The famous half of the idea is the young half. At Semco, as Ricardo Semler has told it, employees around 35 could buy back one day of their working week, at a proportional pay cut, to do then what convention schedules for 65: learn the guitar, raise the child, surf on a Wednesday. The people who took the deal, he says, stayed longer and produced more in four days than others did in five.
The half nobody quotes is the other end of the contract: after 65, if you wanted the money back, you could work one day a week, for as long as you liked. Not a farewell tour. Not a consulting arrangement improvised over a nervous lunch. A standing, priced, symmetrical option, written into the same program, decades in advance.
That second clause deserves its own examination, because it quietly solves most of what is wrong with retirement design.
What the one-day clause actually does
First, it abolishes the cliff by construction. There is no date at which the person's contribution goes from everything to nothing; there is a curve that was negotiated when nobody was emotional about it. The one-day worker at 70 is not a special accommodation or a favor to a beloved veteran. She is executing a contract everyone understood at 35.
Second, it prices the exchange honestly in both directions. The mid-career worker paid, in salary, for time when time was worth most. The older worker earns, in salary, for judgment when judgment is what remains scarce. The symmetry is the point: the program treats a day of life and a day of work as convertible at every age, rather than pretending work is mandatory at 40 and worthless at 70.
Third, it keeps the transmission channel open. The most expensive thing about a cliff-edge exit is not the lost labor, which is one person-year, but the severed access to thirty years of pattern recognition. A weekly day keeps the veteran inside the informal circuits where judgment actually transfers: the hallway consult, the "before you sign that, call the guy in Rotterdam." No knowledge-management system replicates a Tuesday.
The demographics have caught up
What looked like a charming eccentricity now reads as policy foresight. The OECD's Employment Outlook 2025 projects the old-age dependency ratio across member countries rising from 31% in 2023 to 52% by 2060, and estimates that population aging, absent productivity gains or policy change, would slow per capita GDP growth by roughly 40%. The same report calculates that reducing older workers' exit rates to the level of the best-performing countries would add at least 0.2 percentage points of annual per capita GDP growth in about half the OECD. A standing one-day option for people past pension age is exactly the kind of mechanism that moves an exit rate, one person at a time, voluntarily.
The well-being direction is encouraging too, with a caveat. The largest working-time intervention study to date, the 2025 four-day-week trial published in Nature Human Behaviour (2,896 employees, 141 organizations, six countries), found reduced burnout and improved mental and physical health, with an individual-level dose-response between hours reduced and well-being gained. That study ran on mid-career workforces moving from five days to four, not seventy-year-olds working one; extrapolating the magnitudes would be dishonest. But its core lesson transfers: the benefits came from redesigning the work to fit the reduced time, not from the calendar alone. A one-day elder role needs the same redesign, a real slice of work shaped to a day, not a compressed impersonation of a full role.
What organizations would have to change
Almost everything about the one-day clause is administratively trivial, which is why its absence is revealing. Payroll can handle 20%. What actually blocks it is status architecture: organizations do not know what a one-day person is. Whose meeting do they attend? Who do they report to? The habit of equating seriousness with full-time presence runs so deep that anything less reads as semi-departure.
The fix is to define the day by output, not attendance: a named set of judgment calls, reviews, apprenticeships, client relationships that live on that day. The role is small; it is not vague.
Retire-a-little's second half is the rare retirement idea that asks nothing of governments and little of budgets. It asks organizations to believe that a person's last decade of judgment is worth one day a week, and to write that belief down thirty years early. The companies that find this radical might consider what they currently offer instead: everything until Friday, nothing after, and a phone call six months later asking the retiree if she'd consult.
Sources
- OECD (2025). OECD Employment Outlook 2025. Old-age dependency ratio 31% (2023) projected to 52% (2060); aging would slow per capita GDP growth ~40% absent policy change; best-decile older-worker exit rates worth at least 0.2pp annual growth in about half of OECD countries. https://www.oecd.org/en/publications/oecd-employment-outlook-2025_194a947b-en.html
- Fan, Schor, Kelly & Gu (2025). Work time reduction via a four-day workweek (Nature Human Behaviour, DOI 10.1038/s41562-025-02259-6). 2,896 employees, 141 organizations, six countries: reduced burnout, better mental and physical health; individual dose-response between hours reduced and well-being; gains reflect redesign plus reduction. https://www.nature.com/articles/s41562-025-02259-6