Retirement, as currently designed, is a scheduling error. It takes the two things a working life rations most brutally, free time and freedom of attention, and delivers the entire lifetime supply in one shipment, at the end, to a recipient with less energy, fewer surviving friendships and, often, a body that has already filed its complaints. Meanwhile the 35-year-old, at peak health and peak obligation, gets none of it: not the Tuesday morning with a child, not the instrument, not the wave. The system asks people to defer being alive on the promise that the deferral compounds. Anyone who has watched a parent retire into bewilderment knows the compounding is not guaranteed.
Ricardo Semler's answer to this, as he has told it, was a program he called retire-a-little. A Semco employee at 35 could buy back one day of the week, with a proportional pay cut, and spend it on whatever they would otherwise postpone to 65: the guitar, the kid, the surfboard. The other side of the contract ran in reverse: after 65, they could come back and work one day a week if they wanted the money back. The result, in his telling, was not the productivity leak a controller would predict. The people who took the deal stayed longer, and got more done in four days than others did in five.
The cliff was built for a different century
The block-at-the-end model is not ancient wisdom; it is infrastructure from an industrial bargain in which work was depleting, uniform and impossible to do in fractions, and in which few workers were expected to outlive the pension by much. Every assumption in that sentence has expired, and the demography is expiring fastest. The OECD's 2025 Employment Outlook tracks the old-age dependency ratio, people over 65 relative to the working-age population, rising from 19 percent in 1980 to 31 percent in 2023, with a projection of 52 percent by 2060. A system in which each retiree stands on the shoulders of ever-fewer workers cannot keep promising ever-longer blocks of full-stop leisure. The choices on the table are cruder retirement ages, thinner pensions, or a redesign of the shape itself.
The same OECD report carries a quieter number that argues for the redesign: in 2023, only about one third of adults aged 60 to 65 participated in any training, against more than half of those aged 25 to 44. The cliff model produces exactly this curve. Why would a worker invest in year 39 of a career designed to hit a wall at year 40? Why would an employer? A career shaped as a long taper rather than a cliff changes the arithmetic on both sides: the 63-year-old working two days a week is worth training, worth consulting, worth keeping on the ramp.
Buying back Wednesday
Redistribution has a concrete unit: the day. A day a week is 20 percent of a working life, which over three decades amounts to roughly six years of waking, healthy, mid-life time, moved from the far end of the conveyor to the middle. Priced as a proportional pay cut, it is not a perk and not a favor; it is a purchase, which is what makes it dignified and makes it scale. The employee who buys Wednesday is not asking permission to be less committed. They are exercising an option the contract should always have contained.
The objection writes itself: most people cannot afford the cut. For some, at some wages, that is simply true, and no slogan fixes it. But the objection proves less than it seems. People routinely price other purchases into their lives, cars, renovations, commutes chosen for square footage, without anyone calling the trade-off utopian. What has been missing is not affordability but availability: the option has not existed, so it has never been priced, compared or planned for. And the alternative purchase, six years of deferred life delivered at 68, is the most expensive item most workers ever buy. It is bought at par and redeemed, too often, at a steep discount.
The taper is the plan
Retire-a-little reframes retirement from an event into a dimmer switch: down a notch at 35, another in the 50s, never quite to zero while the work still feeds something. The old model's implicit promise, endure wholly now, live wholly later, was always a gamble on health and luck dressed up as prudence. Spreading retirement across a lifetime is not impatience. It is the refusal to schedule all of one's freedom into the years when freedom is cheapest to grant and hardest to spend.
Sources
- OECD (2025). OECD Employment Outlook 2025. The old-age dependency ratio across the OECD rose from 19% in 1980 to 31% in 2023 and is projected to reach 52% by 2060; only about one third of adults aged 60-65 participated in training in 2023, versus more than half of those 25-44. https://www.oecd.org/en/publications/oecd-employment-outlook-2025_194a947b-en.html