The standard account of retirement is that it takes your sense of purpose with it. The best causal estimate available points the other way. A 2021 study in Psychological Science, working with data from 8,113 older Americans, found that retiring raised people’s sense of purpose in life, and that the gain was largest among those with less money who were leaving jobs they had not enjoyed.
We are writers, not clinicians, and what follows is a reading of the research rather than advice about anyone’s own retirement.
The study, by Ayse Yemiscigil, Nattavudh Powdthavee and Ashley Whillans, is worth taking seriously, but it is not the final word.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
What the study actually did
The hard part of studying retirement is that people who retire are not a random group. They differ in health, money, job quality and much else, and any of those differences could explain a change in how purposeful they feel. Compare retirees with workers and you are comparing two groups that were already different.
The researchers got around this using what economists call an instrumental variable. US Social Security rules create financial incentives to retire at particular ages, and those incentives have nothing to do with any individual’s inner life. The rules act as a lever that pushes some people into retirement for reasons unconnected to how purposeful they feel, which allows a closer estimate of what retiring itself does.
This is a quasi-experiment, not a randomized trial. The estimate describes people whose retirement timing actually responded to those incentives, which is not everyone who retires.
The condition attached to the finding is the part most worth sitting with. The gains concentrated among people of lower socioeconomic status who were leaving unsatisfying work. Read plainly, that is a finding about escape. It says less about retirement as a stage of self-discovery and more about what happens when a job that was pressing on someone stops.
Which also means the people the research found benefiting most are not the people most retirement advice is written for.
Purpose does decline later in life, and that is a separate finding
Sense of purpose does fall across later adulthood. A coordinated analysis presented in 2025 by Olivia Duchow and Gabrielle Pfund pooled six longitudinal studies, including the Health and Retirement Study, the Wisconsin Longitudinal Study and three Rush cohorts, covering roughly 31,800 people between them. Their summary is that sense of purpose tends to decline later in the lifespan.
That work is a conference abstract from a Gerontological Society of America meeting, presented in a late-breaking poster session. Conference abstracts are usually not peer reviewed, and the abstract reports the direction of the finding without the detail that a full paper would carry.
So the largest-scale evidence for the decline is the weakest in publication status, while the more carefully identified retirement finding is peer reviewed. That is an awkward pairing, and it is the honest state of the literature.
The two results are not in conflict. One describes a slope across a decade or more of aging. The other describes what happens at one particular transition inside it. Merging them produces the familiar claim that retiring is what drains purpose, and neither study supports that.
A second study found mostly continuity
A three-year panel study of 5,034 Dutch adults aged 60 to 65, published in The Journals of Gerontology: Series B and led by Grünwald, tracked four things people said they valued: self-development, social status, contributing to society, and generativity, which is the term researchers use for investing in people who come after you.
Voluntary full retirement was linked to statistically significant decreases across all four. The effects were very small, between 0.15 and 0.24 of a standard deviation. People who kept working after retirement age, and those who retired involuntarily, changed less still. The authors describe relatively high levels of continuity through the transition.
Neither the crisis story nor the reinvention story predicts that. Most people came out of retirement valuing roughly what they valued going in.
The study has real limits, and the authors name them. It ran in the Netherlands, where retirement norms and mandatory ages differ from American ones. One of the four scales had weak internal consistency, meaning its questions did not hang together well enough to trust the number closely. And 21 percent of participants dropped out between the first wave and the last, which in aging research tends to remove the least healthy people from the sample.
Where this sits against the honeymoon account
The Artful Age has covered the other version of this story, in which retirement opens with a honeymoon that fades into disenchantment as leisure fails to supply what a job supplied. That account and the finding above cannot both be straightforwardly true. How they fit together turns out to depend on what is being measured.
A fixed-effects analysis of 38,344 people aged 50 and over across nine European countries, published in the Journal of Happiness Studies by Sohier and colleagues using the SHARE survey, split the question in a useful way. On life satisfaction, retirement produced no immediate change, and satisfaction then fell relative to the start of retirement about two years in. On what the researchers call agency-freedom, the sense of being able to direct your own time, the effect was immediately positive and had not faded at two years.
The honeymoon pattern shows up in one measure and not the other.
Purpose and agency sit with the measures that held. General life satisfaction sits with the one that slipped. An article reporting either result on its own will sound like it has settled a question it has only half answered. The authors also note that their post-retirement window runs about two years, which is not long enough to say what happens later, and that health and employment status are tangled together in the data.
Why the distinction is worth making
Most retirement advice treats the last day of work as the danger point, and organizes itself around preparing for that day. The research points somewhere else: at the years afterward, when health changes, partners are lost, and the roles that organized a life quietly stop being available.
Those are harder things to prepare for, and they arrive on no particular schedule.
There is also a reason to care about the measure itself. In a 2025 analysis of 9,808 Americans aged 65 and over, again drawing on the Health and Retirement Study, people whose sense of purpose held up over eight years also tended to hold onto cognitive function.
The authors of that paper are careful about the direction of that relationship, and a reader should be too. Cognitive change could support or erode a sense of purpose just as easily as the reverse. They name that bidirectionality as a limitation themselves, along with the fact that people who stay in a long study are not a representative sample of those who started it.
What this does not prove
Two of the strongest studies here draw on the same dataset. Different authors, different waves, different questions, but the Health and Retirement Study is doing a great deal of the work in this literature, and that is not the same thing as independent replication across separate cohorts.
Purpose is measured with short self-report scales. Someone answering a questionnaire in the months after leaving a job they disliked may be recording relief, and relief and purpose are not the same thing even when they move together.
None of this says what any individual should do. A finding about averages across 8,113 people does not tell one person whether to retire, when, or what will happen to them afterward.
What would settle it is the study nobody has run: following people across the transition with something better than a short questionnaire, in more than one country, and testing whether anything deliberately raises purpose rather than only measuring where it happens to sit. The researchers working on the decline list intervention development as future work. That is a careful way of saying it has not been done.