For decades, the calculus seemed straightforward: invest two or three years in graduate school, absorb the tuition costs, and emerge with credentials that would reliably translate into higher lifetime earnings. That equation has broken down. Badly.
A growing body of research now shows that many master’s degree programs — particularly in fields like education, psychology, and the arts — deliver negative financial returns when measured against the cost of attendance. Not modest returns. Not break-even outcomes. Actual losses, compounded over a career, that leave graduates worse off than if they’d never enrolled at all.
Fortune reported on the mounting evidence that graduate school’s value proposition has fundamentally deteriorated for a significant share of students. The story draws on data showing that across a wide range of institutions and disciplines, the net present value of a master’s degree is negative — meaning the degree costs more than it will ever return in additional earnings over a working lifetime. The implications are staggering, both for the roughly 800,000 Americans who earn master’s degrees each year and for the federal student loan apparatus that underwrites much of their study.
The numbers are particularly grim in education and psychology, two of the most popular graduate fields in the country. Education master’s programs have long been the single largest category of graduate enrollment, driven in part by state salary schedules that reward teachers for advanced credentials. But the salary bumps those degrees provide often don’t offset the tuition, lost wages, and loan interest accumulated along the way. Psychology programs face a similar dynamic: the terminal master’s degree rarely qualifies graduates for the higher-paying clinical or research positions that require a doctorate, yet students pay near-doctoral prices for the credential.
The Federal Loan Machine That Fuels the Problem
At the center of this dysfunction sits the federal Grad PLUS loan program. Unlike undergraduate federal loans, which carry annual and aggregate borrowing caps, Grad PLUS loans allow students to borrow up to the full cost of attendance — tuition, fees, living expenses, everything — with virtually no ceiling. The program was designed to expand access. What it has also done is eliminate any market discipline on graduate tuition pricing.
Universities face almost no constraint on what they can charge. If a school raises tuition by $5,000, the federal government will simply lend students an additional $5,000. The student bears the risk. The institution collects the revenue. And the cycle continues.
This matters because graduate students now account for roughly 40% of all federal student loan dollars disbursed annually, despite representing a much smaller share of total enrollment. The average graduate student borrower leaves school with over $70,000 in debt, according to data from the National Center for Education Statistics. For students in lower-returning fields, that debt load can take decades to repay — if it’s repaid at all.
Income-driven repayment plans, which cap monthly payments at a percentage of discretionary income and forgive remaining balances after 20 or 25 years, have become a de facto subsidy for programs that can’t justify their price tags on earnings alone. The Biden administration’s SAVE plan, now entangled in litigation, would have accelerated forgiveness timelines and reduced payments further. But even under existing repayment structures, taxpayers are absorbing billions in losses on graduate loans that will never be fully repaid.
The incentive structure is perverse. Schools profit from enrollment regardless of outcomes. Students are shielded from the full cost of bad decisions by income-driven repayment. And the federal government — meaning taxpayers — picks up the tab for the gap between what was lent and what’s recovered.
As Fortune noted, researchers at the Foundation for Research on Equal Opportunity and other organizations have been quantifying these dynamics with increasing precision. Their findings suggest that at least a quarter of master’s programs leave the median graduate with negative net returns. Some estimates are higher.
Not every field is underwater. MBA programs at selective institutions still show strong positive returns. So do many engineering and computer science master’s degrees. Health-related graduate programs — nursing, physician assistant studies, occupational therapy — tend to clear the bar comfortably. The problem isn’t graduate education writ large. It’s the specific combination of high tuition, modest post-graduation salaries, and unlimited federal lending that turns certain programs into financial sinkholes.
AI Is Redrawing the Map — and Fast
Now layer in artificial intelligence. The rapid deployment of AI tools across white-collar industries is beginning to compress wage premiums for precisely the kinds of analytical and administrative tasks that many master’s-level workers perform. If a generative AI system can draft lesson plans, summarize research literature, write policy memos, or conduct preliminary data analysis, the incremental value of a human with a master’s degree performing those same tasks starts to erode.
This isn’t speculative anymore. Employers across education, consulting, media, and social services are already integrating AI into workflows that previously required advanced-degreed professionals. A recent report from the McKinsey Global Institute estimated that generative AI could automate tasks accounting for up to 30% of hours worked in the U.S. economy by 2030, with knowledge workers bearing a disproportionate share of the disruption.
For prospective graduate students, this creates a double bind. The degree costs more than ever. And the labor market premium it confers may be shrinking in real time.
Some universities are responding by adding AI-focused coursework or certificates to existing master’s programs. Whether that’s enough to restore their value proposition remains an open question. A two-credit elective on prompt engineering doesn’t fundamentally change the earnings trajectory of an education or counseling psychology degree.
Others are going further. Georgia Tech’s online master’s in computer science, priced at under $10,000 total, has become a frequently cited example of what’s possible when institutions rethink delivery models. But Georgia Tech is the exception, not the rule. Most graduate programs remain priced as if the old value equation still holds.
The market is starting to notice. Graduate enrollment in education programs has declined significantly over the past decade, though it remains large in absolute terms. Psychology graduate enrollment has been more resilient, partly because licensure requirements in many states effectively mandate advanced degrees for entry into the profession. When the government requires a credential, demand for that credential persists regardless of its financial return.
That’s a critical distinction. Many of the worst-returning graduate programs exist in fields where the degree is either required by regulation or strongly incentivized by employer pay scales. Teachers earn more with a master’s degree under most district salary schedules. Social workers need a master’s for clinical licensure. School psychologists need a specialist-level degree. These aren’t optional credentials for the workers who pursue them. They’re table stakes.
Which means the negative-return problem isn’t simply a matter of students making uninformed choices. It’s a structural feature of how certain professions are regulated and compensated. Fixing it requires more than better consumer information — though that would help. It requires rethinking licensure requirements, tuition pricing, and the federal lending framework simultaneously.
Congress has shown intermittent interest. Proposals to cap Grad PLUS borrowing, require programs to meet minimum earnings thresholds to maintain loan eligibility, or shift some repayment risk to institutions have circulated for years. None have become law. The higher education lobby is formidable, and any reform that threatens enrollment revenue faces fierce opposition.
Meanwhile, the debt accumulates. The Federal Reserve Bank of New York estimates total outstanding student loan debt at approximately $1.6 trillion. Graduate borrowers hold a disproportionate share. And with income-driven repayment plans ensuring that many of these balances will eventually be forgiven rather than repaid, the fiscal exposure keeps growing.
There’s an uncomfortable question embedded in all of this: Should the federal government be lending unlimited sums for degrees that data shows are likely to leave borrowers financially worse off? The libertarian answer is that adults should be free to make their own choices. The progressive answer is that access to education shouldn’t depend on ability to pay. But neither answer grapples with the reality that the current system is transferring wealth from taxpayers to institutions through the intermediary of students who end up holding debt they can’t service.
Something has to give. Either tuition comes down, earnings go up, lending gets capped, or taxpayers continue absorbing the difference. The rise of AI makes the earnings side of that equation more uncertain than ever. And the political appetite for large-scale student loan forgiveness — while real in some quarters — is nowhere near universal enough to constitute a durable solution.
For now, the burden falls on prospective students to do the math themselves. Organizations like the Georgetown University Center on Education and the Workforce and the Foundation for Research on Equal Opportunity publish program-level return-on-investment data that didn’t exist a decade ago. The Department of Education’s College Scorecard provides earnings data by institution and credential level. The information is out there. But it competes with powerful cultural signals — from parents, employers, and institutions themselves — that still frame the master’s degree as an unambiguous good.
It isn’t. Not anymore. Not for everyone. And as AI continues to reshape which skills command a premium in the labor market, the gap between graduate programs that deliver value and those that destroy it is only going to widen.
The students entering graduate school this fall deserve to know that.
WebProNews is an iEntry Publication