May 22, 2017
A recent study by economists Raj Chetty, John Friedman, and their colleagues calculates economic mobility rates--called Mobility Report Cards--for colleges across the U.S. Their findings offer an important picture of socioeconomic diversity and economic mobility across thousands of colleges and universities. But nearly as interesting as the findings of the study are the data that made these findings possible.
To construct the Mobility Report Cards, the researchers worked with existing student data from the Department of Education and held in federal tax records to construct a near-universal set of all students who entered college between 1999 and 2013. The study highlights the potential of leveraging already-collected data that could be used to provide better, clearer information to help students make smart choices about where to go to college--the goal of a bipartisan, bicameral bill introduced in Congress last week.
But to answer more of the questions on students’ and policymakers’ minds--like at which schools and in which majors they have the best chances to graduate and earn well-paying jobs--we need more and better information. Unfortunately, a federal ban that Congress put in place in 2008 prohibits the Education Department from developing a student-level data network that could leverage data across the federal government to answer those questions. As a result, we’re not able to evaluate an institution’s outcomes with all of its students, or assess gaps in equity within schools, for instance. Many of the most important measures of colleges’ success with their students remain out of reach.
The Chetty study gives us a glimpse of the potential of such an enhanced data system. The researchers compared Social Security Administration data with IRS 1098-T records and Education Department information on Pell Grant recipients to determine which institutions students attended. The 1098-T forms must be completed by colleges for every student who enrolls and pays tuition, and they include data on tuition charges and scholarships received. The Pell Grant data fill in most of the blanks, providing the institutional data for students who received tuition waivers and thus may not have 1098-T forms. From these matches, the researchers compiled a largely complete dataset of more than 30 million students who attended college during the time period they were studying. This dataset is more comprehensive than the Education Department can build today, precisely because of the ban on a unit record data system.
Still, the dataset only scratches the surface on the important questions students have about the colleges they’re considering. For instance, the Chetty data don’t include students’ programs of study, so it’s impossible to say which programs within an institution provide low-income students with the best opportunities. That information is important given that some students don’t have many options for colleges near them. In addition, the data don’t tell us whether students graduated, making it difficult for college-goers to find answers to some of their biggest questions. For instance, they can’t answer key questions like: Are students like me likely to complete their degrees at this institution? Nor can they say how long it typically takes students to complete a degree at their prospective college.
Despite these limitations, the Chetty study goes far beyond the limited data available to the Department of Education. For instance, because of the 2008 ban on a unit record data system, the Department of Education maintains data only for students who receive federal financial aid, like Pell Grants and student loans. But nearly a third of all college students don’t receive any federal aid, and nearly two out of every five community college students pay their own way or cover the costs with other forms of aid. That means data on student outcomes, like the median earnings of alumni at each institution that the Education Department produced and published for the first time in 2015, don’t tell the whole story because they include income data for only a subset of students. The unit-record ban effectively prevents the Education Department from accessing data that would allow it to provide the most comprehensive information to students and accurately represent institutions’ outcomes.
Politicians of all stripes are looking for a solution to the problem. Rep. Virginia Foxx (R-NC), who leads the House education committee and authored the ban, has said her first principle for rewriting the Higher Education Act is empowering students and families to make better-informed decisions. “Information is all but impossible to come by,” she noted at an accreditation conference earlier this year. “And the information that is available is often misleading or simply inaccurate.” Sen. Lamar Alexander (R-TN), who heads up the Senate education committee, has explored the issue in a white paper about consumer information, where he sought ways to provide accurate, comparable information across institutions to help students make choices about where to go to college. And a bipartisan group of lawmakers from the Senate HELP and House Education & the Workforce Committees took a giant leap forward last week when they introduced a bill that would repeal the ban and provide students with the information they deserve as they make one of the most important, expensive decisions of their lives.
Raj Chetty and his research team have shown us the kinds of questions we can answer with more comprehensive data about colleges and their students. Now we need to take the next step and, once and for all, end the ban that has kept students and policymakers in the dark.
This is the ninth post in a series we are running about new, groundbreaking research that looks at how effective different colleges are in providing social mobility to their students. To see previous posts, click here.