The claim that working longer keeps you alive longer shows up often enough to sound settled. A 2020 systematic review and meta-analysis by Ranu Sewdas and colleagues, published in the Journal of Epidemiology and Community Health, pooled 25 separate studies on the relationship between retirement timing and mortality and found something that complicates the claim rather than confirming it. Early retirement was not associated with a higher risk of death. What was associated with higher mortality, across the pooled studies, was retiring at the standard, expected age, sometimes called “on-time” retirement.
The researchers’ explanation is what statisticians call the healthy worker effect. People who are already unwell are more likely to be pushed out of the workforce, through disability, job loss, or a simple inability to keep going, right around the standard retirement age, since that’s the point where leaving becomes easiest to justify or hardest to avoid. That means a chunk of the “on-time” retirees in these studies weren’t choosing to stop working because they’d hit a number on a calendar. They were retiring because they were already declining, and their subsequent mortality reflects the decline that pushed them out, not the retirement itself. Sorting out that kind of reverse causation is exactly why the review’s authors caution that studies on this topic need to account carefully for a person’s health before retirement, not just the age at which they stopped working.
A stronger test of the question comes from a natural experiment rather than an observational study. Economists Cristina Bellés-Obrero, Sergi Jiménez-Martín, and Han Ye examined a 1967 Spanish pension reform, in research circulated as an IZA discussion paper, that changed early-retirement eligibility based on an almost arbitrary detail: the date a worker first started paying into the social security system. People who started contributing before January 1, 1967 could claim a pension voluntarily at 60. People who started just after that date could not claim one until 65. Because which side of that line a worker landed on had little to do with their health or career choices, comparing the two groups isolates the effect of delayed retirement far more cleanly than comparing people who simply chose to retire at different ages.
The results were not simple good news for working longer. The researchers found that delaying exit from the labor force by one year raised the risk of death between ages 60 and 69 by about 4.2 percentage points, a 43 percent relative increase. But that effect was not spread evenly. It was concentrated heavily among people in physically demanding and psychosocially stressful jobs: the mortality increase from an extra year of work was roughly twice as large for blue-collar workers as for others, and nearly twice as large for people in high-psychosocial-burden jobs compared with low-burden ones. Workers who reported feeling fulfilled and recognized in their jobs showed no measurable increase in mortality risk from working longer at all. The harm, in other words, was not coming from working itself. It was coming from working longer in jobs that were physically grinding or emotionally depleting, without much say in the matter.
There is a separate, more encouraging thread of research on a different outcome entirely: cognition rather than mortality. Researchers Dawn Carr at the Stanford Center on Longevity and Robert Willis at the University of Michigan, drawing on the long-running Health and Retirement Study, have found that older adults who remain in cognitively demanding jobs tend to show slower cognitive decline than people of the same age who have retired. That finding, like the mortality research on early retirement, is observational rather than experimental, so it can’t rule out that people with sharper cognition to begin with are simply the ones more able to keep working. But it points in a similar direction as the Spanish study: what seems to matter is not simply whether someone keeps working, but what kind of work it is.
Put together, neither “retire as early as you can” nor “forget retirement, work a little longer” holds up as advice that applies evenly to everyone. The cleanest causal evidence available, from the Spanish reform, found that delaying retirement did carry a real health cost, but one heavily concentrated among people in physically taxing or unrewarding jobs, and close to absent among people who found their work fulfilling. The widely cited claim that early retirement itself shortens lives turns out to rest on studies that struggle to separate cause from a simpler explanation: people who are already