We use the following variables from the US Department of Education’s College Scorecard:

• students’ median earnings 6 years, 8 years, and 10 years after they enroll at an institution
• the average net price of the institution
• the predominant credential awarded at the institution

The predominant credential awarded at an institution is important because we use it to assume how long a student attended a particular institution. If the institution predominantly grants associate’s degrees, then we assume the student attended for three years, whereas if the institution predominantly awards bachelor’s degrees, then we assume the student attended for five years. (Three and five years are roughly the national average time to completion of associate’s degrees and bachelor’s degrees, respectively.)

We calculate ROI at intervals of 10, 15, 20, 30, and 40 years following initial enrollment at an institution. The ROI is the cumulative sum of earnings minus the total out-of-pocket costs based on the average net price.

For example, we calculate the 10-year ROI for a predominantly bachelor’s degree-granting institution as follows:

10-year ROI = year 6 earnings + year 7 earnings + year 8 earnings + year 9 earnings + year 10 earnings – average net price x 5

Because the College Scorecard doesn’t list earnings for years 7 and 9, we estimate them using the average change in earnings between years 6 and 8 and years 8 and 10, respectively.

For horizons longer than 10 years, we assume that real earnings continue to be the same as in year 10 (i.e., we assume no growth in earnings after year 10). This is a very conservative approach, and because earnings actually tend to grow with age, our estimates of ROI can be assumed to be at the lower end of possible outcomes.

According to the College Scorecard, the average net price is the net amount paid to attend a particular institution by students who are receiving federal financial aid (grants, such as Pell Grants, loans, and work study).

The average net price is the average out-of-pocket cost paid by full-time, first-time students who receive federal financial aid to attend one year of school. Institutions calculate the average net price by adding the advertised price for tuition, fees, books, supplies, and the average living costs at the school (on-campus, off-campus not with family, and off-campus with family) and subtracting the average grant and/or scholarship aid from federal and non-federal sources (e.g., Pell Grants, school-based grants, or merit scholarships). A negative average net price indicates that average grant/scholarship aid for students receiving federal aid exceeded the total cost of attendance at an institution. The share of students at a particular institution who receive federal aid is available in the College Scorecard.

For public institutions, the average net price reflects the average out-of-pocket cost for in-state students only. Average costs for out-of-state students attending public institutions are typically higher.

We do not include debt in ROI calculations because the relevant variables, such as current and future inflation rates and their effects on monthly debt payments, make it too difficult to accurately address with a single adjustment.

This reflects the lower cost of obtaining certificates and associate’s degrees, both in terms of average net price and time enrolled in a college or other learning institution. We assume a person earning a certificate will spend one year at a postsecondary institution, and a person earning an associate’s degree will spend three years in college. Since earning a bachelor’s degree takes longer (we assume five years), the certificate and associate’s degree holders will have more years of work and earnings in the first 10–20 years after enrolling. However, the earnings of people with certificates or associate’s degrees are lower on average than the earnings for those with bachelor’s degrees. As a result, the lifetime earnings of those with bachelor’s degrees will almost always catch up to those with certificates and associate’s degrees and eventually surpass them.

ROI can be an important factor in deciding what institution to attend, but it is only one of many factors that prospective students should consider. Students also will want to think about whether an institution offers the program of study that they want, for example, as well as the institution’s cost, location, instructional delivery method (in-person, online, or hybrid), and whether they are likely to be admitted.

Because many institutions share general characteristics and offer similar programs, ROI can make the difference when a prospective student considers all other factors to be roughly equal. In that case, the student may choose to enroll at the institution with the higher ROI.

Note that the ROI reported here represents the average for all students at a particular institution. Some undergraduate programs of study, such as the sciences or engineering, generally result in higher earnings than other programs of study, such as education or social work, even at the same institution. Furthermore, ROI varies from person to person, even among graduates of the same program and institution, due to individual differences in out-of-pocket costs, time enrolled, and post-enrollment earnings. See our report on the returns of college majors for more detail.

We intend to update the data tool annually each January if the relevant data in the College Scorecard have been updated during the previous year. The Department of Education does not have a set timetable for releasing new data, so the College Scorecard may be updated more than once a year or less than once a year. Also, updates are sometimes partial rather than comprehensive.

Yes. Users can export the data from specific rows of interest to them.

No. The data in the College Scorecard cover only students receiving some form of federal financial aid. Therefore, the ROI at institutions with small shares of students who receive federal aid may not accurately represent average student outcomes. Approximately 72 percent of all students get some form of financial aid, but only 55 percent of all students receive federal financial aid.

Although the Scorecard data are limited to federal financial aid recipients, prior research conducted by the US Department of the Treasury shows that the median earnings of financial aid recipients are representative of the median earnings of all enrollees at most colleges.

The vast majority of institutions are included, as long as the College Scorecard was able to gather data about them. Some very small institutions are not included in the College Scorecard for privacy reasons—their cohorts of students are so small that releasing earnings data might violate the privacy of individual students.

Additionally, state public university systems are not consistent about how they report data for individual campuses. Some state universities report data only for the entire institution, including branch campuses. Pennsylvania State University does this for its more than 20 campuses, for example; calculating separate ROI metrics for each Penn State campus is not possible for this reason. Other university systems—such as the University of California, California State University, and the University of Wisconsin—report data separately for each of their campuses.

In order to calculate an institution’s ROI, the following pieces of information about it must be included in the College Scorecard: three years of earnings data (for years 6, 8, and 10) and the average net price.

If any of this information for a particular institution is missing from the College Scorecard, the ROI for that institution cannot be computed.

It is possible that the information needed may be missing from the Scorecard due to privacy considerations. College Scorecard documentation states that data are not released if the number of people in a cohort is so small that information about individuals might be personally identifiable, thereby potentially violating the privacy of those in the cohort. Also, the US Department of Education does not report the results for an institution when the addition of “statistical noise” produces differences between the actual and privacy-protected earnings levels. The College Scorecard routinely introduces statistical noise in order to protect student privacy.

No. The time spans used in institutional ROI calculations begin with the year of enrollment at an institution, not the year of completion. This means that earnings of students who have not completed their degree are also included. To the extent that the data are representative of all students attending the institution, the risk of noncompletion is baked into the ROI computations. For this reason, institutions with low completion rates are more likely to have a low ROI. This is because students who never complete their degree generally have lower earnings potential, lowering the overall ROI of the institutions they attended.

On the other hand, institutions that have a large number of graduates who go on to attain higher degrees (such as those who start at two-year colleges but earn bachelor’s degrees, or those who earn bachelor’s degrees but go on to earn master’s degrees, doctoral degrees, and professional degrees) can see a boost in their ROI because of the additional earnings power that recipients of higher degrees experience in the labor market.

No. The data for each institution are the combined results from the entering classes of two consecutive years. For example, in the July 2024 release of the “Most Recent Cohort” data, the earnings 10 years after enrollment are for students who entered an institution in the 2009–10 and 2010–11 academic years. The earnings six years after enrollment are for students who entered a particular institution in the 2013–14 and 2014–15 academic years, and the earnings eight years after enrollment are for students who entered in the 2011–12 and 2012–13 academic years. The earnings for all of these cohorts correspond to the calendar years 2020 and 2021.

Because two years of data are involved, the ROI does not reflect the outcomes for a single year of students.

While the College Scorecard does not explain why it uses information about students from two years to report one year of data, it is a common practice in statistics and data gathering. Combining the results from two years tends to balance the results, lessening the chance of outlier results from a single year of data.

A single institution can award multiple types of credentials: certificates, associate’s degrees, bachelor’s degrees, master’s degrees, and doctoral degrees. The predominant credential indicates which of these were awarded by an institution in the greatest abundance in the year corresponding to the most recent Integrated Postsecondary Education Data System (IPEDS) data used by the College Scorecard. This categorization can affect the computed ROI over different years if the institution’s predominant credential changes over time.

In computing the ROI for an institution, we use the predominant credential to assume how long a student was enrolled. If the institution predominantly grants associate’s degrees, then we assume the student was enrolled for three years. If the institution predominantly awards bachelor’s degrees, then we assume the student was enrolled for five years.

About 40 percent of institutions in the College Scorecard predominantly grant certificates. These institutions are further classified by the length of time required to earn the relevant certificate: less than two years, two years, or four years. We assume that students were enrolled for one year if the data suggest that the institution requires less than two years of study. If a predominantly certificate-granting institution is classified as requiring two years of study, we assume the student was enrolled for three years, the same amount of time as a student enrolled in a predominantly associate’s degree-granting institution. In the rare cases where a predominantly certificate-granting institution is classified as requiring four years of study, we assume the student was enrolled for five years, the same amount of time as a student enrolled in a predominantly bachelor’s degree-granting institution.

No. In earlier reports, we used a discount rate of 2 percent on future cash flows to account for the fact that money today is more valuable than money in the future. We called the result the “net present value.” For the current version of this data tool, we use a discount rate of zero percent. In other words, we do not discount future cash flows.

In retrospect, we decided that the concept of net present value added an unnecessary source of confusion to our calculations. A 2 percent discounting rate fails to factor in that different people will have different views on how heavily the future value of money should be weighed, depending on their individual circumstances.

By doing away with the discount rate, we are attempting to eliminate subjectivity by treating future earnings the same as current earnings.

Many factors can influence the ROI of an institution. We analyzed the importance of five institution-level variables that research indicated might be closely associated with ROI:

• acceptance rates
• graduation rates
• percent of degrees offered in STEM fields
• the share of undergraduates who receive any federal aid
• the share of undergraduates who receive Pell Grants

Institutions are not monolithic: A public institution that predominantly awards bachelor’s degrees, for example, is categorically different from a for-profit institution that predominantly awards certificates. Therefore, in addition to aggregate-level correlations, we looked at correlations between the five institution-level variables listed above by predominant credential awarded and institutional control.

Our analysis revealed the following:

• Graduation rates are most strongly correlated with ROI measures at institutions that predominantly award bachelor’s degrees, with the notable exception of those that are for-profit. Private, nonprofit institutions that predominantly grant associate’s degrees also exhibit a moderate correlation between graduation rates and ROI.
• The share of STEM degrees awarded is moderately correlated with ROI at predominantly bachelor’s degree-granting institutions (except the for-profit institutions), but otherwise is not strongly correlated with ROI for other institution types.
• The share of students who received any federal aid is not strongly correlated with ROI, except for a 0.58 correlation with 10-year ROI at public institutions that predominantly award certificates.
• Across institution types, acceptance rates are only modestly negatively correlated with ROI, though the correlations are highest for predominantly associate’s degree-granting, private, nonprofit institutions.
• Pell Grant recipient rates are negatively (if a bit unevenly) correlated with ROI, including moderately strong correlations with 10-year earnings of -0.50 and -0.66, respectively, at private, nonprofit institutions that predominantly award associate’s and bachelor’s degrees.