The ageing ‘crisis’ has not and will not happen

Bunkum - it can and will happen as Demography and Debt collide. Ifyou believe as i do, that most if not all countries are technically "insolvent" if they took even a cursory look at their "forward liabilities/promises", there comes a point where they cannot meet those liabilities aka pensions and health care. To me, that is a full blown crisis getting worse by the day as global pollies pretend and extend the fairy tale.

Some good points made. It’s difficult to appreciate in retrospective data, but medical advancements recently are so impressive I don’t think politicians and social services are aware how big a change will come to ageing.

We can now do gene editing on PCSK9 as a once-off treatment that will lower someone’s LDL cholesterol lifelong. This will reduce heart attack and stroke. We have wearable devices that can pick up atrial fibrillation before a stroke occurs. There are now disease modifying treatments for Alzheimer’s.

My point is there are limited ways that people can die and we’re doing very well at treating them. Unforeseen downstream consequences may come as a shock to policy makers and service providers. For instance, I wonder if at some stage we have socially acceptable forms of voluntary assisted dying for individuals or couples toward the end of the life, with a mandatory “cooling off period.” Regardless, we will be forced to adapt.

Brilliant analysis that shows how simplistic use of statistics can be misleading. Of course we work longer and more dual incomes as costs are so much higher as are lifestyle expectations (reinforcing system that most don’t consider when looking at housing costs…much higher expectations now than 50 years ago). Anyway, great article, thanks

Whilst obviously people (baby boomers) working till later in life reduces the baby boomer hump. However the graph shows that a higher percentage of overseas born are in the key working age deciles which also reduces the hump as predicted. The percentages in the graph are of the total population and overseas born are about a third of the population. So you have to multiply the overseas percentages by two to get a comparison with natives to get their figure relative to the cohort.

Disagree with this analysis claiming higher wrok force participation of retirees precludes the need for immigration till now?

False as it's not just about workers and a demographic bulge due to immigration (false), but budgets and tax revenue to deliver public services for more seniors vs fewer taxpaying workers.

It ignores accepted OECD old age dependency formula of retirees 65+ vs working age 20-64 (OECD); 20% 2000, 30% now and 40%+ mid century.

There is no definition of 'immigration' nor 'migration' provided that includes the NOM net OS migration of high temporary visa holder churn over* eg. international students, backpackers etc.

The latter cohorts' populate working age data in the chart, but are temporary with employment restrictions.

The theme or argument that higher or existing workforce participation will preclude the need for immigration, is not explained, not forecasted into the future and not proven?

The same could lead to a UK PM Truss type government, budget and taxes which rattled the bond markets, otherwise could have crashed budgets, PS delivery, government and institutions, the desired outcome for radical right libertarians?

Interesting article but the graphs clearly show that economic dependency in both the UK and USA have not fallen by very much, if at all -both around 50% and selected graphs like these always raise my suspicion that the graphs have been cherry-picked (aka Ancel Keys and his fat/heart disease relationships).

But I do agree in many respects - the sky hasn't fallen down in places like Japan, despite both an aging and falling population.

Good stuff. Perhaps the bigger story ii seems both husband and wife have to work to maintain lifestyle. Are our kids more materialistic or has there been a significant drop in the standard of living since the sixties? Bigger houses, two cars, overseas holidays, children at private schools, coffees, eating out; even children’s activities. Go figure.

Thank you for an excellent article. I love it when established ideas get challenged by facts in the way you have.
Another aspect worth adding here is that the growth of many people's superannuation and wealth in recent times means that more of the aging population is still contributing to the economy than say, 20 or 30 years ago.
My parents retired in their 50's and were self funded and still paid tax for most of the 30+ years they lived on. I, on the other hand kept working til 70, but am also self funded and still paying tax.
Both my parents and I also contribute/d as volunteers.
No ageing crisis here!

I’ll separate the critique into data/measurement, causal claims, policy implications, and omissions/edge risks.
1. Data and measurement issues
• Age dependency vs economic dependency definitions. The article uses a conventional age dependency ratio (0–14 and 65+ divided by 15–64) and contrasts it with a simple economic dependency measure (non workers/population). Both are blunt instruments:
o Age brackets are arbitrary. The 15–64 window was historically meaningful but less so now; yet replacing it with labour market measures introduces other biases.
o Economic dependency as non workers/population conflates many groups: students, unpaid carers, disabled people, retirees, discouraged workers, and those not seeking work for other reasons. It does not distinguish between non working by choice (e.g., full time students) and non working due to lack of jobs or poor health.
o Labour force participation rates hide intensity and productivity. More people working at older ages does not automatically offset costs if they work fewer hours, in lower productivity roles, or require workplace accommodations. The article notes full time vs part time shifts but does not quantify hours or output per worker.
2. Causality and composition
• Working later ? same economic output. The article assumes that increased participation of older cohorts offsets the fiscal and care burdens of ageing. But fiscal pressure depends on tax revenue per worker and age specific public spending (healthcare, aged care, pensions). Older workers may pay taxes but also consume more healthcare; net fiscal impact depends on the balance.
• Health and productivity heterogeneity. Lifespan increases are not uniform. Gains in life expectancy may be concentrated among higher income groups; poorer cohorts may experience more years in poor health (morbidity). If additional years are unhealthy, healthcare and long term care costs rise even if participation increases.
• Timing and cohort effects. Baby boomers have distinctive lifetime labour and savings patterns. Their higher participation now may reflect cohort-specific preferences, wealth, or policy (pension eligibility changes). Future cohorts might behave differently.
3. Policy and institutional assumptions
• Assumes labour markets can absorb older workers. The article treats higher participation as a natural offset. But labour demand, age discrimination, retraining availability, and job design matter. If employers prefer younger workers or automation replaces routine older worker roles, participation gains may stall.
• Pension and healthcare systems are path dependent. Fiscal sustainability depends on benefit design, indexation, and political choices. Even with higher participation, pension liabilities accrued under previous rules can create transitional fiscal pressures.
• Immigration argument is oversimplified. It’s true immigrants age too, but immigration affects the age profile at arrival, labour supply, and fiscal flows (young immigrants can raise the working age share and contribute taxes). The article dismisses immigration as a solution too quickly; the short to medium term effects can be meaningful even if long term ageing persists.
4. Omitted risks and uncertainties
• Health care cost growth and technology. Medical innovation can both raise costs (new expensive treatments) and reduce morbidity. The net effect on public budgets is uncertain. The article treats medical progress as unambiguously positive without grappling with cost inflation.
• Care needs and informal care. Increased participation of older people may reduce informal care supply (older adults often provide childcare or care for spouses). If both middle aged and older cohorts work more, informal care networks may weaken, increasing demand for formal care services.
• Distributional effects. Aggregate ratios can mask distributional burdens. Even if aggregate economic dependency falls, some households or regions may face concentrated ageing related costs (rural areas, low income groups).
• International comparisons and measurement periods. The article shows trends over 35 years; structural breaks (policy reforms, recessions, pandemics) can change trajectories. The COVID shock, for example, affected labour participation and health in ways that may not be fully captured.
5. Rhetorical and normative framing
• “No crisis” vs prudent planning. Declaring there is no crisis risks complacency. The article’s rhetorical stance (ageing is a success) is valuable to counter alarmism, but it underplays legitimate planning needs: adapting workplaces, retraining, pension reform, and health system capacity planning.
• Selective evidence. The article emphasizes participation gains but gives less attention to hours worked, productivity, healthcare spending trends, and pension liabilities—metrics crucial to the fiscal story.
Mechanisms by which AI could change the ageing story
1. Raise output per worker: If AI increases productivity, the same number of workers can support more dependents. This is the clearest route to offset ageing costs without raising participation.
2. Change the nature of work for older adults: AI tools can make jobs less physically demanding and more flexible (remote work, assistive tech), enabling older workers to remain productive longer.
3. Reduce care costs via automation and diagnostics: Robotics, remote monitoring, and AI diagnostics can lower the labour intensity and unit cost of eldercare, shifting the fiscal calculus.
4. Create new employment opportunities: New industries (AI maintenance, data annotation, human AI collaboration roles) could absorb displaced workers, including older workers if retraining is effective.
5. Widen inequality and political backlash: If AI concentrates gains, governments may face pressure to expand social safety nets for displaced or low income older people, increasing fiscal burdens.
Likely frictions and constraints
• Retraining and lifelong learning capacity. Older workers need accessible, effective retraining. Without it, automation can increase unemployment among older cohorts.
• Institutional inertia. Pension rules, labour regulations, and healthcare funding are slow to adapt; benefits of AI may not be captured quickly enough to offset near term liabilities.
• Distribution of gains. Whether AI helps the ageing problem depends on tax policy and redistribution: GDP growth alone doesn’t pay pensions unless revenue is collected and allocated.
• Technological limits in care. Some aspects of care (emotional support, complex physical assistance) are hard to fully automate; human labour will still be needed.
Practical policy levers to manage the transition (brief, actionable)
• Tax and benefit reform to capture productivity gains. Consider progressive taxation of capital and high incomes, and mechanisms to tax AI driven rents (e.g., digital services taxes, robot taxes debated carefully).
• Invest in lifelong learning and mid career retraining. Subsidies, portable training accounts, and employer incentives to upskill older workers.
• Redesign work and retirement norms. Flexible hours, phased retirement, and job redesign to accommodate older workers’ strengths.
• Support AI in care with regulation and funding. Pilot programs for AI assisted care, with evaluation of outcomes and cost impacts.
• Targeted support for vulnerable cohorts. Safety nets for low income older adults and regions hit by structural change.
SUMMARY
The fiscal and social impact of ageing depends on hours worked, productivity, health status, care needs, and distributional outcomes—not just headcounts by age.
AI could materially change the calculus: if it raises productivity broadly, reduces care costs, and its gains are captured by public revenue, ageing becomes far less of a fiscal problem. If AI instead polarises labour markets and concentrates gains, ageing could become harder to finance for many households and governments.
Policy choices matter. Whether ageing is a manageable transition or a political economic crisis will depend on labour market policy, tax design, retraining systems, and how AI’s gains are distributed.