Is college worth it? For many people who complete a degree at a manageable cost, yes. Anyone building a critique of higher education on the claim that degrees are generally worthless is starting with the wrong evidence.
But the opposite slogan is also an evasion. A positive return for the typical graduate does not validate every program, every price or every admissions promise. The question should not be whether higher education deserves unconditional praise or blanket contempt. It should be why institutions selling a costly opportunity are so often defended with averages that cannot describe the buyer.
The average defense is stronger than the cynics admit.
New York Fed economists Jaison Abel and Richard Deitz estimated a 12.5 percent internal rate of return for the typical college graduate in 2024. Their model incorporates direct costs and forgone earnings. It concerns graduates, not everyone who starts college, and is not a guaranteed investment yield for an individual student.[1]
That finding matters. Advising an entire generation to abandon university because several famous entrepreneurs did not finish would be a spectacular misuse of anecdote. For a prospective student, the relevant alternative is an actual available path, not somebody else’s exceptional career.
Still, a distribution is not a promise. In the companion analysis, the same economists estimated a return of 2.6 percent for graduates at the 25th percentile of the college earnings distribution under their baseline cost assumptions. They also examined how higher costs and longer completion times reduce returns.[2]
That is not proof that precisely one quarter of students lose money. A low modeled return is different from a negative return, and earnings are not the only benefit of education. It is enough, however, to invalidate the claim that the broad graduate premium makes program-level scrutiny unnecessary.
The degree can create value and still function as a gate.
College teaches real knowledge. It can also give employers a convenient screening device. Those functions are not mutually exclusive.
A February 2024 study by the Burning Glass Institute and Harvard Business School examined more than 11,000 roles for which employers had removed degree requirements. The share of workers without degrees hired into those roles rose by about 3.5 percentage points overall. Around 45 percent of the studied firms showed little meaningful change in actual hiring after changing their stated requirements.[3]
The study does not prove that all degrees are empty signals. It does expose the distance between announcing access and providing it. A recruitment system can stop demanding a degree in its advertising while continuing to favor one in practice.
When an employer can assess a skill directly but instead insists on an expensive credential, my objection is not that education has no value. It is that the employer may be outsourcing the cost of screening to applicants. The firm saves effort; the candidate must finance the ticket to be considered.
That arrangement deserves scrutiny even when buying the ticket remains individually rational. A toll can be worth paying because the road beyond it is valuable. That does not prove the toll is justified.
Do not sell economic security, then retreat into abstraction.
Education is more than earnings. Intellectual independence, cultural knowledge, friendships and the ability to understand a complicated society all matter. A university should be able to defend those goods without apology.
But a program should not emphasize career rewards when recruiting students and invoke the immeasurable value of learning whenever asked about poor economic outcomes. Those are different claims. Institutions should be clear about which promise they are making and what evidence supports it.
The same standard should apply to critics. Someone selling an alternative course should face questions about completion, outcomes and total costs too. Replacing university marketing with boot-camp marketing does not automatically produce a more honest market.
Make the institution carry an evidence burden.
Before asking students to commit years of work and substantial money, programs should disclose net costs, completion outcomes, relevant earnings distributions and the limits of the data. Outcomes should be reported in ways that do not quietly discard people who leave without a credential.
Where courses, advising or scheduling make timely completion harder, the institution should be accountable for the obstacle it controls. Where employers remove credential requirements, they should examine who actually gets hired. Where policymakers subsidize a program, they should ask whether it opens opportunity rather than merely expanding enrollment.
These are standards I am proposing, not findings that every institution currently violates. Their purpose is to stop treating the student as the only party responsible for evaluating a product whose sellers possess much more information about it.
Students should ask difficult questions. So should the people lending, subsidizing, regulating and hiring. It is unreasonable to demand perfect foresight from an eighteen-year-old while accepting slogans from organizations with decades of operating experience.
College can be worth the price. That is precisely why its reputation should not become a hiding place for weak programs and lazy hiring practices. Keep the education. Challenge the gatekeeping. Make the institutions prove what they are charging for.
nn
Sources
n
- Jaison R. Abel and Richard Deitz, Is College Still Worth It?, Federal Reserve Bank of New York, Liberty Street Economics, 16 April 2025. Estimates concern the typical graduate and 2024 economic conditions.
- Jaison R. Abel and Richard Deitz, When College Might Not Be Worth It, Liberty Street Economics, 16 April 2025. See the 25th-percentile earnings comparison and completion-time scenarios.
- Matt Sigelman, Joseph Fuller and Alex Martin, Skills-Based Hiring: The Long Road from Pronouncements to Practice, Burning Glass Institute and Harvard Business School, February 2024. Study of changed requirements and observed hiring, not a census of all employers.