Measuring the Return on Investment of Higher Education: Breaking Down the Complexity

The most recent iteration of Third Way’s analysis finds that a typical student at 55% of four-year institutions will recoup the costs of their enrollment within five years of graduating, while 5% of institutions generate a negative ROI, leaving the typical student unable to ever recoup the costs of their education.

Third Way also provides estimates for ROI by degree program, reporting that 41% of bachelor’s programs provide a return within five years, while 10% of bachelor’s programs fail to provide a positive ROI.

The Foundation for Research on Equal Opportunity’s ROI Calculator (2021)

FREOPP adds additional variables and considerations into its calculation of ROI, which focuses solely on program-level outcomes, concluding that a student’s program of study can matter as much or more than their institution in determining ROI. FREOPP’s model calculates the difference between estimated lifetime earnings for a graduate of a postsecondary program and the net cost of college (tuition, fees, and supplies after grants and scholarships) together with what students might have expected to earn otherwise, after taking opportunity costs, demographics, and student characteristics into consideration. FREOPP then makes additional adjustments to better assess variations in the risk associated with different degree programs and institutions, typically reflected in completion rates (e.g., dropping out or not completing a degree on time).

FREOPP’s analysis finds that 72% of programs at bachelor’s degree-granting institutions generate a positive ROI for the typical graduate.

The Bipartisan Policy Center’s College ROI Calculator (2022)

BPC provides three versions of its ROI model. Like FREOPP, BPC estimates ROI for students by assessing their total lifelong financial gains from attending a postsecondary institution relative to the costs incurred by attending. (Unlike FREOPP, BPC looks only at institution-level outcomes.)

To calculate the ROI of the median student at an institution, BPC’s baseline model subtracts the net price of attendance from the estimated lifetime college earnings premium (the positive earnings bump from attending college for the median student compared to counterfactual earnings). Unlike FREOPP, BPC calculates counterfactual earnings based on median earnings for high school graduates, rather than estimating expected counterfactual earnings based on demographics and other characteristics. Building on this baseline analysis, BPC’s intermediate model then incorporates discount rates, selection adjustments (i.e., accounting for the portion of the college earnings premium that actually comes from family resources and networks), and state-specific opportunity costs. The full model further adjusts for public subsidies for higher education and the negative effects of labor market discrimination (i.e., accounting for the earnings gaps that women and people of color experience).

When using the full model, BPC’s analysis finds that 97.6% of public four-year institutions, 92.0% of nonprofit four-year institutions, and 51.7% of for-profit four-year institutions provide a positive ROI.

Key Findings from Models

These models broadly demonstrate that most four-year institutions of higher education provide value for the typical student, with the public sector most likely—BPC (97.6% of institutions), Third Way (99.4% of institutions), and FREOPP (76% of programs)—to generate a positive return, as seen in Table 1. Significantly smaller shares of institutions and programs in the for-profit sector generated a positive ROI for students in each model.

Although the ROI models generally show similar trends, differences among their findings underscore the importance of the variables included. BPC’s model, for example, shows a significantly lower percentage of for-profit institutions providing a positive ROI compared to Third Way’s approach (51.7% versus 77.4%). At the program level, meanwhile, Third Way finds a higher percentage of all bachelor’s programs providing a positive ROI compared to FREOPP (90% compared to 72%).

The three versions of BPC’s model further illustrate this point. When the baseline model is used, public (99.8%) and private nonprofit (97.4%) four-year institutions fare very well. When accounting for discount rates and selection factors, however, the percentage of schools with positive ROI drops across the board, especially at private nonprofit (79.2%) and private for-profit institutions (16.7%). Then, with additional adjustments in the full model, estimated ROI improves substantially for private nonprofits (92.0%) and private for-profits (51.7%). The decline in ROI in the intermediate model stems, in part, from adjustments that account for the opportunity cost of higher education (through application of a discount rate) and the fact that the college earnings premium cannot be entirely attributed to students’ experience of higher education, as opposed to family connections or students’ own preexisting abilities. Consideration of labor market discrimination in the full model, conversely, produces significantly stronger estimates of institutional performance.

Integrating additional variables into ROI calculations carries both advantages and risks. Factoring in estimates for opportunity costs, demographic inputs, and projected lifetime earnings can provide a fuller picture of the costs and benefits of postsecondary education. It also introduces additional uncertainty into calculations, however, because these estimates and projections rest on assumptions that may not hold equally true for each institution or student body.

Equity and Economic Mobility Considerations when Measuring ROI

Researchers note that it is important to account for economic mobility in postsecondary ROI, though they propose different approaches for doing so. As discussed above, FREOPP and BPC make important adjustments for demographic factors and selection bias when calculating ROI. BPC’s full model adjusts the results to offset the impact of labor market discrimination faced by women and underrepresented minorities. FREOPP’s model similarly accounts for demographic characteristics of students and local labor market impacts.

Conversely, Third Way produces an Economic Mobility Index (EMI) that focuses on measuring institutions’ ability to uplift low-income students. Building on Third Way’s earnings premium ROI model, the Postsecondary Value Commission (PSV) further developed a framework to assess how well institutions serve students of color and students from low-income backgrounds. In addition to measuring the earnings premium for different student populations, PSV’s approach also considers whether an institution enables disaggregated student populations to achieve median earnings in their field of study and reach the fourth income quartile.

Conclusion

Insights from these models demonstrate the larger challenge faced by policymakers: ROI is more than a simple calculation; it is how you measure value that informs decisions based on returns on postsecondary education. This is influenced by what information and metrics matter most to the research organization (e.g., what outcomes they are trying to observe). ROI estimates are also constrained by what data and information are available.

As such, users should remain cautious in putting too much stock in any one ROI measurement to definitively predict student outcomes, and should instead use these models to enhance a more comprehensive decision-making process. Even the researchers behind ROI models advise that ROI calculations may be too difficult and impractical to use directly for accountability. Instead, the results can provide additional transparency and inform discussions on students’ outcomes and institutional accountability.