College ROI by Earnings and Debt

When families ask whether a college is "worth it," they're rarely asking an abstract philosophical question. They're asking something concrete: if my kid borrows this much, will the paycheck on the other side make the loan feel small or feel like a weight? That's the question this analysis tries to answer with the only large public dataset that reports both numbers at the school level.

The honest framing matters here. Earnings is an annual figure. Debt is a cumulative total. Subtracting one from the other does not produce a true rate of return, and we do not pretend it does. What it produces is a directional signal: which schools send graduates into the working world earning far more each year than they owe in total, and which leave the two numbers uncomfortably close.

The Leaderboard: Earnings Far Above Debt

This is the CollegeHelpGuide analysis of U.S. Department of Education College Scorecard data, ranking schools with at least 500 students by the gap between median 10-year earnings and median federal debt at graduation. Every figure below comes directly from College Scorecard.¹

The list reads, at first glance, like a roll call of the most selective private universities in the country. MIT, Stanford, Harvard, Penn, Yale, Princeton, Duke, Columbia, and Dartmouth all appear, and they appear because their graduates earn six figures a decade after enrolling while borrowing relatively modest amounts of federal debt. Notably, the median federal debt at these elite privates is often lower than at far less prestigious schools, because generous need-based aid means many students borrow little or nothing through federal loans.

But look more carefully and a second story appears.

The Public Health-Science Schools Hiding in Plain Sight

Four of the top thirteen schools are public institutions, and they share a trait: they are health-science campuses. The University of Maryland, Baltimore ranks ninth with median earnings of $114,200 against $15,000 in debt. Oregon Health & Science University sits tenth at $114,900 in earnings. UT Health Science Center at San Antonio comes in eleventh with $111,500, and the University of Nebraska Medical Center ranks thirteenth at $108,500.

These are not household names the way the Ivy League is. They appear here for a structural reason: their student bodies are concentrated in medicine, nursing, pharmacy, dentistry, and allied health fields where ten-year earnings run high and federal borrowing at the undergraduate or entry level stays moderate. A graduate-heavy, clinically-focused campus produces an earnings figure that competes directly with Stanford and Harvard.

This is the part that gets lost in "best ROI college" lists built around prestige. A specialized public university with a heavy clinical mix can deliver an earnings-to-debt picture nearly as strong as the most selective privates, often at a fraction of the sticker price for in-state students. If you are weighing where the financial math lands best, the field mix of a school can matter as much as its name.

National Context: The Typical School Looks Nothing Like the Top

The leaderboard is the exciting part. The national picture is the sobering one.

Across all 3,972 schools that report both metrics, the average median 10-year earnings figure is $40,919, and the average median federal debt is $16,419.¹ The median school posts earnings of $38,700 against $14,906 in debt. In other words, the typical American college does not send graduates into the world earning $150,000. It sends them into the world earning somewhere in the high $30,000s, carrying around $15,000 in federal debt.

That gap between the headline schools and the average school is the single most important thing to take from this analysis. MIT's $138,832 earnings-minus-debt figure is more than five times the gap at a school posting national-average numbers. The leaderboard is real, but it describes a sliver of the system, not the experience most students will have.

It's also worth saying plainly: a $38,700 annual salary against $14,906 in total borrowing is not a bad outcome. A year of earnings comfortably exceeding total debt is the position most borrowers hope to reach. The danger zone isn't the average school. It's the corner of the market where earnings stall and the gap closes.

Public vs Private vs For-Profit

Sector tells you a lot about where a school is likely to land.

Private nonprofit schools post the highest average earnings at $49,149, pulled upward by the selective universities on the leaderboard. Public schools follow at $42,305, above the all-school average. For-profit institutions trail at $30,978, with average debt of $12,958.¹

The for-profit number deserves attention. These schools do not, on average, saddle students with the largest debt; their average debt is actually below the national figure. The problem is the other side of the ledger. When earnings land near $31,000, even a moderate debt load eats a larger share of what graduates take home. The earnings-minus-debt gap that looks comfortable at a public flagship gets thin fast when the salary side is suppressed.

This is why we treat the earnings-minus-debt gap as a comparison and not a verdict. Two schools can both report $15,000 in median debt and produce wildly different financial outcomes, because the earnings on the other side are doing most of the work.

What "ROI" Really Means Here

The phrase "return on investment" carries a precise financial meaning: the gain from an investment relative to its cost, usually expressed as a percentage or an internal rate of return over time. This analysis does not produce that. We want to be direct about it.

Median 10-year earnings is an annual figure measured roughly a decade after a student first enrolls. Median federal debt at graduation is a one-time cumulative balance. Subtracting the second from the first answers a useful but limited question: how does a single year of typical earnings compare to the total federal debt a graduate carries out the door? When earnings dwarf debt, the loan looks manageable against the income it helped produce. When the two numbers sit close together, repayment is a heavier lift.

What this figure cannot tell you:

• It does not account for private loans, parent loans, or out-of-pocket costs that never appear in the federal debt figure.
• It does not measure how earnings grow (or stall) over a full career.
• It does not isolate the school's contribution to earnings. A campus full of future doctors and engineers will report high earnings no matter how it teaches, because of who enrolls and what they study.
• It does not adjust for geography, where the same salary buys very different lives.

For a deeper look at the earnings side of this question across the whole system, see our companion study on the gender pay gap by college, and for the cost side, the college net price by state analysis. The three together describe the same trade-off from different angles.

The Honest Caveats

Three limitations shape how far you should push these numbers.

First, the earnings figure reflects who enrolls, what they study, and where they live, not the school's "value-add." A school that admits high-achieving students from advantaged backgrounds into high-paying fields will post strong earnings even if a comparable student would have earned just as much elsewhere. The data describes outcomes, not causation.

Second, the two metrics measure different things on different clocks. Annual earnings versus cumulative debt is an apples-to-oranges comparison made useful only by acknowledging exactly what each number is.

Third, averages across institution types blur enormous internal variation. A single school's reported median hides the gap between its highest- and lowest-earning programs. A business graduate and a fine-arts graduate from the same university can have very different financial lives, and the school-level median washes that out. If you're choosing a school, the question that matters most is the earnings-and-debt picture for your specific program, not the campus-wide median.

None of this means the data is useless. It means the data is a starting point. It can flag a school where earnings reliably outrun debt, and it can flag a corner of the market (low-earning programs at any sector) where the math gets risky. What it cannot do is replace the program-level homework every family should do before signing a loan.

How to Use This for Your Own Decision

If you take one thing from this analysis, let it be the comparison habit. Don't look at a single school's earnings number in isolation. Look at it against the debt students borrow, against the national averages above, and against schools with a similar field mix. A $45,000 earnings figure means something very different at a school where students borrow $10,000 than at one where they borrow $30,000.

Then narrow from the campus level to the program level. The whole-school median is the blurriest version of the truth you'll ever see. Your major, your borrowing, and your local job market will move you well off that average in one direction or the other.

For more on the cost half of the equation, our guide to the average cost of college per year breaks down where the money actually goes, and our list of the cheapest colleges in every state shows how much the debt side can shrink with the right choice. If you're still weighing the bigger question, is college worth it in 2026 walks through the trade-off without the prestige bias that warps most rankings.

Methodology

This analysis uses the U.S. Department of Education's College Scorecard, the federal dataset that publishes outcome data for institutions receiving Title IV financial aid.¹

For each school we pulled two fields: median earnings of students measured approximately ten years after entry, and median federal debt of students at the point of completion or graduation. The ranking metric is a simple subtraction, median 10-year earnings minus median federal debt, applied only to schools with at least 500 students to avoid distortion from very small programs.

The sample includes 3,972 schools that report both metrics. Schools missing either field are excluded, which removes some institutions from the analysis entirely. National averages and medians (average earnings $40,919, average debt $16,419, median earnings $38,700, median debt $14,906) and the sector averages (private nonprofit $49,149, public $42,305, for-profit $30,978, for-profit average debt $12,958) are computed across this same set of reporting schools.

The honest limitations: median earnings is an annual figure while median debt is a cumulative total, so the difference between them is a directional comparison and not a true ROI or internal rate of return. Federal debt excludes private loans, parent loans, and out-of-pocket spending. Earnings reflect who enrolls at a school, the mix of fields it offers, and the geography of where its graduates live and work, none of which the data isolates from the school's own contribution. Averages that span public, private, and for-profit institutions blur wide differences within each type and within each campus. Treat the figures as a well-sourced starting point for comparison, not a final verdict on any single school. For broader context on enrollment and cost patterns we draw on the National Center for Education Statistics,² and for the labor-market backdrop on earnings by education level we reference the U.S. Bureau of Labor Statistics.³

FAQ

Which college has the best ROI by earnings versus debt?

In this analysis of College Scorecard data, the Massachusetts Institute of Technology has the largest gap between median 10-year earnings ($153,600) and median federal debt ($14,768), at $138,832.¹ Stanford and Harvard follow. Keep in mind this is an earnings-minus-debt gap, a directional comparison, not a formal rate of return.

Are public colleges ever competitive with elite private schools on ROI?

Yes. Four public health-science campuses (University of Maryland, Baltimore; Oregon Health & Science University; UT Health Science Center at San Antonio; and University of Nebraska Medical Center) rank in the top thirteen, with median earnings between $108,500 and $114,900.¹ Their strong showing reflects a program mix concentrated in medicine and allied health.

Why do for-profit colleges rank so low?

Not because of high debt. For-profit schools in this data carry average debt of $12,958, below the national average. They rank low because average median earnings are $30,978, the lowest of any sector.¹ When earnings are suppressed, even a modest loan becomes a heavier burden.

Is "earnings minus debt" the same as return on investment?

No. Return on investment has a precise financial meaning that accounts for cost and gain over time. This metric subtracts a cumulative debt total from a single year of earnings. It's a useful comparison for judging whether a loan looks manageable against income, but it is not an internal rate of return.

What does the average college look like on this measure?

Far from the leaderboard. Across the 3,972 schools reporting both metrics, average median 10-year earnings are $40,919 and average median debt is $16,419.¹ The typical school sends graduates into work earning in the high $30,000s with roughly $15,000 in federal debt, not into six-figure salaries.

Should I pick a school based on its ROI ranking?

Use it as a starting point, not a verdict. School-level medians blend high-earning and low-earning programs together. Your major, your borrowing, and your local job market will move you off the campus average. Ask schools for program-level earnings and debt before deciding.

Related Articles

• Gender Pay Gap by College
• College Net Price by State
• Is College Worth It in 2026?
• Average Cost of College Per Year
• Cheapest Colleges in Every State

Footnotes

1. U.S. Department of Education. (2026). College Scorecard. National Center for Education Statistics. https://collegescorecard.ed.gov/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

2. National Center for Education Statistics. (2025). Digest of Education Statistics: Postsecondary Enrollment and Costs. U.S. Department of Education. https://nces.ed.gov/ ↩

3. U.S. Bureau of Labor Statistics. (2025). Education Pays: Earnings and Unemployment by Educational Attainment. U.S. Department of Labor. https://www.bls.gov/ ↩