We hear it all the time: we have to work longer. This leads to tensions, strikes, and demonstrations. Just this week, among the pilots in the country where I live. But… are there also personal benefits to working longer? It is perhaps a question that some might find offensive, but it is one that Noah Arman Kouchekinia and colleagues sought to answer concretely in a new NBER working paper: Does working longer lead to less cognitive decline? The answer is, as always… nuanced.

In their study, the authors do not simply compare who works and who does not. Instead, they examine what happens when people are effectively “pushed” by labour-market changes. That distinction matters. People who work differ systematically from those who do not. Health, education, motivation… all of these play a role. Ignore that, and you can no longer tell cause from effect.

So the researchers turn to local employment shocks as a natural experiment. When jobs disappear or expand in a region, people’s likelihood of working changes, largely independent of their individual characteristics. That allows them to move closer to a causal effect.

At first glance, the result looks straightforward: when employment falls, cognitive scores decline more quickly. In other words, less work appears to go hand in hand with faster cognitive decline. But the story does not stop there. The effect is mainly seen among men aged 51 to 64. Among women and older men, it is largely absent. That alone makes clear that there is no universal rule.

The effects are also modest. We are not looking at dramatic changes, but at gradual shifts in cognitive scores over time. That fits with what we already know: cognitive decline often starts early and unfolds slowly.

More interesting, perhaps, is the question of why work might matter at all. The obvious explanation is cognitive stimulation. Work often involves solving problems, planning, and communicating. These are exactly the kinds of activities linked to building “cognitive reserve”.

But that is likely only part of the story. Work also provides structure, social interaction, and a sense of purpose. It may well be that this combination drives the effect.

What this study offers is not a definitive answer, but a shift in confidence. The idea that work can help slow cognitive decline gains some support. At the same time, the effect depends on context, age, and, likely, the nature of the work itself.

That makes the findings relevant for policy. Debates about working longer usually focus on pensions and affordability. But there may also be a health dimension.

Still, this is no simple argument for extending working lives. Not every job is cognitively demanding. Not every career allows people to keep working in good health. And not everyone has the same opportunities. The more relevant question, then, is not whether we should work longer, but under what conditions doing so actually makes sense.