Real ROI of a College Degree: NPV Analysis of 40 Majors (2026, Inflation-Adjusted)

Most "is college worth it" calculators answer the wrong question. They show you that a college graduate earns roughly $1.2 million more over a lifetime than a high school graduate — the figure popularized by Georgetown's Center on Education and the Workforce. That number is technically accurate. It is also financially misleading.

The problem is that it adds up dollars across a 45-year career as if a dollar in 2065 is worth the same as a dollar today. It is not. Inflation, opportunity cost, and the time value of money mean that a future dollar is worth meaningfully less than a present dollar — and any honest answer to "is this major worth the cost" has to account for that.

This is what corporate finance has done for decades when evaluating capital projects. Every major company assessing whether to build a factory, acquire a competitor, or fund a research program runs a Net Present Value calculation: it discounts future cash flows back to today's dollars at a chosen rate, sums them, subtracts the upfront investment, and accepts the project only if the result is positive. We are going to apply that same machinery to the household decision of paying for a college degree.

What This Analysis Does That Others Don't

The Federal Reserve Bank of New York has been computing the IRR of a college degree since 2014. Their most recent updates put the bachelor's-degree IRR at approximately 14% on average, with major-level variation from roughly 9% (education, leisure and hospitality) to 21% (engineering, math and computers). The Hamilton Project at the Brookings Institution, in their 2014 analysis "Major Decisions: What Graduates Earn Over Their Lifetimes," computed present-value lifetime earnings by major using a 3% real discount rate, finding engineering and math majors at roughly $3.4 million PV against early-childhood education at roughly $800,000.

What we are doing here is extending that academic methodology with three changes that matter to a parent making a real decision in 2026:

1. Three discount-rate scenarios. The 3% rate Hamilton Project used in 2014 reflected a different macroeconomic environment. In 2026, with 10-year Treasury yields around 4.3% and persistent inflation in the 2.5–3.5% range, a more realistic real discount rate for a household is closer to 6%, with sensitivity at 4% and 8%. We show all three.
2. Cost-tier sensitivity. A petroleum engineering degree at a public in-state university is a fundamentally different financial product than the same degree at a private nonprofit. We compute NPV at three cost tiers using NCES IPEDS net-price data.
3. Both NPV and IRR, side by side. NPV tells you the dollar value created. IRR tells you the percentage return. Parents intuitively understand the second one because it is comparable to investment returns — you can directly compare a major's IRR to the historical 7% real return on the S&P 500. We show both.

The Methodology in Plain English

If you have ever looked at a mortgage amortization schedule, you already understand most of what is happening here. Every future dollar gets multiplied by a discount factor that shrinks the further out it sits. A dollar one year from now at a 6% discount rate is worth about 94 cents today; a dollar 30 years out is worth about 17 cents. We sum those discounted dollars across the full career arc.

The cash flows we are summing have four components:

• Cost during the four college years (years 0–3, ages 18–21). The student pays net-of-aid tuition and earns a small amount from part-time work. We use NCES IPEDS data, which puts net price after average grant aid at roughly $23,000 per year for public in-state, $42,000 for public out-of-state, and $58,000 for private nonprofit four-year institutions in 2026 dollars.
• College-graduate earnings (years 4–44, ages 22–62). Starting salary, mid-career salary at age 40, and modest growth through the age-50 peak, drawn from BLS Occupational Employment and Wage Statistics, NACE Salary Survey 2025, and Census ACS Field of Bachelor's Degree tables.
• Counterfactual earnings (years 0–44). What the same person would have earned over the same 45 years if they had skipped college and entered the workforce at 18. We use BLS Education Pays median earnings for high-school-only workers, growing at 1.3% real per year. This is the opportunity cost most calculators ignore entirely.
• The discount factor. Every cash flow above gets divided by (1 + r)^t, where r is the real discount rate and t is the year.

The Net Present Value is the present-value sum of college-graduate cash flows minus the present-value sum of counterfactual cash flows. If that number is positive, the degree creates household value at the chosen discount rate. If negative, it destroys household value — the family would have been financially better off skipping that specific major-and-school combination.

The Internal Rate of Return is the discount rate at which NPV equals zero. It is the breakeven hurdle rate. If a major has an IRR of 12%, it creates value at any discount rate below 12% and destroys value above. Conceptually it is the "interest rate" the degree pays on the investment of foregone wages and tuition.

The Results: NPV and IRR for 40 College Majors

Below is the full table, sorted by NPV at public in-state cost and a 4% real discount rate. NPV is the present-value lifetime premium of the degree over the high-school-only counterfactual. IRR is the discount rate at which that premium equals zero.

Sources: BLS Occupational Employment and Wage Statistics 2024 (inflated to 2026 dollars), Census American Community Survey Field of Bachelor's Degree tables, NACE Salary Survey 2025, NCES IPEDS net-price data. Methodology described in detail above. Discount rates are real (inflation-adjusted). Time horizon: ages 18 to 63.

Five Findings That Should Change How Parents Think About This

1. The "$1 million premium" headline overstates real economic value by roughly half.

Across the 40 majors in our sample, the average undiscounted lifetime earnings premium is approximately $1.1 million, broadly consistent with Georgetown CEW's published number. But at a 6% real discount rate, the average present-value premium drops to about $410,000. The number you see cited in news headlines is roughly twice the number that actually matters to your household balance sheet.

This is not a quirk of our model. The Federal Reserve Bank of New York's framing of college as a financial investment, and the Hamilton Project's 2014 NPV-by-major analysis, both arrive at similar discount-adjusted figures. The undiscounted nominal number is the version that gets repeated because it is bigger and easier to communicate — not because it is more accurate.

2. At private-school cost, more than 40% of majors destroy household value.

At public in-state cost and a 6% discount rate, 14 of 40 majors (35%) have negative NPV. At public out-of-state cost, that rises to 15 of 40 (38%). At private nonprofit cost, 17 of 40 (43%) destroy household value in present-value terms. The same major can be a strong financial decision at one price and a poor one at another. A psychology degree at a state university with strong financial aid is a different financial product than a psychology degree at a $58,000-per-year private school with minimal aid.

This is the single most important practical insight from the analysis: the major matters, but the cost matters just as much. Two students choosing the same major can have NPV outcomes that differ by $300,000 or more depending purely on where they enroll. For families weighing school choice, this is the lever that has the largest financial leverage in the entire decision.

3. Engineering majors return roughly the same as long-run equity markets — with less risk.

Petroleum engineering, software engineering, computer engineering, and computer science all show IRRs in the 20–25% range at public in-state cost. The historical real return on the S&P 500 over long horizons is approximately 7%. A petroleum engineering degree, financially speaking, is roughly equivalent to investing the same capital in an asset that has returned 25% real per year for 45 years. There is no equity index that has done that.

The risk profile is also different. The earnings curves underlying these IRRs are far more stable than equity returns: they reflect median earnings of working professionals, not market prices. The downside scenarios are job loss or career change, not a 50% drawdown in a single year. Looked at as a financial product, a STEM degree at a public university is one of the highest risk-adjusted returns available to a U.S. household.

4. The bottom of the table is mostly arts and education — but the reasons differ.

Theater, fine arts, and music show negative NPV across every cost tier and every discount rate. The economic reason is simple: median earnings in these fields are roughly equal to or below the high-school-only counterfactual after accounting for four years of foregone wages and tuition. The earnings premium does not exist on average. Individuals in these fields can earn far more, but the median outcome does not justify the investment in present-value terms.

Elementary and secondary education show small negative NPVs at public in-state cost and larger negative NPVs at private cost. The reason here is different: the earnings premium does exist, but it is small (roughly $400–$700 per month over the high-school counterfactual) and the four years of foregone wages are not recovered until late in the career. At a 6% discount rate, those late earnings get heavily discounted. This is not a statement about the social value of teaching — it is a statement about the financial mechanics of paying tuition for a low-earning career.

The practical implication is not "do not study these majors." It is: do not study these majors at high-cost institutions while taking on debt. A psychology or elementary-education degree from a state university with substantial financial aid can be neutral or positive NPV. The same degree at a private school at sticker price is reliably negative.

5. The IRR-to-equity-return comparison is the most useful frame for parents.

Parents understand investment returns. They have 401(k)s and mutual fund statements. The framing that consistently lands in our reader research is not "this major has $300,000 NPV" — it is "this major returns 12% real, against the 7% historical real return on the stock market." That comparison reframes the entire question. A 12% IRR major is not just "worth it." It is, financially, a better long-run investment than putting the same capital in equities.

By the same logic, a 4% IRR major is a worse investment than buying a Treasury bond. A negative-IRR major is a worse investment than holding cash. This is a stark frame, and it is meant to be: when families are weighing $200,000 of borrowed money against the alternative of investing that money in literally anything else, the IRR comparison is the honest answer.

What Changes the Picture

Financial aid and scholarships

Every dollar of grant aid reduces the year-zero cost line, which has an outsized effect on NPV because it is being discounted by the smallest factor. A $20,000-per-year merit scholarship at a private school can convert a negative-NPV humanities degree into a positive-NPV one without changing any other input.

Time-to-degree

The model above assumes a four-year completion. Each additional year extends the cost-and-foregone-wages window by another roughly $80,000 in present value at a public university and $115,000 at a private one. Students who take five or six years to graduate — which is now the majority outcome at four-year institutions — see meaningfully lower NPV. We cover this dynamic in detail in our analysis of what the wrong major actually costs.

Graduate school

The 45-year horizon assumes the student begins earning at age 22. Medicine, law, dentistry, and pharmacy require additional years of school and additional cost. Some of those advanced degrees have very high IRRs once computed across the full timeline; others do not.

Student debt and interest costs

The numbers above assume tuition is paid out of household savings rather than borrowed. Adding student loan interest to the cash flow waterfall reduces NPV further. Every dollar borrowed at a 7% federal rate, repaid over 10–25 years, has an interest-cost overhead of roughly 30–60% of principal in present-value terms. A psychology major borrowing $80,000 at private-school cost is not paying $80,000 — they are paying closer to $115,000 in present-value terms once interest is added.

Major-to-career match

Our model assumes graduates work in fields aligned with their majors. In practice, only about 27% of college graduates work in jobs directly tied to their undergraduate major, according to Census ACS data. A finance major working as a barista earns barista wages, not finance wages. The major-level NPV figures above are therefore best-case averages that assume the student successfully enters and remains in their field.

The Three Practical Decisions This Analysis Should Inform

Decision 1: Major selection

If your student is genuinely undecided between two or three majors, the NPV and IRR figures should be one input — not the only input, but a substantial one. A 10-percentage-point difference in IRR between two majors of similar interest to the student translates to hundreds of thousands of dollars over a career. Our framework for working through this decision is in how to choose a college major.

Decision 2: School-cost tier

The same major delivers radically different NPV at different cost tiers. The honest question is not "should we send our daughter to her dream school" but "is the additional NPV destroyed by attending the higher-cost school worth the qualitative benefits of attending it." For some majors and some schools, that calculus works. For most, the public in-state option creates substantially more financial value, and the qualitative benefits do not close the gap. The starting point for thinking about this is our analysis of community college versus university outcomes.

Decision 3: College versus the alternative

For students who would land in the bottom third of our table — arts, lower-paying education, some humanities — the financial case for a four-year degree at any cost above public in-state is genuinely weak. The alternative paths worth comparing are skilled trades (high IRR, very low cost) and direct workforce entry with later education. Our cluster on trade school versus college applies this same NPV-style framing to the trades comparison.

Limitations and What This Analysis Cannot Tell You

An NPV analysis treats a college degree purely as a financial investment. It does not capture the qualitative dimensions of college: the network, the personal development, the credential signal, the campus experience, the access to a different social class. Some of these have economic value that is real but hard to quantify. A degree from a top-25 institution opens doors that a degree from a regional public university does not, in ways that occasionally show up in lifetime earnings and often do not. The model above does not adjust for institutional prestige.

The model also assumes the median outcome. Real students live on the distribution, not at the median. A computer science graduate who lands at a top-tier tech firm earns 2–3x the figures we used; a computer science graduate who never finds work in the field may earn less than the high-school counterfactual. The IRR figures above describe the median path. Risk-adjusted analysis — what NPV looks like across the 25th, 50th, and 75th percentile of outcomes — is the natural next step, and one we will tackle in a follow-up.

None of these limitations invalidate the analysis. They just mean it should be one input into the decision, not the only one. A negative-NPV major can still be the right choice for a student passionate about it, particularly if cost can be controlled. A positive-NPV major can still be the wrong choice for a student who will be miserable doing the work. The numbers tell you what the financial price of those non-financial choices is. They do not tell you whether to pay it.

How to Use These Numbers in Your Family's Decision

If you are using this analysis to inform a real decision, three suggestions:

1. Start with your student's likely major or 2–3 candidates. Not the major you wish they would pick. The major they will actually study and complete. Look up that row in the table and note the NPV and IRR at your expected cost tier.
2. Compare to the financial alternative your family would otherwise pursue. Investing the tuition money? Buying a different house? Funding a younger sibling's degree? The IRR figure is the threshold the alternative needs to clear for the alternative to be financially better.
3. Stress-test the cost assumption. The biggest variable you control is the cost tier. Run the same major at public in-state, public out-of-state, and private. The spread between those three is often $300–$500K of NPV. That spread is the financial price of the school choice.

For a structured way to walk through this with your student, our free college planning checklist includes a worksheet for the financial-decision side of the major-selection process.