Research - 09.09.2026 - 10:00
How long will I live? And how much money will I need in old age? As life expectancy rises, this question is becoming increasingly important, both personally and financially. After all, anyone planning for retirement today is shaping a future whose duration is unknown.
The “Beliefs, Behaviour and Longevity” research project at the University of St.Gallen (HSG) is investigating how people deal with this uncertainty. Two studies show that people’s estimates of their own lifespan are by no means neutral. Younger people underestimate their probability of survival, whilst older people overestimate it. At the same time, despite these biases, these subjective expectations are surprisingly revealing: they contain information about the actual risk of death that cannot be fully deduced from objective health data alone.
For the first study “A Behavioural Gap in Survival Beliefs” Enrico G. De Giorgi, Professor of Mathematics at HSG, and Giovanna Apicella from the University of Udine investigated how people estimate their life expectancy. Their analysis is based, amongst other things, on extensive longitudinal data from the European SHARE study.
A characteristic pattern emerges over the course of a lifetime: younger people tend to be too pessimistic in their assessment of their chances of reaching a certain age. As people get older, this relationship is reversed: older people tend to overestimate their chances of survival.
One explanation lies in how people process changes in their health. A new diagnosis or a deterioration in health does not factor equally strongly into the life expectancy forecast for everyone. For younger people, a health-related shock can be more surprising and alter their assessment more significantly. In later life, health problems occur more frequently and are therefore perceived differently.
Added to this is a personal outlook on the future. Optimism or pessimism influences how health-related information is processed. “People do not process information about their health independently of their expectations,” says De Giorgi. A personal outlook on the future plays a particularly important role when it comes to estimating one’s own lifespan.
The second study “Do Subjective Survival Beliefs Improve Survival Prediction?”, by HSG researchers Florian Benkhalifa, Maximilian Jakob Arrich and Enrico G. De Giorgi, goes one step further: if people systematically misjudge their chances of survival, are their estimates of any use at all? The surprising answer is: yes. When subjective assessments are incorporated into models for predicting mortality, their predictive power improves. This holds true even when extensive information on physical health and other observable characteristics is already taken into account.
Extreme assessments are particularly revealing. People who estimate their probability of survival to be very low tend to actually have a very high risk of death. Extremely high figures also provide a signal. Estimates of around a 50 per cent chance of surviving the period in question, by contrast, are considerably less informative.
A person’s own assessment thus appears to consolidate information that is difficult to capture through individual health characteristics alone. Simple forecasting models that combine subjective survival probability with age, gender and background can therefore already be surprisingly informative. However, this also demonstrates that the reliability of models increases with a more comprehensive information base.
It is also interesting to see what people base their judgements on. Physical health plays a central role in both subjective assessment and actual risk of mortality. Where other factors are concerned, however, perception and statistical risk diverge.
For instance, people place comparatively greater weight on their mental health when considering their life expectancy, even though its significance for actual mortality is lower. Demographic and behavioural factors, such as physical activity, smoking and drinking habits, on the other hand, are more important for the real risk of mortality, yet are given less weight in personal assessments.
The crucial question is therefore not merely whether people view the future too optimistically or too pessimistically. It is equally important to consider what information they base their judgement on, how they weigh it up, and which factors they tend to underestimate.
This is where the economic relevance of research into life expectancy perceptions comes into play: those who believe they have only a few years left to live might save less for old age, retire earlier or use up their assets more quickly. Conversely, those who expect to live a very long life might work longer, save more or take out more comprehensive cover against what is known as the ‘longevity risk’.
Subjective life expectancy can thus influence decisions regarding saving, retirement, insurance and healthcare provision. However, the consequences may extend beyond traditional pension provision. A further study by the HSG research team shows that people's estimates of their own life expectancy are also linked to their investment behaviour: particularly pessimistic people underestimate their chances of survival and, at the same time, invest more cautiously. Misjudging one's own life expectancy can thus affect one's financial situation in old age via two interlinked channels – through pension provision and through one's willingness to invest assets.
At the same time, the results also suggest that simply providing people with more information about health and average life expectancy is not enough. What is crucial is how such information is understood, weighted and processed. For insurers and pension providers, this could mean taking psychological biases more fully into account when providing information about pension and life insurance schemes or longevity risks.
The research also opens up new perspectives for risk forecasting: subjective assessments could complement objective health data, particularly where comprehensive medical information can only be collected at great expense.
Living longer is, at first glance, a positive prospect. However, this can create a dilemma when it comes to financial planning. This is because we have to make decisions today for years of life that we do not know whether, or to what extent, we will actually live. The St.Gallen research shows the systematic biases to which people are subject in this regard.
Both studies are available to download online:
‘A Behavioural Gap in Survival Beliefs’ published in the ‘Journal of Risk and Insurance’ (onlinelibrary.wiley.com)
‘Do Subjective Survival Beliefs Improve Survival Prediction?’ accepted in ‘Demographic Research’ (papers.ssrn.com)
The project also resulted in two further papers:
‘Gender-inclusive financial and demographic literacy: Monetizing the gender mortality gap’ (onlinelibrary.wiley.com)
‘A Leveraged Gender Gap: The Combined Effect of Longevity Risk (Mis)-Perception and Financial Risk-Taking’ (papers.ssrn.com)
