The MBA Math Just Got More Complicated

Tomas Chamorro-Premuzic – chief science officer at Russell Reynolds Associates, a professor of business psychology at University College London and Columbia University, an associate at Harvard’s Entrepreneurial Finance Lab, and the author most recently of Don’t Be Yourself: Why Authenticity Is Overrated (and What to Do Instead) (Harvard Business Review Press, 2025) – revives an old question in Harvard Business Review: Is grad school worth it? 

His answer is the usual “it depends,” weighed against three arguments in favor – salary bumps, career pivots, following a genuine passion – and three against: free alternatives online, the gap between theory and doing, and the debt. Chamorro-Premuzic notes that top MBA programs can lift earnings by 60% to 150%, dwarfing the 10% to 15% bump from degrees like museum studies or human services.

That view holds up reasonably well, in the abstract. But for the specific case of the MBA, facts reported in recent months complicate nearly every plank of the argument – sometimes reinforcing it, sometimes undercutting it.

THE SALARY CASE IS SOFTENING WHILE AI EATS THE ENTRY-LEVEL RUNGS

Chamorro-Premuzic’s strongest argument for grad school is the pay premium, and on that front the MBA still wins relative to almost any other master’s degree. But the premium is compressing. GMAC’s 2026 Corporate Recruiters Survey found projected median U.S. starting salaries falling across every degree type this year: MBA hires down to $120K from $125K, non-MBA business master’s grads to $82,500 from $92,500, and bachelor’s hires to $72K from $75K. Seems small – but GMAC cautions the year-over-year shifts sit within the margin of error, and that the dip lands against a backdrop of real wages already trailing inflation, so the erosion may be sharper than the headline figures suggest.

The HBR piece treats reskilling toward AI-resistant work as a matter of picking the right major. The same GMAC survey suggests the ground is shifting faster than that view implies: 1 in 3 employers now say they’ve already replaced at least some entry-level roles with AI, and skill in using AI tools has topped employers’ list of future-critical capabilities for the second year running. That is a complication for Chamorro-Premuzic’s “set a career change in motion” argument – the entry point he assumes awaits graduates on the other side of a degree is itself being automated.

APPLICANTS ARE PRICING IN THE RISK; EMPLOYERS AREN’T SO SURE

Historically, uncertain job markets have pushed more, not fewer, professionals toward business school, a pattern that held through the 2008 financial crisis, subsequent recessions, and the coronavirus pandemic. But this cycle looks different. We know about double-digit application declines at several unnamed top-ranked programs, with one top-10 MBA seeing first- and second-round applications fall 30%. Analysts attribute part of the shift to AI’s effect on candidate psychology – prospective students are unsure not just about the job market but about the durability of white-collar credentials generally. That is close to Chamorro-Premuzic’s “you may be wasting your time” argument, just arriving a few years earlier than he might have expected.

Set against all that, we reported this month that MBB firms are describing MBA hiring as steady rather than in retreat, even as the composition of who gets hired shifts toward candidates who’ve gone deep in one industry rather than broad across many. And GMAC’s own survey, however smaller and less Fortune-500-heavy than past years, shows employer confidence in the MBA credential holding up even as the specific skills they prize evolve. 

That tension – rising confidence in the degree, falling confidence in what comes after it – is arguably the more precise 2026 version of the dilemma Chamorro-Premuzic describes in general terms.

Read the HBR article, Should You Go To Graduate School? here.

DON’T MISS GLOBAL MBA DEMAND IS UP; U.S. PROGRAMS ARE LOSING GROUND 

© Copyright 2026 Poets & Quants. All rights reserved. This article may not be republished, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Poets & Quants, please submit your request HERE.