Aggregate statistics can sometimes mask important information.
Ben Bernanke, speech on economic measurement, August 6, 2012
The Georgetown University Center on Education and the Workforce published a report in 2011 that, in my view, changed how a generation thought about college.1
Its central finding is simple: by the report’s own median figures, a bachelor’s degree delivers a lifetime earnings premium of $964,000 over a high school diploma.1
The number is real.
The method behind it is not wrong for what it measures. It is incomplete for the choice a family faces.
Incomplete data, applied to personal decisions, causes serious harm.
What the CEW Measured
The Georgetown University Center on Education and the Workforce (CEW) is a legitimate research institution. Its data sources are credible. Its analysts are serious. The problem is not bad faith. It is four methodological choices that, taken together, systematically overstate the financial return on a college degree.
One piece of context matters before examining those choices. The CEW is a university research center. Its reports are written by its own staff, and I have found no sign that they go through independent peer review. That is not unusual in policy research. But it is relevant when weighing its conclusions. My book is also not peer-reviewed. What both have in common is simple. Their assumptions and data sources can be examined, tested, and replicated. That is the standard that applies here.
The CEW calculates lifetime earnings by summing annual median earnings for full-time, full-year workers ages 25 to 64, comparing college graduates to high school graduates. The 2011 report shows college graduates earning $2,268,000 over that window, compared with $1,304,000 for high school graduates. A $964,000 gap.1 The gap shows up in wages, too. San Francisco Fed researchers Leila Bengali, Marcus Sander, Robert Valletta and Cindy Zhao put the college wage premium, how much more college graduates earn in average wages than high school graduates, at about 75 percent in 2022.2
The lifetime gap is real for the population it describes. Whether it describes the decision a student faces at 18 is another matter.
It does not.
What the CEW Left Out
CEW published an update in 2021. It does not correct the four structural omissions described here.
The first omission is the head start. The CEW starts its comparison at age 25. A high school graduate who goes directly to work at 18 has seven years of earnings before the CEW starts counting, four of them while the college student is still in school.
During four years of full-time college, the high school graduate is earning and accumulating. Those four years alone are worth approximately $148,000 in earnings before the comparison even begins.
None of it appears in the CEW’s model.
Figure 6-B in Chapter 6 of my book illustrates the cumulative effect.
The second omission is college costs. The CEW discusses college costs in its technical appendix but leaves them out of its lifetime figures.1
The College Scorecard’s national median net price for bachelor’s institutions is $20,081 a year after aid. With only 44 percent of bachelor’s graduates finishing within four years and a median time to degree of 52 months, total costs can grow significantly.
Excluding four years of direct costs from a lifetime earnings comparison removes approximately $80,000 from the calculation before discounting. That is four years at today’s median net price, and most graduates take longer than four years.
By age 22, when the college graduate starts work, the median high schooler is about $228,000 ahead on the model’s own figures, counting four years of earnings plus four years of today’s median net price (see Figure 6-C in Chapter 6 of my book). The graduate’s higher pay from 22 to 24 closes part of that gap. At 25, when the CEW starts counting, the high schooler is still about $173,000 ahead. The CEW’s lifetime figures leave all of it out.
Taxes are the third omission. The federal tax system is progressive. Under the 2025 tax brackets and the $15,000 standard deduction the model uses, a high school graduate earning $35,872 a year pays roughly $2,300 in federal income taxes annually. A college graduate earning $56,943 pays roughly $4,800, rising into the 22 percent bracket as earnings grow. On the model’s own salaries, as both incomes grow over a career, that gap adds up to roughly $227,000 in additional federal income taxes paid by the college graduate. A lifetime-earnings count like the CEW’s treats that money as earnings, but it never reaches the worker’s bank account.
The fourth omission is the time value of money. A dollar earned 30 years from now is not worth a dollar today.
Table A-A · Time value of money: what $100 received in the future is worth today at various discount rates
| Discount rate | In 10 years | 20 years | 30 years | 40 years | 50 years |
|---|---|---|---|---|---|
| 5% | $61.39 | $37.69 | $23.14 | $14.20 | $8.72 |
| 7.8% | $47.19 | $22.27 | $10.51 | $4.96 | $2.34 |
| 10% | $38.55 | $14.86 | $5.73 | $2.21 | $0.85 |
Source · Author calculation using the standard present-value formula at three discount rates.
The CEW acknowledges this criticism in its technical appendix and shows a version discounted at 2.5 percent, while keeping undiscounted dollars for its headline.1 At that rate, the $964,000 premium falls to $593,000. But the CEW chose 2.5 percent because it is the real interest rate on long-term government bonds. That is a risk-free rate applied to one of the riskier investments a family can make. From July 2023 through the 2023-24 school year, the Federal Reserve held its benchmark rate at 5.25 to 5.5 percent. Parent PLUS loans made that year carried 8.05 percent. At either of those rates, the CEW’s premium shrinks further still.
What the Research Shows
Other research also finds a college premium. The CEW’s own report cautions that its figures are “not exact representations of an individual’s lifetime earnings.”1 In my experience, its headline number is still quoted as if it were one.
Kristen Broady and Brad Hershbein found in a 2020 Brookings analysis that the premium depends heavily on field. Measured in present value at a 3 percent discount rate, lifetime earnings for the median graduate range from $770,000 in early childhood education to $2.28 million in aerospace engineering. Late in their careers, the median graduate in two majors, early childhood education and visual and performing arts, no longer earns more each year than the median high school graduate.3 A median hides the spread it summarizes.
Michael Itzkowitz of The HEA Group, analyzing College Scorecard data for 3,887 institutions in 2024, found that at 26 percent of those institutions, most former students earn less than a typical high school graduate ten years after enrolling.4 That count covers certificate and two-year schools as well as four-year colleges. It also counts former students who left without a degree. Among the 1,657 colleges that mainly award bachelor’s degrees, the share is about 5 percent, or 84 schools. At the rest, a median above the line can still leave many former students below it.
Both findings point to the same conclusion: the CEW’s aggregate number is accurate for the population it measures and dangerously misleading for the individual making a choice. It is a sum of typical earnings, built from outcomes ranging from highly positive to materially negative, read as if it describes each person’s likely result.
The Complete Model
The model developed in my book corrects all four omissions. It starts at age 18, not 25. It includes full college costs. It taxes earnings at federal rates. It applies a 7.8 percent discount rate, a little above the 7 percent the federal government uses as its benchmark for benefit-cost analysis (OMB Circular A-94, reinstated in 2025).
The result: the Net Present Value (NPV) of lifetime after-tax earnings for the median high school graduate is $484,376. For the median college graduate, after paying the full cost of attendance, it is $402,005.
The CEW’s figures show a $964,000 lifetime advantage for college. The complete model shows an $82,371 disadvantage. At the median net price after aid, the graduate is still about $17,500 behind.
At full cost, that is a swing of over one million dollars. The earnings data are not the CEW’s: the model uses College Scorecard earnings four years after graduation. But count those same earnings the CEW’s way, from 25 to 64 with no costs, taxes, or discounting, and college shows a gain of about $1.16 million.
The method, not the data, turns that gain into a loss.
What This Means
The college wage premium exists. What the report cannot do is tell any individual student whether a specific degree, at a specific school, in a specific field will beat the alternative. No aggregate study can do that.
In my view, the CEW’s headline number is the kind of figure used to justify lending decisions that skip any individual calculation. Today there is $1.7 trillion in outstanding educational debt. And I think many borrowers have not earned enough to justify what they borrowed. The aggregate statistic did not cause this result on its own. Treating a median as individual advice helped.
About this analysis
This critique appears as Appendix A (Georgetown, The College Payoff) in We Need To Talk About Higher Education by Leon Shivamber.
Get the book → Read the argument in full → Run your own numbers →
Notes
- Carnevale, Anthony P., Stephen J. Rose, and Ban Cheah. “The College Payoff: Education, Occupations, Lifetime Earnings.” Georgetown University Center on Education and the Workforce, 2011, pp. 3, 21-22. https://cew.georgetown.edu/wp-content/uploads/collegepayoff-completed.pdf
- Bengali, Leila, Marcus Sander, Robert G. Valletta, and Cindy Zhao. “Falling College Wage Premiums by Race and Ethnicity.” FRBSF Economic Letter 2023-22, August 28, 2023. https://www.frbsf.org/research-and-insights/publications/economic-letter/2023/08/falling-college-wage-premiums-by-race-and-ethnicity/
- Broady, Kristen, and Brad Hershbein. “Major Decisions: What Graduates Earn over Their Lifetimes.” Brookings Institution, October 8, 2020. https://www.brookings.edu/articles/major-decisions-what-graduates-earn-over-their-lifetimes/.
- Itzkowitz, Michael. “Ensuring a Living Wage Through Higher Education.” The HEA Group, February 2024. https://www.theheagroup.com/blog/ensuring-a-living-wage-through-higher-education. Data download: https://www.theheagroup.com/s/22124FinalEnsuringaLivingWageSpreadsheet-1.xlsx.
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 ROI ranking, examined.
- IHEP’s Rising Above the Threshold, 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?