The conversation everyone avoids goes like this: a 61-year-old wants fewer hours and less pressure; the organization wants to keep her judgment; and between them sits the question neither will raise, which is what happens to her pay and her title. So nothing is designed, she remains nominally full-throttle, and the actual downshift happens invisibly, as disengagement. The org loses the honest version of her and keeps the expensive silhouette.
The blockage is not financial. It is that late-career pay is governed by taboo where it needs arithmetic.
What the comparison research licenses
The fear behind the taboo is social: a visible reduction reads as demotion, and pay cuts are assumed to poison morale, the cuttee's and everyone's. The field evidence is more precise about when pay differences hurt.
Card, Mas, Moretti and Saez, in a field experiment with University of California employees, found that learning peers' salaries reduced satisfaction and raised job-search intentions among those paid below their unit median, with no symmetric gain above it. Comparison pain is real and asymmetric. But Breza, Kaur and Shamdasani, randomizing pay structures among Indian manufacturing workers, found the crucial moderator: pay inequality damaged output, attendance and cooperation only when the productivity difference behind it was not observable. When workers could see that the better-paid colleague was in fact more productive, the inequality had no adverse effect. Justifiability, not equality, is what morale runs on. And Zoë Cullen's 2024 review of the pay-transparency evidence adds the design corollary: transparency is neither good nor bad in the abstract; its effects depend entirely on what kind of comparison it enables.
Translate to the late career: a senior person moving to three days at three-fifths pay is not a morale hazard if the logic is public and the arithmetic is visible. What corrodes trust is the current default, where compensation and contribution drift apart in the dark, and everyone suspects everyone else's deal.
Decoupling the two currencies
The deeper design problem is that conventional careers weld two currencies together: money and hierarchical position. You cannot reduce one without appearing to lose the other, so people cling to titles they no longer want to fund lifestyles they could afford to shrink, and organizations pay leadership premiums for leadership no longer being exercised.
Semco, as Ricardo Semler has told it, unbolted the weld at one specific joint. Managers there were rated anonymously by their teams twice a year, and a manager the team would not keep lost the function. The designed exit mattered as much as the rating: returning to technical work at the same salary was a legitimate outcome, not a disgrace. Some of the managers who went back, he says, simply returned to doing good technical work.
Keep-the-salary-drop-the-title inverts the usual deal, and the inversion is instructive. It says: the organization was paying for your capability all along, not your altitude, so capability redeployed is not capability discounted. For late careers the applicable versions are several: the expert track that pays like management without being it; the phased arrangement where pay tracks days transparently; the advisory role priced per day at an honest rate rather than smuggled through a bloated retainer.
Making the trade discussable
A few practices convert the taboo into a menu.
Publish the shapes. If the organization has three sanctioned late-career forms (say: full role, three-day expert, one-day advisor), with the pay arithmetic attached, then choosing one is picking from a catalog, not confessing decline. The Breza finding operationalized: visible logic drains the poison from visible difference.
Separate the title conversation from the money conversation. Much late-career clinging is about standing, not cash. Organizations that offer durable professional standing (senior fellow, master craftsman, the named expert roles) make it cheap to release the managerial title, which is the one blocking the pipeline.
Time-box everything, renegotiate on schedule. An honest trade this year may misfit in three; standing reviews keep the arrangement matched to reality without requiring anyone to initiate the awkward conversation.
The quiet finding across the comparison literature is that people are not enraged by differences; they are enraged by stories that don't add up. Late-career pay design is, at bottom, the discipline of making the story add up in public: less time, less money, same respect, stated plainly. Most organizations have the spreadsheet for this. What they lack is the sentence.
Sources
- Card, Mas, Moretti & Saez (2012). Inequality at Work: The Effect of Peer Salaries on Job Satisfaction (American Economic Review, DOI 10.1257/aer.102.6.2981). Field experiment: below-median earners' satisfaction fell and search intentions rose after learning peer salaries; no symmetric gain above median. https://doi.org/10.1257/aer.102.6.2981
- Breza, Kaur & Shamdasani (2018). The Morale Effects of Pay Inequality (Quarterly Journal of Economics). Randomized field experiment: pay inequality reduced output and cooperation only when productivity differences were unobservable; justifiable differences had no adverse effect. https://academic.oup.com/qje/article-abstract/133/2/611/4430649
- Cullen (2024). Is Pay Transparency Good? (Journal of Economic Perspectives, DOI 10.1257/jep.38.1.153). Effects of pay transparency depend on its type (horizontal, vertical, cross-firm); transparency is a design variable, not a virtue in itself. https://doi.org/10.1257/jep.38.1.153