Guest Post: Looking Under the Hood

Using Mobility Report Card Data to Understand What Works in Reducing Educational Inequalities

By Kelly Rosinger

This is the 10th and last post in a series we have run on groundbreaking research that looks at how effective different colleges are in providing social mobility to their students. To see previous posts, click here. And stay tuned, as we will be releasing a paper in the fall that includes the posts in this series as well as additional analyses of the Chetty data. 

Higher education has long been viewed as a vehicle of upward mobility, a pathway into the middle- and upper-middle class, for students from less-advantaged backgrounds. But higher education is not as reliable as we might want for a vehicle that is supposed to transport students up the income distribution. College entry, choice, and completion are deeply embedded in students’ economic backgrounds. In fact, the gap in college entry and completion rates between students from high- and low-income households continues to widen. These patterns become more troubling when we consider the types of colleges in which low-income students disproportionately enroll: less selective colleges, often with fewer financial resources and relatively poor graduation rates. When it comes to shuttling students up the income distribution, we’re leaving many stranded on the side of the road. 

The mobility report card data constructed by Raj Chetty and colleagues give researchers a place to look to find out what works and what doesn’t work in mitigating educational inequalities, something like a user’s manual for vehicle performance that can be used to improve upward mobility. For scholars like myself who study the impact of policies and interventions aimed at expanding educational opportunities for students from less-advantaged backgrounds, these data provide the clearest picture we have to date of the role that colleges play in access, success, and upward mobility. In previous work, researchers have largely examined enrollment among students who receive the Pell Grant, the federal government’s largest source of student grant aid, to examine college access. The Pell Grant, however, is a rough proxy for low-income status. By contrast, the Chetty data provide a more complete picture of students enrolled at particular colleges across the entire income distribution.

Until recently with the introduction of the U.S. Department of Education’s College Scorecard, there has also been little publicly available information about students’ earning outcomes after leaving a particular college. The college mobility data additionally link students’ earnings outcomes to parents’ income, expanding our understanding of how particular colleges contribute toward upward mobility.

Importantly, the mobility report card data provide a snapshot of student enrollment at most U.S. colleges over the first decade of the 2000s. This timeframe corresponds to a number of shifts in the higher education landscape. For instance, there have been significant changes in higher education finance at the federal and state levels and growing concerns over college access and affordability amidst rising tuition levels at many institutions. These shifts – as well as campuses’ responses to a changing landscape – are likely to influence the opportunities that students from various economic backgrounds have to attend particular colleges. Many of the nation’s most selective institutions, ones that appear to have particularly good earnings outcomes for low-income students but that enroll relatively few such students, have taken substantial steps over the last decade to alter admissions and financial aid processes in an attempt to expand access for students from less-advantaged backgrounds by, for example, making SAT or ACT test scores optional and replacing loans with grant aid. The full impact of these and other admissions and financial aid efforts on enrollment patterns and upward mobility are still relatively unknown.

At the same time, we know less about how policies and programs at many of the campuses that Chetty and colleagues identified as high mobility colleges, namely mid-tier public colleges, influence enrollment of students across the income distribution. This is particularly important given the declining shares of students coming from the lowest-income families that Chetty and colleagues document. For researchers, the Chetty data provide an opportunity to examine what policies programs, and interventions can support low-income students and provide pathways to the middle and upper class.

To be sure, there are limitations to the mobility report card data, many of which have been noted in this blog series. One clear limitation is the exclusion of non-traditional aged college students from the data. This population of students represents a large and growing proportion of college students, and their exclusion means future studies using the data will offer little insight into how to support upward mobility for this population. The clustering of some colleges into one, a function of how some colleges report information to the Internal Revenue Service, creates an additional challenge. Nearly one-in-five students represented in the data are enrolled at a college that is grouped with other colleges, making it difficult to understand what policies and conditions support upward mobility within these clustered campuses.

Despite these limitations, the Chetty data offer a look under the hood – to carry through with the vehicle metaphor – into what works to mitigate educational inequalities and promote upward mobility. They also show where we might shift gears to improve outcomes for students, especially students from lower-income households who may be stuck in neutral.

Kelly Rosinger is an Institute of Education Sciences’ Postdoctoral Research Associate in EdPolicyWorks at the University of Virginia and will be an assistant professor in the Department of Education Policy Studies and research associate in the Center for the Study of Higher Education at the Pennsylvania State University this fall.