RDP 2006-06: Ageing, Retirement and Savings: A General Equilibrium Analysis 1. Introduction

It is well known that there will be significant changes in the demographic structure of the world over coming decades. It is also well known that these demographic changes will have consequences for many aspects of the economy. Perhaps less well appreciated is that the economic consequences of ageing depend on the strength of the two factors that drive the ageing process: falling fertility and rising longevity. There are numerous studies addressing the economic effects of ageing.^[1] Bryant (2004) discusses some differences between declining fertility and increasing longevity in terms of their macroeconomic effects. However, no studies (to the best of our knowledge) have analysed the combined implications of changes in fertility and longevity in a general equilibrium model with endogenous retirement decisions.

While it is difficult to get a precise measure of the relative contribution of fertility and longevity to population ageing, we can get a rough idea by examining projections derived from demographic models. Figure 1 shows a projected rise in the median age for Australia's population of about 13 years from the start of the baby boom^[2] to 2051.^[3] The elevated level of the total fertility rate over the two decades or so of the baby boom actually helped delay population ageing in Australia, while one of the genuine causes of ageing is the lower levels of the fertility rate that prevailed from 1965 onwards. If fertility had not declined in this way, the Productivity Commission's model would have projected a rise in the median age of around 4 years – suggesting that a sizeable portion of the estimated rise of 13 years over 1946–2051 can be attributed to falling fertility. Somewhat less of the rise in the median age can be attributed to the rise in longevity. If longevity had been constant throughout the period, the demographic model would have projected a rise in the median age of about seven years.^[4] In short, both the drop in fertility and the rise in longevity will contribute significantly to ageing, but in Australia's case, the effect of the former is projected to be somewhat larger than the latter.

These two demographic factors have potentially quite different effects on the economy, largely because of their different implications for labour supply and demand. In particular, a decline in fertility, other things equal, reduces the size of the working-age population relative to the stock of capital, putting upward pressure on wages. In contrast, rising longevity – to the extent that it is also accompanied by improvements in health (at all ages) – encourages individuals to delay retirement, increasing the size of the working-age population and putting downward pressure on wages.

Of course, these immediate effects are not the full story. In the case of a drop in fertility, we need to account for adjustments in wages and rates of return to capital when determining the ultimate impact on labour supply and capital accumulation. And in the case of a rise in longevity, the initial downward impact that a longer lifespan has on the wage rate could even be reversed in the long run. The key feature of longevity is that, in addition to providing extra labour (with healthier workers able to work for longer), it can also add substantially to labour demand in the long run. This is because longevity is likely to raise the absolute number of years a person spends in retirement. To finance this longer retirement, individuals need to accumulate more wealth during their working years. These extra savings add to the capital stock, raising the demand for labour. So it is the potential for a relatively larger capital stock that means increased longevity could lead to higher real wages in the long run.

Whether the changes in labour supply or demand contribute most to the change in wages as a result of increased longevity will depend on a range of factors, including the extent of changes in fertility and longevity, the relationships between longevity and health, and preferences over consumption and leisure. This paper incorporates these factors in a general equilibrium model. Our goal is to investigate responses to ageing in an economy populated by rational, fully informed individuals, in which there is no taxation and no government intervention. In this way, our results can be seen as suggestive of ‘first-best’ responses to ageing.

To study how combined changes in longevity and fertility affect the economy, we augment an otherwise standard overlapping generations (OLG) model by incorporating endogenous retirement. We study the response of the economy to changes in fertility, longevity, and the two combined. We calibrate the model's parameters to match key features of the Australian economy and find that ageing, driven by either falling fertility or increased longevity, ultimately raises the capital-to-labour ratio, increases wages, decreases rates of returns, and increases income and consumption per capita. However, the transition to this new steady state is more complex than the long run suggests, with wages falling initially (and rates of return initially rising) in the cases where a rise in longevity contributes to population ageing.

The rest of the paper is structured as follows. Section 2 reviews related literature. Section 3 describes the OLG model. Section 4 discusses the calibration of the model. Section 5 analyses the results and Section 6 concludes.

Footnotes

See, for example, Kotlikoff (1989), Yoo (1997), Miles (1999), Brooks (2002), Bloom, Canning and Graham (2002), Bloom, Canning and Moore (2004) and Poterba (2004). [1]

This is the period 1946–1965 when the level of the total Australian fertility rate is generally considered to have been very high relative to its own past (Productivity Commission 2005). [2]

These projections were derived from data and models provided by the Productivity Commission. [3]

Notice that the sum of the implied changes in the median age due to longevity and fertility do not sum to the projected change in the median age of 13 years. The difference arises because the two demographic factors interact. Also, any estimate that isolates the contribution to ageing from individual factors is sensitive to the assumptions about the demographic variables over the forecasting period. [4]