Psychology says procrastination about retirement may be less about discipline than identity — brain scans found people often represent their future selves more like strangers than like themselves, and experiments using age-progressed faces made tomorrow’s person feel real enough for participants to save more money for them

Retirement procrastination is usually described as a failure of discipline. The forms have been sitting in a folder for months. The pension dashboard remains unopened. The intention is sincere, but the action keeps being handed to next week.

There is another way to read that delay. The person who benefits is not quite experienced as the person making the sacrifice.

A line of psychology research calls this future self-continuity: the degree to which someone feels that the person they will become is still meaningfully them. When that connection is weak, setting money aside can feel less like looking after oneself and more like transferring resources to a distant stranger.

That is a psychological account, not personal financial guidance, and it cannot explain every delayed retirement decision. Income, debt, housing costs, pension access, financial literacy, administrative friction and distrust of institutions all matter. Identity may be one part of the behaviour without being the whole explanation.

The distance is personal as well as financial

Retirement is an unusually difficult object for the imagination. It may be decades away, its dates are uncertain and the person living through it will have routines, relationships and priorities that cannot be known in detail now.

Psychologists describe the preference for smaller rewards now over larger rewards later as temporal discounting. It is not peculiar to pensions, nor is it evidence of a defective character. Present costs are concrete. Future benefits are both delayed and abstract.

Hal Ersner-Hershfield and colleagues added identity to that familiar account. In a 2009 paper in Judgment and Decision Making, people who reported a stronger sense of overlap with their future selves also chose more delayed rewards in laboratory tasks. In another part of the work, perceived continuity was associated with accumulated financial assets after age and education were taken into account.

Those results were correlational. They did not establish that feeling close to a future self caused wealth, and they certainly did not remove the role of earnings, inheritance or circumstance. What they offered was a measurable version of an ordinary experience: the future can belong to us intellectually without feeling fully personal.

What the brain scan actually found

The strongest version of the headline says the brain treats the future self exactly like a stranger. The original study was subtler than that.

In research published in Social Cognitive and Affective Neuroscience, participants underwent functional magnetic resonance imaging while judging whether personality traits described their current selves, their selves ten years in the future, another person in the present and that person in the future.

The researchers focused on differences in activity in the rostral anterior cingulate cortex, a region previously implicated in judgements about oneself and other people. People whose neural response differed more between present-self and future-self judgements also tended to discount delayed monetary rewards more strongly in a behavioural task one week later. The reported relationship was moderate, with a correlation of 0.47.

The full 2009 paper does not show that every future self is neurologically indistinguishable from a stranger. It shows individual differences in one task, one region of interest and one small experiment. Brain imaging added another measure to the identity hypothesis; it did not turn a metaphor into a universal biological fact.

An older face made the recipient visible

The next question was whether that sense of distance could be changed. Hershfield and a larger team tested age-progressed images in four randomised experiments, reported in the Journal of Marketing Research in 2011.

Some participants entered immersive virtual environments and saw an avatar with their current face. Others saw a computer-generated version of their face at around retirement age. Later experiments used interactive tools in which a current or older face responded as participants adjusted the share of hypothetical income directed towards retirement.

Across the experiments, participants exposed to their older likeness generally showed more willingness to favour later rewards or allocate more hypothetical income to retirement. In one early task, the future-face group allocated an average of $172 from a hypothetical $1,000 windfall, compared with $80 in the current-face group. Other tasks produced smaller differences.

The effect was not simply a generic reminder that old age exists. One experiment compared a participant’s aged face with another person’s aged face, attempting to isolate the role of identification. The authors’ account, available through Microsoft Research, was that a visible future self made the distant recipient of today’s choice more concrete.

But participants did not move money into verified retirement accounts. They made choices in virtual environments, surveys and hypothetical allocation tools.

The experiment was cleaner than the real world

The US Department of Labor’s CLEAR evidence review rated the experiments highly for causal evidence within their designs. Random assignment supported the conclusion that seeing the manipulated images produced the observed differences inside those studies.

The same review was direct about generalisation. The first three experiments involved between 21 and 50 university students. The fourth used 40 adults recruited online. The outcomes were hypothetical rather than actual saving, and the participants were not a representative sample of people making pension decisions under ordinary financial pressure.

Later work also cautions against treating any reminder of ageing as sufficient. A 2020 UK replication study with 219 participants found no evidence that textual ageing primes increased retirement allocations, even when the researchers tried to raise future self-relevance. That intervention was not the same as a personalised aged face, but the null result matters. Vividness, personal resemblance and context may not be interchangeable.

A 2025 systematic review of future self-continuity interventions reached a measured conclusion: the approach is promising, but methods and outcomes vary enough that firm claims about effectiveness remain difficult.

Retirement products often hide the person

The research has an implication for the way pension systems and financial products are designed. Most interfaces present the future as a number: a balance, a projection, an age or a percentage. These can be necessary, but they leave the identity problem untouched.

An age-progressed portrait is one attempt to make the beneficiary harder to overlook. A less technologically elaborate version might use a concrete description of an ordinary future day, a letter addressed across time or language that treats retirement income as support for a continuing person rather than a score to maximise. Research on future self-continuity in a large nationally representative US sample suggests the underlying sense of connection varies considerably across people, so no single prompt should be expected to work for everyone.

There are design risks too. Ageing filters can turn an intimate anxiety into a sales device. Face processing raises privacy questions, and an older avatar that smiles or frowns in response to a contribution choice can slide from clarity into emotional pressure. A pension provider using this research would still need to show fees, uncertainty, trade-offs and assumptions plainly. Identification should not be used to disguise a poor product.

Procrastination can protect the present identity

Retirement planning asks for more than arithmetic. It asks someone to accept that their working identity will change, that their future needs cannot be known precisely and that today’s resources may have to serve a person they cannot yet picture.

Delay can therefore be a way of avoiding an identity question as much as avoiding paperwork. This does not make every postponed decision wise or harmless. It does make shame a blunt response. Calling the behaviour lazy adds moral pressure while leaving the psychological distance intact.

The older-face studies are useful because they make that distance visible. Their limits are just as useful: an image can briefly change a hypothetical choice, but it cannot create disposable income, simplify a pension system or decide what any individual should do with money.

Sometimes tomorrow’s person becomes easier to consider once they stop being an abstraction.