Economic implications of an ageing population

We project that the average age of the UK population will rise over the next 50 years. This box explores the implications that may have both for the structure of economic activity and the performance of the UK economy.

A common metric used to assess the degree to which a population is ageing is the old-age dependency ratio, which is defined as the population above a typical retirement age divided by the rest of the adult population. For the purposes of international comparison, we use 65 as the retirement age benchmark and 15 as the start of working age. Over the past 50 years, the old-age dependency ratio has increased relatively little in the UK, despite rising life expectancy (Chart A). This reflects substantial net inward migration by working-age people. This contrasts with most other G7 nations, which have experienced larger increases in the old-age dependency ratio over the same period. Over the next 50 years, in line with ONS projections for the UK, the UN projects the old-age dependency ratio to increase substantially across other G7 nations and for the global population.

Chart A: UN projections for old-age dependency ratios across G7 countries

Note: UK data is taken from ONS outturn and the 2024-based principal population projection. World and other countries use United Nations data and projections.

Source: ONS, United Nations, OBR

Measures such as the old-age dependency ratio give a broad indication of the degree of population ageing. But the economic implications of population change depend on a much wider set of factors, particularly around individuals’ participation and employment choices. An ageing population could influence several key labour market drivers:

• Older individuals have a lower average labour participation rate, reflecting health and retirement choices. All else equal, this means an older population could be associated with lower aggregate GDP, as proportionally fewer people are in work. But participation has been trending upward among older cohorts, in part due to policy decisions, such as a rising state pension age. And this effect could continue to provide a significant offset to the cohort effect of an older population. How far future increases in life expectancy are matched by longer healthy life expectancy will also be pivotal. In Chapter 3 we assume that half of all gains in life expectancy will translate to higher healthy life expectancy.
• Changes in demand for goods and services in an ageing population, in tandem with technological change, may also influence the composition of the workforce. There may be significant growth in demand for some occupations, like caring roles, while employers may face increasing constraints on hiring for physically intensive work from a smaller pool of younger workers. These changes will influence employers’ choices on substituting labour for capital, and the development of new labour-saving technologies.^a Whether older workers face frictions in switching occupations and sectors will also be important.
• As discussed in Briefing Paper No.9: Forecasting productivity, the impact of population ageing on labour productivity is highly uncertain. On average, older workers tend to be better-paid, which could be associated with higher productivity, but this varies significantly across individuals and occupations. Compositional changes in the workforce could both increase and decrease productivity. For example, individuals who exit the labour force early could be above or below average productivity, and future labour-saving technologies may affect roles with above- or below-average productivity levels.

OLG simulation of the economic impact of an ageing population

In our long-term projections, we account for the impact of an ageing population on labour participation using our cohort model. We do not make an explicit judgement about the impact of population ageing on productivity growth since, as described above, the evidence for the sign and size of the effect is not clear. However, we can produce a stylised scenario using the UK Overlapping Generations Model (UK OLG).^b OLG models are useful for analysing long-term demographic trends because they explicitly model households of different ages that make forward-looking decisions about their labour supply and consumption in light of wages, taxes, welfare transfers and their probability of survival. These household decisions feed back into economy-wide outcomes, with wages, real interest rates, and the supply of labour and capital determined by the responses of households, firms and government in general equilibrium. This allows the model to capture the wider macroeconomic implications of demographic change.

The left panel of Chart B sets out the results of a UK OLG simulation that takes a baseline steady state equilibrium in 2025 and layers on the projected demographic structure for the UK in 2075. To isolate only the impact of demographic changes, the simulation leaves total factor productivity (TFP), the state pension age, and the size of the state pension unchanged from 2025 values. But all these factors would also materially affect the economic incentives faced by an ageing population.

The key findings are that with the 2075 demographic structure, total average hours worked fall by nearly 8 per cent, primarily because on average older workers tend to work fewer hours. A higher proportion of older, more asset-rich households also increases the capital-to-labour ratio by over a tenth, pushing up productivity. This also raises the marginal product of labour, increasing equilibrium real hourly wages by just over 4 per cent.

Average hours worked conditional on age rise in this simulation (Chart B, right panel), with a pronounced increase among older households. This means the total fall in labour supply is entirely due to an older population structure. The rise in average hours worked by age is largely driven by a higher life expectancy, which increases the years spent in retirement that households must finance out of their savings. This effect is strongest for the oldest households, as their lower starting hours mean they face a lower utility cost of additional work.

In these model results, the impact of higher productivity and households working longer hours conditional on age, particularly at older ages, is still not sufficient to offset the compositional effect of an older population on total hours worked. As a result, real GDP per person with the 2075 demographic structure is 5 per cent lower. This does not mean that faster population growth is guaranteed to generate higher GDP per person. When the population rises, the capital stock takes time to adjust, which can temporarily lower the capital-to-labour ratio and drag on productivity growth. This happens in our higher population scenario discussed later in this chapter, in which higher births, lower deaths, and a slightly younger age structure still produce lower GDP per person than in the baseline scenario.

Chart B: Simulated impact of 2075 demographics relative to 2025 demographics

Note: This simulation holds TFP, the state pension age and the state pension level constant to estimate the marginal impact of the projected demographic change in our baseline scenario.

Source: OBR

Observational evidence on the economic impacts of an ageing population is limited, given this is largely expected in the future across advanced economies. Japan has already experienced substantial population ageing, and this trend is projected to continue in future decades (see Chart A). But despite a shrinking working-age population in Japan since the 1990s and relatively weak growth in aggregate GDP and GDP per person, growth in GDP per working-age adult and GDP per hour have been relatively resilient.^c In the UK, as in Chart A, the old-age dependency ratio has increased by around 8 percentage points since 1975, but there is not yet clear evidence that this has slowed economic growth.