Takeaways from New Program-Level Data on the College Scorecard

The Department of Education recently released a new, more user-friendly version of its College Scorecard website, which is an important consumer resource that provides key information about college costs and student outcomes. Alongside improvements to the website’s user interface, including features that let students more easily compare different institutions, federally collected, program-level earnings data are widely available for the first time. The Department also finalized the program-level student debt data released provisionally earlier this year.

The release of these data, and their integration on the consumer website, marks a milestone in over six years of work across two administrations to provide students with more granular information about college outcomes, most recently as directed in the President’s March 2019 Executive Order. And while the data come with some noteworthy limitations, their availability is an important step forward in allowing practitioners, researchers, and policymakers to examine federal student debt and earnings for individual programs and credential levels.

About the New Data

While program-level[1] information on earnings and debt are useful complements to the institution-level data on earnings previously included on the Scorecard, the new data have several limitations that call for caution. The data only include students who receive federal loans and/or grants, a minority of students in some schools and programs. Because they focus on outcomes after graduation, they exclude students who left with no credential, despite those students being most likely to struggle with any debt incurred.[2] Also, while the newly available earnings and debt data cover most of the graduates who receive federal aid (slightly less than 84% for both measures), they exclude most programs because they are too small to assess given the Department’s privacy standards.[3]

Beyond limited coverage of students and programs, the earnings data have several limitations that are important to understand. Earnings are measured only one year after graduation and likely understate how much graduates will make in the long-term. Census Bureau data show that earnings grow considerably over the first ten years after graduation—and especially quickly for bachelor’s degree completers.[4] The data also can’t speak to likelihood of being employed – reflecting only those employed and excluding graduates who are looking for work or pursuing graduate education.

The debt data, too, have important limitations and do not include non-federal student loans. Because the federal government does not collect private student loan data, College Scorecard debt information includes only federal student loans. Yet, college-reported data demonstrate that private loan borrowing varies by school and can make up a meaningful share of a graduate’s total student loan debt. For example, the Scorecard shows that graduates leave the College of New Jersey with between $19,382-$27,000 in student loans, depending on the program. However, the college itself reported that indebted graduates leave school with nearly $40,000 in debt (available at College Insight); these same self-reported data show that 41 percent of their graduates’ debt is made up of loans from banks, schools, or their state – debt which is not included in the much lower Scorecard figures. The federal government is not currently able to include private loan debt in its calculations, but this is an important limitation not to overlook.

Finally, in addition to the limitations of both the program-level earnings and debt data, there are a couple important reasons why comparing the two can be problematic. One is that the data measures include different groups of students: program-level earnings data include all federally-aided students, while the debt data only include students who borrowed federal student loans. Also, debt and earnings data are measured at slightly different points in time (immediately after graduation vs. one year after graduation, respectively).[5]

Key Findings

Despite limitations, the program-level debt and earnings data still offer the opportunity for new insights into the outcomes of recent college graduates and where programs with relatively poor outcomes are concentrated. They also underscore the continuing need for more comprehensive, federally-collected student debt data.

• Lower-earning programs are concentrated in the for-profit sector. After one year, the typical (median) graduate makes less than $30,000 in 36 percent—or just over 16,000—of all programs with available data.[6] Low earnings are concentrated in the for-profit sector with 58 of programs falling below the $30,000 threshold compared to 31 percent of programs at public institutions and 26 percent of programs at private, not profit institutions (Table 1). While only 24 percent of programs are operated by for-profit institutions, the for-profit sector represents 39 percent of all programs where the typical graduate makes less than $30,000 (see Table 2 for program counts).
• Lower earnings are especially common among graduates of certificate programs, except for those at public institutions. The typical graduate makes less than $30,000 in 88 percent of for-profit certificate programs and 78 percent of private non-profit certificate programs, but only 46 percent of public certificate programs. The percentage of low-earning programs is also higher at for-profit associate programs (64%).
• Both earnings and debt levels vary by field of study. The new program-level earnings data in particular add more evidence of well-documented variations in earnings after graduation by field of study. Less well-known is that debt can also vary a lot by program within certain sectors. For example, the most common for-profit bachelor’s program (Business Administration, Management and Operations) leaves the typical borrower with $38,592 in federal debt compared to $23,995 for the second most common program (Registered Nursing, Nursing Administration, Nursing Research and Clinical Nursing).[7] Another example within the community college sector is that the typical borrower graduates with $16,925 at Health and Medical Administrative Services associate degree programs, more than double the typical debt amount ($7,575) at Biological and Physical Sciences associate degree programs.

To be sure, the release of program-level information on debt and earnings represents important progress for students, families, researchers, and policymakers interested in learning more about the outcomes of college graduates.  However, the Department of Education has touted this revamped transparency tool as a replacement for more direct federal accountability mechanisms like the federal gainful employment rule, a move that holds little water give the well-known limits to disclosure as a form of accountability. Transparency and accountability work together to make sure all students have access to quality education that leaves them better off and able to achieve their career goals. Indeed, the Scorecard update underscores the need both for continuing data improvements, such as the inclusion of private debt, as well as robust policy to protect students from costly, low-quality programs that are more likely to leave their students worse off than better.

Table 1: Share of Programs with Median Earnings Below $30K by Credential Level and Control

Table 2: Number of Programs with Median Earnings Below and Above $30K by Credential Level and Control

[1] Program-level information refers to data reported by the unique combination of institution (6-digit OPIED), location (UnitID), field of study (4-digit CIP), and credential level. For institutions with multiple branches (UnitIDs) that map to the same institution, the U.S. Department of Education calculates data at the aggregated institution level and data points are repeated across any branches that offer those fields of study. We did not remove these duplicates from our analysis so that our figures would remain consistent with program percentages and counts published by the Department. More in-depth analysis may require examination of unduplicated data.

[2] Other exclusions include students who were subsequently enrolled in school during the measurement year, died prior to the end of the measurement year, had loan deferments for military service during the measurement year, or did not work during the measurement year.

[3] https://collegescorecard.ed.gov/assets/FieldOfStudyDataDocumentation.pdf

[4] Ibid.

[5] Scorecard provides first-year earnings for those who graduated in 2014-15 and 2015- 16, measured in calendar years 2016 and 2017. Debt is measured immediately after graduation in 2014-15 and 2015-16.

[6] We analyze the $30,000 threshold because this represents the lowest income category that institutions report to the U.S. Department of Education on the breakdown of net price by income quintile. It also roughly approximates how much employed high school graduates make, adjusting to same inflation index as program-level earnings (2018 dollars).

[7] Figures on borrower debt are based on the amount graduating borrowers owe at the median program within each sector and field of study. For the same two programs, the amount of debt students typically graduate with also differ at the associate degree level ($29,437 for business vs. $24,200 for registered nursing) and certificate level ($18,034 for business vs. $10,725 for registered nursing).