a lot better than those that started at other places.

A second kind of measure is what percentage of the students entering are actually graduating. That’s a more stripped down—but logical—way to measure the performance of institutions. If you’re just bringing in a lot of people, but they’re not graduating, there’s something wrong there.

What you have told constant though is quality of the students. I worked on an effort to see if we can statistically correct for the inputs—the kind of students schools are getting. If you have three colleges that have students with the same kind of preparation, then how are they doing in the labor market and are they graduating at comparable rates? That’s when you can say you have a measure that’s good for making comparisons.

We have a ways to go in doing this. It would be nice if the president could come up for a measure of bang for the buck, but my guess is he can’t, because there’s no good measure of the bang. We can measure the buck, but we can’t measure the bang.

A lot of the conversation has focused on big institutions like my alma mater, the University of Virginia. But community colleges are arguably more important given the large population they serve. What kind of work has been done in measuring value at that institutional level?

We have just finished a paper on comparing community colleges in North Carolina. We ended up using measures of outcomes that look a lot like graduation rates. We said a community college student could succeed in two ways. One is to get an associate’s degree in applied fields—health sciences or public safety—or they could get enough units to transfer to a four-year university.

We measured all of the 58 community colleges in North Carolina, and looked at each student. We looked at whether they met any of these criteria within six years, and got a success rate for these two measures at each community colleges. Then we corrected for the strength of the students.

We found some community colleges do a better job in producing students that do well on these measures than others, even when you control for the strength of the students. We feel that’s a step in the right direction for doing this. For the community college system as a whole, we hope this will lead people to ask what particular colleges are doing right, and what is it that makes them do well?

This Q&A was edited for length and clarity.