Statistics can be so misleading. It is funny, though, how often at the moment you see one team had 60 per cent of the ball but still lost.

Neil Warnock, The Independent, November 10, 2012

The Institute for Higher Education Policy (IHEP) published a 2023 report applying a minimum standard for college financial value.¹ The benchmark is called Threshold 0. Walk through IHEP’s own example using their data and arithmetic. The school they describe passes. Apply complete accounting to the same school and its typical student still ends up behind a high school graduate in lifetime present value, even measured against California’s own high school baseline.

That gap between IHEP’s conclusion and a complete accounting runs through the entire report. By IHEP’s measure, 83 percent of institutions meet Threshold 0.

They say: “At the majority (83 percent) of institutions—representing 93 percent of undergraduates—students receive at least a minimum economic return on their investment. In other words, students’ typical earnings meet or exceed the Threshold 0 benchmark within 10 years of starting college … In these terms, nearly all public and private nonprofit institutions leave students better off financially in comparison to similar adults who did not pursue postsecondary education.”¹

The method behind that number does not support that conclusion.

IHEP also published Threshold 0 results for more than 4,000 colleges in its Equitable Value Explorer, which in May 2023 began breaking earnings out by gender and family income. The five gaps described here apply equally to the tool’s data.

What Threshold 0 Measures

IHEP, through the Postsecondary Value Commission it helped lead, designed Threshold 0 to answer a specific question: does a college’s typical student, graduate or not, earn enough to recoup the cost of their education within 10 years, compared with a high school graduate who never attended college?

The mechanics are straightforward. Take the median high school salary in the state where the institution is located. Add the annual repayment cost of the degree, amortized over ten years with loan interest (the 3.73% federal undergraduate rate in the report’s analysis, 2.75% in its worked example). If students’ median earnings ten years after they enter college clear that sum, the institution passes.

The example IHEP provides uses California. The state median high school salary is $28,297. A hypothetical institution with a net annual cost of $18,500, or $74,000 over four years, generates a repayment burden of $8,472 per year. Threshold 0 is $36,769. A college whose typical student earns $45,750 clears it. Pass.

The method is the right instinct.

Setting a minimum required return before recommending an investment is basic financial discipline.

But the conclusion collapses because of what the method leaves out.

What Threshold 0 Leaves Out

The first gap is the baseline itself. IHEP uses state-level high school salaries, not the national median. A degree clearing Threshold 0 in Mississippi does not necessarily clear it in Massachusetts. Graduates are not confined to the states where they attended college. Some move. The benchmark this book uses throughout, a national median wage for full-time workers aged 22 to 27 with only a high school diploma, sets a higher bar than California’s $28,297. Part of that difference is who gets counted: IHEP’s state figure includes high school graduates aged 22 to 40 with any earnings who are not enrolled in college, full time or not. Part is the year: IHEP’s figure is in 2022 dollars, and this book’s benchmark is in 2025 dollars.

The second gap is cost inflation. IHEP holds each year’s net price flat across the enrollment period. A student’s actual costs can rise from freshman year to senior year. The assumption overstates the economic case in the short term, precisely when it matters most for students who need the math to work.

This is where the flat-cost assumption hurts most. The students the Commission focused on, including students from low-income backgrounds with little financial cushion, are the students least able to absorb the short-term gap between costs paid now and earnings realized later. A flat-cost assumption does the most damage to the people it most needs to protect.

The third gap is opportunity cost. Threshold 0 measures earnings ten years after a student enters college. For a four-year degree finished on time, that leaves six years of actual work at most, and fewer for the many students who take longer. During those college years, a high school graduate working full time earned a salary that a full-time college student gave up, creating an opportunity cost that Threshold 0 ignores.

In this book’s model, a high school graduate working full time from age 18 earns $147,850 before taxes during the four years a typical college student is enrolled. After federal income and payroll taxes, that drops to roughly $127,000 in take-home pay. That $127,000 in actual take-home earnings does not appear in the Threshold 0 comparison.

Even if a college graduate meets Threshold 0 within ten years, they may not catch up to the cumulative earnings of a high school graduate who started working immediately. The structure systematically flatters the college outcome.

The fourth gap is taxes. IHEP measures gross earnings against a Threshold 0 designed to screen for economic benefit. But the roughly $227,000 in additional federal income taxes the median college graduate pays over a lifetime never reaches the graduate’s bank account. Gross earnings across a progressive tax system are not a defensible proxy for economic benefit. The accounting is inconsistent.

The fifth gap is the time value of money. Apart from student loan interest on the cost, Threshold 0 treats a dollar earned in year ten as equivalent to a dollar paid in year one. It is not. College costs are front-loaded. The earnings premium arrives later. Not applying a discount rate to those future earnings systematically inflates the apparent return. This design choice is the difference between measuring an investment and flattering it.

What the Numbers Show

The model used throughout this book applies all five corrections: a national high school salary baseline, college costs rising 5.6 percent a year, full opportunity costs starting at age 18, federal taxes at 2025 rates, and a 7.8 percent discount rate applied to lifetime earnings projected to age 65.

As of March 2026, the College Scorecard put median earnings four years after graduation at $134,794 for Harvard and $60,428 nationally for bachelor’s institutions. The New York Fed puts the median wage of full-time workers aged 22 to 27 with only a high school diploma at about $40,000. On those salaries, both college groups would clear IHEP’s Threshold 0, which itself uses earnings ten years after students enter college.

Now look at cumulative after-tax earnings through the first ten years, net of college costs at net price: Harvard’s own, and the national median for the median graduate. The high school graduate, who has been working since age 18, has accumulated $336,000. The median college graduate has accumulated $214,000. Harvard graduates, earning more than twice the median college salary, have accumulated $529,000.

Figure C-A · Cumulative after-tax earnings net of college costs, first ten years

Source · Author analysis · Earnings from College Scorecard · high-school median from the Federal Reserve Bank of New York.

After taxes and college costs, the high school graduate is more than $120,000 ahead of the median college graduate within the ten-year window IHEP uses. Harvard’s premium is real in raw cash terms, but it leans on the elite salary.

Net Present Value (NPV) at a lifetime horizon tells the same story. At the College Scorecard’s national median net price for bachelor’s institutions, $20,081, the median college graduate generates $467,000 in after-tax NPV. At full sticker, $402,000. The median high school graduate generates $484,000. A median Harvard graduate generates $1,037,000 at Harvard’s own net price ($798,000 at full sticker). For the median college graduate, the lifetime NPV never recovers.

The ten-year window IHEP uses tells a sharper version of the same story. Figure C-B puts it all together by deducting costs and taxes and calculating the NPV.

Even Harvard, with graduates earning more than double the median college salary and paying a net cost below the national median for bachelor’s institutions, only modestly outperforms the high school baseline within the IHEP window once time value enters the calculation.

Figure C-B · Net present value after taxes, costs, and time-value of money, first ten years

Source · Author NPV model · Same population as Figure C-A

The NPV of the high school graduate’s 10-year earnings is $225,000. The median college graduate sits at $100,000. Harvard reaches $281,000 over that same window, about $56,500 above the high school baseline after six years of higher earnings.

The median college graduate clears Threshold 0 yet sits about $125,000 behind the high school graduate in that same window, and even Harvard’s elite premium buys only a $56,500 margin. The threshold is not measuring what it claims to measure.

IHEP’s model says 83 percent of institutions pass. This model says the median institution fails.

Consider what those numbers mean for a student choosing between working after high school and attending a median-cost college.

Discounted at 7.8 percent, the high school graduate’s after-tax earnings over the first ten years after graduating at 18 are worth $225,000 today. The college graduate, starting work four years later, ends the same calendar window at $100,000 on the same basis, after paying for college.

Threshold 0 looks at year ten salaries and sees a pass. A family looking at the bank accounts sees something different.

The Massachusetts Finding

That gap is not unique to this analysis.

In December 2021, a College101 study by Stig Leschly and Yazmin Guzman, with Michael Itzkowitz, looked at degree-specific earnings for graduates of colleges in Massachusetts. The findings were direct: 55 percent of degree programs allowed graduates to earn more than non-college peers and recover costs within ten years. The other 45 percent did not. Nineteen percent of graduates, in 26 percent of the programs, are either worse off or need 20 years or more to break even.²

The Massachusetts study used a similar method to IHEP’s. It did not account for taxes or the time value of money. Correcting for those two factors, which this book’s model does, would push more programs into the failure column, not fewer.

IHEP’s national claim: 17 percent of institutions fail.

Massachusetts, without correcting for taxes or time value: 45 percent of programs fail. This book’s model, correcting for all five gaps: the median graduate falls behind high school at 79 percent of 1,679 bachelor’s institutions at full cost of attendance, and at 58 percent at each school’s own net price. The Georgetown University Center on Education and the Workforce (CEW) ROI rankings, re-cut in this book’s analysis against the CEW’s own figure for a high school graduate’s lifetime earnings: on a rough test, about two-thirds of the institutions in the CEW’s 2025 update fail to outperform it.³

These are not apples to apples. They span institutions and programs, state and national samples, graduates and all students who enrolled. But the direction is consistent. Each of these analyses finds a sizable share failing to deliver, and each of the others finds far more failures than IHEP’s 17 percent.

What the Statistic Misses

The line is about football. A team controlling possession for most of the match can still lose. The final score, not the time-on-ball, is what counts.

IHEP’s Threshold 0 is a possession metric. State-specific baselines rather than national ones. Flat costs. No opportunity cost. No taxes. No discount rate. Under those conditions, 83 percent of institutions pass. Change the conditions to match how investment returns are actually calculated, and the result shifts by thirty percentage points or more.

That is not a methodological quibble. It is what this book’s model shows once it counts the costs Threshold 0 leaves out. IHEP and the Commission behind Threshold 0 chose every one of those parameters, and the 83 percent pass rate is what those choices produce. Strip out the choices that flatter the result and the number changes. The headline travels. The report never states what the metric leaves out next to the number.

---

About this analysis

This critique appears as Appendix C (IHEP, Rising Above the Threshold) in We Need To Talk About Higher Education by Leon Shivamber.

Get the book →   Read the argument in full →   Run your own numbers →

---

Notes

1. Dancy, Kim, Garcia-Kendrick, Genevieve, and Cheng, Diane. “Rising Above the Threshold: How Expansions in Financial Aid Can Increase the Equitable Delivery of Postsecondary Value for More Students.” IHEP.org. Institute for Higher Education Policy, June 2023. IHEP analysis applying the Postsecondary Value Commission’s Threshold 0 to 2,921 institutions and modeling how doubling the Pell Grant and free college programs would change how many meet it.
2. Leschly, Stig, Guzman, Yazmin, and Itzkowitz, Michael. “Degree-Specific Earnings Outcomes of Graduates From Colleges in Massachusetts.” College101, December 2021. Massachusetts-specific analysis of degree-level earnings outcomes showing significant variation across programs within institutions.
3. Cheah, Ban, Van Der Werf, Martin, Morris, Catherine, and Strohl, Jeff. “Ranking 4,600 Colleges by ROI (2025).” Georgetown University Center on Education and the Workforce, 2025. Carnevale, Anthony P., Cheah, Ban, and Wenzinger, Emma. “The College Payoff: More Education Doesn’t Always Mean More Earnings.” Georgetown University Center on Education and the Workforce, 2021. The College Payoff puts a high school graduate’s median lifetime earnings at $1.6 million. This book’s re-cut sets each institution’s 40-year return in the 2025 data tool (2021-22 data) against that $1.6 million: 2,982 of 4,476 institutions (66.6%) fall at or below it. The $1.6 million is for a steady full-time, full-year career from age 25 to 64, while the data tool uses former students’ median earnings and holds each college’s earnings flat after year 10, so treat the share as a rough test.

Related critiques

This is one of four close readings of the major college ROI studies. See the others, then read the audit that weighs all four together and the plain-language case beneath them.

• Georgetown’s College Payoff, examined.
• Georgetown’s ROI ranking, examined.
• The Fed’s college ROI math, examined.
• All four studies, audited together: Auditing the Numbers That Say College Pays.
• The plain-language version: College Is a Bet.
• The same problem priced against the alternative: Compared to What?
• The full argument: Is College Worth It?