A growing body of economic research is identifying specific graduate degree programs that are producing negative returns on investment for their students, a finding that cuts through the general cultural assumption that any graduate degree is worth the cost. The clearest example is psychology: a comprehensive analysis by economists at Yale, led by professor Joseph Altonji whose prior work on education returns has shaped labor economics for decades, found that psychology graduate programs are delivering a cost-adjusted ROI of approximately -8 percent for master's degree holders and -5 percent for those completing clinical psychology doctoral programs. Those negative figures account for the opportunity cost of time in school as well as tuition, and measure actual post-graduation earnings against the earnings trajectory of comparable individuals who did not pursue the degree.
The Altonji analysis, published in the Journal of Human Resources in , is not the only data pointing in this direction. The Georgetown Center on Education and the Workforce has been tracking degree-level ROI for years and has published updated figures for 2026 that show an expanding gap between degree programs producing strong earnings premiums and those that are failing to deliver earnings sufficient to offset the investment. The gap between the highest and lowest ROI programs has widened significantly since 2022, a period that corresponds with the acceleration of AI-driven automation in several professional service categories.
Why Psychology's Returns Are Negative
The psychology ROI problem is driven by the intersection of three structural factors. The first is credential inflation: the clinical psychology field has accumulated degree requirements faster than the earnings available to credentialed professionals have grown. A licensed clinical psychologist today typically holds a doctoral degree, has completed a one- to two-year supervised internship, and has passed multiple licensing examinations. That credential pathway takes seven to ten years after undergraduate completion and costs between $150,000 and $300,000 in direct tuition at most programs. The median salary for a licensed clinical psychologist in the United States is approximately $89,000.
The second factor is AI-driven substitution in adjacent roles that have historically provided employment pathways for psychology degree holders while they worked toward licensure. Entry-level positions in counseling support, assessment assistance, and behavioral data analysis, jobs that psychology graduates held while accumulating supervised clinical hours, have been automated or substantially reduced at several major healthcare systems and corporate wellness programs. The result is a longer and more uncertain path to the supervised hours required for licensure, which extends the investment period without increasing the eventual earnings at its end.
"The credential requirement for clinical psychology has not changed meaningfully in 20 years. The earnings premium for holding that credential, adjusted for the cost of obtaining it, has declined substantially. Those two things in combination create the negative ROI we are measuring."
Professor Joseph Altonji, Yale University, in the Journal of Human Resources paper, March 2026
The third factor is the specific labor market for licensed psychologists, which is geographically concentrated, insurance-reimbursement-dependent, and operating in a compensation environment that has not kept pace with the cost structure of the degree path. States where clinical psychology is well-compensated, primarily New York, California, and Massachusetts, have an excess of credentialed applicants relative to available positions. States with more open markets for psychology practice often have lower reimbursement structures that depress earnings below national averages even for fully licensed practitioners.
The High-ROI Comparison: Medicine and Law
The contrast with medicine and law is instructive for understanding why psychology's negative returns are not simply an artifact of measuring the wrong variables. Medical degrees, which involve a comparable or longer training pathway and substantially higher tuition costs, produce a 173 percent positive ROI in the Altonji analysis. Law degrees produce approximately 41 percent ROI on average, though the variance within law is substantial: graduates of top-14 law schools in large-city markets produce significantly higher returns, while graduates of regional law schools entering low-competition markets may produce near-zero or negative returns depending on debt levels.
What medicine has that psychology does not is a compensation structure that scales with credential depth: an attending physician earns multiples of what a resident earns, which earns multiples of what a medical student earns, with clear escalation at each credential stage. The total investment is large, but the eventual earnings premium is large enough to produce a substantial positive return even after accounting for the investment period.
| Degree Path | Typical Duration After Undergrad | Approximate Cost | 2026 Estimated ROI |
|---|---|---|---|
| M.D. (Medicine) | 4 + 3-7 years residency | $250,000-$400,000 | +173% |
| J.D. (Law) | 3 years | $100,000-$220,000 | +41% (average) |
| M.B.A. (Business) | 2 years | $80,000-$180,000 | +29% (top programs) |
| M.S./Ph.D. Psychology | 5-10 years | $120,000-$280,000 | -5% to -8% |
| M.A. Social Work | 2 years | $40,000-$90,000 | -3% to +2% |
The AI Layer: Entry-Level Elimination
The degree-level ROI analysis intersects with a distinct but related trend: AI automation is reducing entry-level hiring in several professional service categories that have historically absorbed new degree-holders. This is not uniformly true, and the sectors most affected vary considerably from those least affected, but the pattern is visible in the data across multiple professional fields and is accelerating the negative ROI dynamic for degree programs that feed specifically affected labor markets.
Legal document review, a category of work that absorbed large numbers of recent law school graduates in the 2000s and 2010s, has been substantially reduced by AI document review tools. The downstream effect is not that lawyers are unemployed: it is that the entry-level hiring pathway has narrowed, which increases competition for the positions that remain and puts downward pressure on entry-level salaries in a way that compresses the early-career component of the ROI calculation.
Psychology and social work adjacent roles in corporate settings, including roles supporting employee wellness programs, HR behavioral analytics, and customer experience research, were growing prior to 2023 and have since contracted as companies deployed AI tools for sentiment analysis, employee engagement monitoring, and similar functions that previously required behavioral science expertise to execute. The professional labor that those functions consumed has been partially absorbed by AI, and the roles remain but in reduced numbers.
Financial analysis and accounting roles have followed a similar trajectory, with AI tools handling portions of work that previously required credentialed professionals at the entry level, though the impact has been more pronounced in the functions that involve pattern recognition and less pronounced in those requiring professional judgment and client relationship skills.
What the Data Does Not Tell You
ROI analysis of graduate degrees has well-known limitations that matter for how its findings should be applied. The most important is that average returns conceal enormous individual variation. A clinical psychology doctoral graduate who works in a high-reimbursement specialty area, develops a private practice in a high-income market, and maintains full clinical hours produces a very different financial outcome than the median clinical psychologist. The negative average ROI reflects the distribution of outcomes, including the many graduates who enter the field with high debt and encounter the career barriers described above, not a prediction of any individual's outcome.
The analysis also does not fully capture the non-financial dimensions of degree value that many graduates, including many psychology graduates, consider central to their decision. The ability to practice a profession that is aligned with personal values and that provides direct human benefit is not captured in an ROI calculation, and for some portion of every graduate cohort, that dimension of the choice is more important than the earnings trajectory.
What the research does clarify is that the assumption that graduate degrees are financially sound investments by virtue of being graduate degrees is not accurate across all programs and fields. The decision to pursue a psychology graduate degree in 2026, with full information about the negative average ROI and the AI-driven labor market headwinds in adjacent entry-level roles, is a different decision than it was in 2015, when the labor market was growing, AI displacement was theoretical rather than observable, and the negative ROI pattern was not yet visible in longitudinal earnings data.
Pathways That Are Producing Better Returns
The same research that identifies negative ROI degrees also identifies the fields where returns have held or improved. Technology-adjacent graduate programs, particularly in data science, computer engineering, and machine learning, continue to produce strong positive returns in most market conditions. Healthcare professional programs outside of clinical psychology, including physical therapy, occupational therapy, and physician assistant training, have maintained positive returns through the current period of AI automation by combining credential requirements with hands-on patient contact skills that have proven resistant to direct automation.
Trade certifications and apprenticeship-backed credentials, a category that overlaps with but is distinct from traditional graduate programs, are producing some of the strongest ROI in the current labor market according to data from the skilled trades sector. The jobs that require a physical presence, manual dexterity, and contextual problem-solving in variable environments have retained and in some cases expanded their earnings premium as AI has absorbed portions of the knowledge economy that were once assumed to be the exclusive territory of credentialed professionals.
The practical implication for students currently deciding on graduate programs is not to avoid psychology as a field of study but to enter the degree pathway with accurate information about what the labor market for psychology credentials currently looks like, and to make intentional choices about specialization and location that give the best chance of landing in the portion of the distribution where positive returns are achievable. The negative average does not describe everyone. It does describe enough graduates to make the decision worth examining carefully rather than by default.
Frequently Asked Questions
Does this mean psychology is not worth studying?
Not necessarily. The negative ROI reflects the financial return relative to cost and opportunity cost, not the intrinsic value of psychology as a field or as a career. Graduates who are aware of the labor market conditions and who are making the decision primarily on grounds of personal fulfillment or values alignment are making a different calculation than the ROI analysis is measuring.
Which psychology specializations have better outcomes?
Neuropsychology, industrial-organizational psychology, and health psychology tied to specific reimbursable medical contexts tend to produce better financial outcomes than general clinical psychology. I-O psychology, in particular, has maintained stronger earnings trajectories because its applications in corporate settings have been less directly automated than some clinical functions.
Is the negative ROI pattern the same outside the US?
The degree of negative ROI varies by country depending on tuition costs and public versus private funding of graduate education. Countries with heavily subsidized graduate education, including most of Western Europe, produce smaller ROI gaps because the investment side of the calculation is lower. The earnings premium pattern for psychology versus medicine is broadly similar across countries, but the negative return specifically is more pronounced in countries with high graduate school tuition costs.
Sources
- Journal of Human Resources — Altonji et al., "Graduate Degree Returns in an AI-Adjacent Labor Market," March 2026
- Georgetown Center on Education and the Workforce — Graduate Degree ROI Rankings 2026
- Bureau of Labor Statistics — Occupational Outlook Handbook: Psychologists
- American Psychological Association — Workforce Research and Labor Market Data 2026













