The requisition was hers. The manager scoped the role, fought for the budget, ran the interviews. Then the offer letter came back from compensation, and the number at the bottom was higher than her own. She approved it anyway, because the market is the market, and spent the commute home doing arithmetic she will not share with anyone.
This scene, familiar to a generation of managers hired before the recent pay run-ups and now supervising people hired during them, is usually filed under personal grievance. The research suggests it is a structural problem that reaches well past one bruised ego.
The boss's salary is not private
Cullen and Perez-Truglia's field experiment inside a large Asian bank, with 2,060 employees and administrative records, established two things. First, employees systematically underestimate their manager's salary, by 14.1 percent on average. Second, the direction of a salary comparison determines its motivational sign. When employees learned their manager earned more than they had assumed, they worked more: a 10 percent increase in the perceived manager salary raised hours worked over the following 90 days by about 1.5 percent, with the mechanism traced to career aspirations, the sense that the ladder pays. Horizontal comparisons ran the opposite way: a 10 percent increase in perceived peer salary reduced hours worked by 9.4 percent.
Read through a compression lens, this is uncomfortable. The gap between a manager's pay and her team's is not just her reward; it is the posted prize of the tournament everyone below her is running. Compress it to nothing and the signal reads: the ladder does not pay here. The manager who earns less than her new hire is a walking advertisement against ambition, visible to everyone with a salary benchmarking tab open.
Why it stings, precisely
Card, Mas, Moretti and Saez showed what peer comparison does when it goes badly: University of California employees who learned they earned below their unit's median reported lower job satisfaction and higher intent to leave, while those above the median gained no symmetric boost. The effect of pay information is asymmetric and mostly downside. Breza, Kaur and Shamdasani sharpened the point in a field experiment with Indian manufacturing workers: pay inequality damaged output, attendance and cooperation when workers could not observe the productivity difference behind it, but when the higher-paid colleague was visibly more productive, the same inequality caused no harm. The active ingredient is justifiability. People do not demand equal pay; they demand a legible reason.
A market hire who out-earns her manager can be justifiable. Scarce skills, hot market, documented benchmarks. What organizations typically do instead is hide the anomaly, swear HR to silence, and let it leak, unexplained, into the exact conditions the research says are most corrosive.
Transparency mandates alone do not solve this either. Mas studied California's disclosure law for municipal manager pay and found that forced publication compressed managers' compensation and pushed their quit rates up. Cullen's review for the Journal of Economic Perspectives concludes that transparency is neither good nor bad in the abstract; horizontal, vertical and cross-firm transparency have different, sometimes opposite effects, and design is everything.
The Semco test
Semco's approach in the 1980s was to make the arithmetic itself public. As Ricardo Semler has told it, a fixed share of each unit's profit, 23 percent, went to the people who produced it, with the books open so anyone could check the calculation; his favorite interview question for candidates was what percentage of last year's profit went to non-executives. Whatever one thinks of the specific numbers, the design principle survives: pay peace comes not from hiding differences but from letting people verify the logic that produced them.
For the compressed manager, that principle yields a practical agenda. Get the compression named: an anomaly acknowledged with a correction path is survivable, one denied is radicalizing. Push for pay logic the team can inspect, bands, criteria, review cadence, even where exact figures stay private. And when the market hire's premium is real, say so, because Breza's workers have already demonstrated how people respond to differences nobody explains.
The manager who earns less than her newest report has a legitimate personal claim. But her stronger argument to the organization is not "pay me more because it hurts." It is that an unexplained inversion at the center of the team quietly reprices ambition for everyone watching. Companies do not get to skip the explanation. They only get to choose whether they give it, or the rumor mill does.
Sources
- Cullen, Z. & Perez-Truglia, R. (2022). How Much Does Your Boss Make? The Effects of Salary Comparisons. Journal of Political Economy 130. Employees underestimate manager pay by 14.1%; higher perceived manager pay raises effort (elasticity 0.15 on hours), higher perceived peer pay lowers it (elasticity -0.94). https://doi.org/10.1086/717891
- Card, D., Mas, A., Moretti, E. & Saez, E. (2012). Inequality at Work: The Effect of Peer Salaries on Job Satisfaction. American Economic Review 102(6). Below-median earners lost satisfaction and looked to leave; above-median earners gained nothing symmetric. https://doi.org/10.1257/aer.102.6.2981
- Breza, E., Kaur, S. & Shamdasani, Y. (2018). The Morale Effects of Pay Inequality. Quarterly Journal of Economics 133(2). Unobservable pay inequality cut output (~0.45 SD) and attendance (18 pp); justifiable inequality caused no harm. https://academic.oup.com/qje/article-abstract/133/2/611/4430649
- Mas, A. (2017). Does Transparency Lead to Pay Compression? Journal of Political Economy 125(5). Forced disclosure of California municipal manager pay compressed compensation and increased quits. https://doi.org/10.1086/693137
- Cullen, Z. (2024). Is Pay Transparency Good? Journal of Economic Perspectives 38(1). Horizontal, vertical and cross-firm transparency have distinct, sometimes opposite effects; outcomes depend on design. https://doi.org/10.1257/jep.38.1.153