Graduate Education: Where Is The Data?

Graduate education is increasingly in the spotlight, but despite the attention paid to student debt headlines and newly established graduate loan caps under the One Big Beautiful Bill Act (OBBBA), a more foundational problem remains: the absence of timely, comprehensive, and comparable data on graduate school outcomes. 

Graduate students hold roughly halfof all federal student-loan debt, and evidence suggests that even with higher earnings, many face long-term loan debt that offsets the premiums associated with graduate degrees. Graduate programs at elite institutions are not immune to this trend either. Yet, our ability to assess and compare graduate program outcomes (completion rates, post-degree earnings, debt-to-income ratios) remains limited, hindering opportunities to create adequate solutions through policy and programs.

One dataset that includes a small sample of students is the Baccalaureate and Beyond survey, which tracks a representative cohort of bachelor's degree recipients 10 years post-undergraduate completion. It includes those who enrolled in graduate school during that time; however, the latest data is from the 2016 cohort. Much has changed since then (e.g. COVID-19). Additionally, this does not represent all of those who started a graduate degree. Consequently, we lack rigorous and timely data on a large portion of the graduate student population.

Another dataset that includes undergraduate and graduate students is the 2020 National Postsecondary Student Aid Study, which helps us answer questions about the types of financial aid various graduate students receive. AccessLex uses this data to analyze the percentage of students enrolled in various graduate programs by education attainment and undergraduate Pell status. One notable finding is that in 2019-2020, over half of students enrolled in Master's in Public Policy programs received a Pell Grant during their undergraduate studies. This information is useful, but it is still not the most timely, and it does not analyze completion outcomes or post-graduation outcomes. 

Some studies do mention graduation rates. For instance, Georgetown's Center for Education and Workforce study of Graduate programs shows that Ph.D. students in all graduate programs have the lowest completion rates, at 56% (though many obtain a different graduate degree in the process or transfer). That data is only in aggregate and does not allow us to see completion rates by individual colleges or by subgroups, such as those who received a Pell grant during their undergraduate tenure (a proxy for low-income students). Missing data on them in graduate school makes them invisible to policymakers.

The College Scorecard provides some information on earnings, but it suppresses post-graduation earnings data for a large share of graduate programs: approximately 77 percent of master's degree programs, about 70 percent of professional programs, and nearly 93 percent of doctoral programs. And while some colleges contribute voluntarily to the Post-Secondary Employment Outcomes Explorer from the U.S. Census Bureau, one study shows that the dataset covers only 16–19% of master's holders, 14–19% of doctoral program holders, and 10–13% of professional degree holders. These gaps limit our ability to create effective policy that supports positive student outcomes.

In comparison, we know much more at the undergraduate level. The College Scorecard and the Integrated Postsecondary Education Data System (IPEDS), both maintained by the National Center for Education Statistics (NCES), allow policymakers, institutions, and students to analyze undergraduate access, completion, and outcomes in detail.  One recent study used this data to rank colleges that both accept a high rate of Pell students and graduate them, a measure of success. This framework is also utilized in the new Carnegie Classifications. While IPEDS is not perfect and could be improved, many states rely on this data for policymaking.

Wouldn't it be great if we had data like this at the graduate level?

Under the OBBBA passed in July, schools will now have to collect earnings data on both undergraduate and graduate programs. Under this regulation, graduate-level programs must demonstrate that their graduates earn more than those with only a bachelor’s degree. While this is helpful, an analysis by the PEER Center shows that relatively few graduate programs are expected to fall below this earnings threshold (earn less than a bachelor’s degree recipient). It only accounts for earnings and does not take into account the debt students may incur to finance the program, particularly for low-income students, and the impending graduate loan caps set to take effect this upcoming summer. 

Another aspect to consider is that aggregate earnings gains may not account for disparities in pay by Gender or Race. Pay gaps by both gender and race are substantial among graduate degree holders across many fields. These aspects should be considered when measuring earnings gains at the program level. 

The administration has been more focused on data on admissions, particularly by race and gender, under an August Executive Order. However, new details about the order also include collecting data on graduation rates. This order is specifically targeted at selective colleges. Considering that the NCES has significantly fewer staff due to government layoffs and the speed at which the data needs to be produced, mistakes are not out of the question, both for institutions providing the data and for staff analyzing the data at the Department of Education. The data can also be misinterpreted. 

If we truly care about graduate access, completion, and return on investment, especially for students from under-resourced backgrounds, then policymakers must stop seeking outcomes without allocating the necessary resources.

Below are three potential recommendations for elevating data (and student success) higher on the agenda. 

Adopt legislation to mandate national data collection on graduate programs

Orders are not the way. Legislation is a better option. As with the undergraduate level, we need legislation that clearly mandates data collection on enrollment, completion, debt, earnings, and other outcomes for graduate programs, disaggregated by Pell status, parental education attainment, discipline, degree type, and race/ethnicity. For instance, the College Transparency Act (CTA) would enable data collection not only on undergraduate outcomes but also on graduate programs, making the data publicly available on an easy-to-use website. This data collection should continue on a regular basis. 

This legislation is not new. It was first introduced in 2017 and reintroduced in 2023, neither of which became law. In previous introductions, it has endured widespread support, not only amongst lawmakers, but also over 150 prominent organizations, including AccessLex Institute, Achieving the Dream, Braven, Center for First-generation Student Success, Jobs for the Future (JFF), Third Way, National Association for College Admission Counseling, The Education Trust, and many more.  

Although it has broad support, unfortunately, neither the Senate nor the House version left the Committee.

Involve States and Accreditors

States and Accreditors can also play a role. The Federation of American Scientists recently published an article suggesting that the National Center for Education Statistics establish State Cooperative Groups to collect this data instead of the Federal Government. Some states and Accreditors are already collecting more data on student and program outcomes that account for factors such as earnings and debt. The Western Association of Schools and Colleges Senior College and University Commission (WSCUC) is one such example. 

State policymakers can provide incentives and funding to integrate graduate student data into statewide longitudinal data systems, which, as of 2024, over 33 states already have. The Federal Government should also incentivize the collection of graduate school data by continuing to fund the Statewide Longitudinal Data Systems Grant Program, offered by the NCES, and making it a requirement for funding. Currently, many states are concerned about the sustainability of their Longitudinal Data systems due to funding issues. For states seeking to develop more sustainable models of these systems, consider this new resource from the Data Quality Campaign. 

Incentivize Institutional Reporting

Given that very few institutions report their data to the Post‐Secondary Employment Outcomes Explorer from the U.S. Census Bureau, the federal government, states, or philanthropic organizations can provide financial incentives for more institutions to contribute to this dataset, making it more representative of earnings. The Census Bureau should expand the data set to include other measures, such as earnings-to-debt ratios or net earnings after debt, as well as completion rates by institution and by subgroups (e.g., Pell status, family education attainment, race). 

If the Administration is serious about helping all students and families navigate the terrain of higher education, it would allocate the funds to have a healthy (full) staff of analysts at these federal agencies (Census Bureau, NCES, etc.) to help collect, analyze, and visualize the data. 

Final Word

We can no longer accept a system of graduate education that operates largely in the dark. For too long, we have tolerated students enrolling in advanced programs without the same level of transparency afforded to undergraduates, despite high debt levels and widely varying outcomes. 

If the nation cares about expanding graduate access, supporting low-income, first-generation, and other underresourced students, and ensuring that advanced degrees remain a viable path to economic mobility, then we must start with the fundamentals: better data, greater transparency, and aligned accountability.

Policymakers must act. Congress should pass robust transparency legislation (such as the College Transparency Act) and allocate adequate resources (funding) to equip NCES and other federal agencies to collect and publish graduate‐level data. Institutions, states, and Accrediting agencies should adopt transparent publishing of debt/earnings/completion metrics for graduate programs. Only then can we transform graduate education into a system that is accessible, successful, and grounded in evidence rather than assumptions.