0 topics selected
17 August, 2026
Beyond words goes here
Published in The Sunday Times on 16th August 2026.
For years, the demographic debate has been dominated by a single, ominous narrative. As populations age, workforces shrink, dependency ratios rise, government budgets come under increasing strain, and economic growth slows. Demographics, are destiny—a time bomb heralding an era of slower growth and tough economic trade-offs.
But a recent report from Goldman Sachs challenges this view with broad implications.
To be clear, the developed world is ageing and there is a serious demographic challenge. Over the last 50 years the median age in the developed world has increased from 30 to 43 years. In emerging economies it has increased from 19 to 30 years.
The challenge is most acute in places like China. Its population shrank in 2025, the fourth-year running. China had fewer births in 2025 than in 1776, when its population was a fifth of what it is now. As a measure of how serious ageing is, an app called “Are you dead” became the most popular paid Apple store download in China earlier this year. It was designed for people living alone: users must check in regularly, and if they fail to do so for two consecutive days, the app alerts a designated emergency contact.
Ageing is a consequence of two developments:
(i) Increased longevity – we are living longer. In the last fifty years, in the developed world, life expectancy has increased from 72 to 82 years. This means that an average individual born 82 years ago (in 1943) would live to be 82. But in 1943, the official life expectancy was only 61. Only time will tell how long an average person born today will live.
(ii) Declining fertility – global fertility peaked in 1963 at 5.4. It stands at 2.1 today. In the developed world it’s 1.5.
In China, even though the one-child policy was scrapped in 2016, the fertility rate is below 1. This has prompted pronatalist policies like state sponsored mass weddings and more bizarrely - higher taxes on contraceptives.
It’s commonly assumed that the replacement fertility rate is 2.1. But this assumes that mortality rates are fixed at current levels. If life expectancy increases over time, the replacement fertility rate could be much lower.
GDP is a product of the number of people in employment and the amount each produces. If the population growth rate declines, so will GDP growth. Population growth is 1% today, going to zero over the next fifty years. The main worry is that ageing will result in a decline in the working age ratio (those aged 15-64), i.e. more dependents than workers. It has declined from 67% in 1985 to 63% today and to 57% by 2075. The antidote to this is for people to extend their working lives. According to GS average effective working lives would need to increase by 5 years by 2075 to offset the impacts of population growth decline.
The average person who was 70 in 2022 had the same cognitive ability of a 53-year-old in 2000. So biological ages are coming down rapidly - 70 is the new 53. The good news is that the average effective working life has already increased by four years (from 34 to 38) since 2000. Part of this improvement has been through reducing the female participation fall off post childbirth.
So, extending working lives has more than offset the effects of population ageing on developed world employment. Employment rates (as a share of population) has risen materially since 2000 despite all the negative headlines around the decline in the working age population.
This has implications for a market that is positioned for an old age economy rather than a long-life economy where people consume leisure, education, financial products, digital services etc rather than simply hospitals, nursing homes and pharmaceuticals.
Early retirement tends to be concentrated in jobs that are more physically demanding. But if the number of jobs in this area decreases due to automation, the need for early retirement diminishes too. This of course plays into the big issue currently around the capex boom in Artificial Intelligence.
Jevons Paradox was first observed in the 19th century following improvements in the efficiency of steam engines. Rather than reducing coal consumption, greater efficiency lowered costs and expanded the range of economically viable uses, leading to an increase in overall coal demand.
If software engineers can accomplish significantly more with AI than was previously possible, firms may choose to employ more engineers to pursue projects and opportunities that were previously uneconomic or beyond reach. So far, the evidence points in that direction according to JP Morgan, with job postings for software engineers rising rather than falling.
Major technological advances—from industrial robotics and personal computing to the internet and smartphones—have generally expanded economic activity rather than simply substituting for labour. The result has often been a larger economic pie, creating new industries, jobs and opportunities alongside productivity gains.
It is still far too early to draw firm conclusions about AI’s long-term impact on employment. But if it's premature to declare that AI will usher in a new era of prosperity, it is equally premature to assume a doomsday scenario in which human labour becomes obsolete.
The nature of life (and economics) is complex, where cause and effect are hard to pin down. The demographic changes across the world over the course of the next fifty years will be profound. But to conclude based upon analysis of the first order effects that this is universally bad, is intellectually lazy. Profound positive change beyond what we are capable of imagining is equally possible.
Gary Connolly is Investment Director at Davy. He can be contacted at gary.connolly@davy.ie or on X at @gconno1.
It all begins with a simple, no obligation conversation.
For investors who are comfortable making their own investment decisions.
