Abstract
The earlier sections have given a description of past and expected future population changes as well as a partial equilibrium analysis of the channels via which ageing is likely to impact, including a detailed assessment of the implications for external capital movements. The present section synthesises this information, using ECFIN’s general equilibrium ageing model, and presents its central view of the global economic and financial market impact of ageing populations over the next 50 years. In terms of the overall effect of ageing, figures for both potential output and income per capita (i.e. living standards — GDP growth rates adjusted for changes in trend population) need to be considered since there are significant differences between both these economic performance indicators for the different world regions covered by the ageing model.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
The results presented in this section suffer from 2 main sources of uncertainty, firstly regarding the population projections (where recourse can be made to the pessimistic and optimistic variants as well as the central scenario) and secondly the general issue of model uncertainty. A number of different models (life cycle, bequest, buffer-stock) can be used to try to fit the historical data. These models are also crucial in assessing how consumers will react to the type of policy measures which governments will put in place to overcome the economic consequences of ageing. In this context, in additional simulation work carried out for the analysis in this book, the ageing model was run using • firstly, an extreme form of the life cycle model, where consumers are expected to exhibit perfect forward looking behaviour and to set in place an optimal planning strategy for handling the consequences of ageing; and • secondly, what is termed a myopic / Keynesian view of the world, where working age households do not adjust their savings behaviour in the light of the looming ageing crisis. These two views of the world were simulated since they represent in effect the opposite ends of a spectrum of possible outcomes which includes, somewhere along the continuum, the “buffer stock” and “bequest” theories of savings. In other words the most likely effect of ageing on the EU and world economies would probably lie somewhere in the middle (i.e. a weighted average of the two extreme cases) since as explained in Box 1, there is a strong case for suggesting that the life cycle view is limited in certain respects. It was because of the uncertainties attaching to the life cycle view that it was decided to do a series of simulations which would try to encompass the wide spectrum of consumption theories which presently exist in the mainstream economic literature. While simulations have been run using other theoretical specifications, such as for example the “bequest” motive, the results fell into the middle of the range provided by the life cycle and “Keynesian” models. The conclusion of these simulations, in terms of growth rates, is that ageing is expected to reduce the annual average rate of growth in living standards (i.e. GDP per capita) in the EU over the next 50 years, relative to the baseline, by between 0.3 and 0.5 of a percentage point, depending on which of the two models one finds the most credible, with the Keynesian specification giving the most pessimistic assessment. The most significant point to emerge from these simulation exercises is the fact that whichever consumption theory one adopts, it is clear that there will be significant negative growth implications from the ageing of EU populations. Given the greater degree of theoretical and empirical support for the life cycle view, this is the specification which has been used for all of the simulations presented in the present book. It is also the view of the authors that with longer life expectancy and with financial market liberalisation ensuring that more and more people can borrow on the basis of their lifetime wealth (i.e. the proportion of liquidity constrained households is falling over time) economic agents will increasingly be living in a life cycle world.
In the case of Japanese living standards, the decline in the growth rate of GNP per capita is 25% less than that of GDP per capita due to the cushioning effect of income from their foreign investments.
With more sluggish growth in the developed world and little global income convergence, such a relatively low rate of world potential growth is not unexpected given global interdependencies.
Differences in TFP growth rates for all of the 5 areas in the 1990’s are very small, ranging from 0 to a 1/4 of a % point. Large persistent differences in TFP growth rates over long periods of time would be unrealistic, with long-run convergence a more credible assumption.
Effects on capital intensity are likely to occur due to changes in savings and investment behaviour, with further changes being possible in the event of a partial shift to funding due to alterations to the proportion of the overall change in savings which is invested abroad or is invested in the EU’s domestic capital stock. These changes in capital intensity however only affect labour productivity in the model and not TFP which is assumed to remain unchanged.
Another interesting feature of Graph 36 is the fact that the increase in the savings rate of workers is so small, especially when one considers that workers are faced with the prospect of having to fund a significantly larger number of years in retirement. Perhaps part of the explanation is due to the other influences on savings which are at work, most notably the real interest rate and income growth rate effects. However, it is probably more likely that the real reason for this muted response of workers to the ageing crisis is the fact that for the bulk of workers in the EU (at least in terms of the EU as a whole since individual country differences are large), they will in effect be forced to save for their retirement in the coming decades via much higher social security contributions (SSC’s) which they will automatically have to pay into the public pension system to keep it solvent. When Graph 36 is adjusted for the additional SSC’s which workers will mandatorily have to pay to finance the higher public pension expenditure, then the household savings rate for workers increases dramatically to reflect the higher cost associated with retirement income provision in an ageing world. This issue of replacing a mandatory PAYG system for a regime based on voluntary contributions to privately run pension funds is an issue of fundamental importance to future savings and growth trends and is a subject which is addressed in greater detail later on in this book.
Unlike the 1960–2000 period this reduction in growth is due to labour force changes rather than productivity. The labour force effects are partly offset by changes in the capital-output ratio (i.e. capital intensity) since the predicted fall in the real interest rate should lead to an increase in investment. A potentially negative factor for investment could be the shift to services production which is generally less capital intensive compared with industry. In addition, even allowing for “broadband”, it is difficult to envisage equivalently large demands over the coming decades, compared with the past, for new public infrastructure investments.
This graph shows only the change in the age-related component of savings and investment relative to their share of GDP in 1970. For the historical period 1970–2000, there are therefore differences with the actual changes in both the savings-to-GDP and the investment-to-GDP ratios. The investment to GDP ratio cannot fall as much as the savings/GDP ratio since we assume capital intensity is increasing.
This view is corroborated on the basis of results from the London Business School’s “Global Entrepreneurship Monitor” which covers 29 countries around the world (2 1/2 billion people in total) and suggests that older societies tend to be less entrepreneurial, with evidence that the majority of “entrepreneurially active adults” are aged between 25 and 44.
Vintage models are a methodology for linking changes in technical progress to changes in the average age of the capital stock.
Health expenditure estimates are determined exogenously on the basis of the expected changes in the demographic structures of the various geographical regions. In addition, allowance is made for differences in terms of income elasticities and age-specific health expenditure profiles for the different areas. It should be underlined that the uncertainties with regard to health care expenditures are much greater compared with pension expenditures.
Whether real interest rates fall or rise will depend on the relative strength of the effects of reduced labour force growth (on investment) and a lower private savings rate: Labour Force Effects: Reduced labour force growth leads to lower output growth, which in turn results in a decline in the proportion of GDP which needs to be directed towards investment as the required capital stock is proportionately lower. In addition, reduced labour force growth would be expected to lead to an increase in the capital intensity of economies, with this resulting in a fall in the marginal productivity of capital and the directly linked real interest rate. Private savings effects: On the basis of the life cycle hypothesis, ageing populations would be expected to result in a lowering of the private savings rate, due to the change in the structure of the population towards a greater share for the low saving / dissaving pensioner population. Most empirical studies support the contention of a negative relationship between dependency ratios and private savings rates. If this is true then ageing should result in a lower savings rate and a scarcity—induced increase in both the marginal productivity of capital and real interest rates. In this central scenario, the empirics are pointing to the dominance of the labour force effects, with real interest rates predicted to fall from from 5 3/4 to 5% over the next 50 years.
This is an important point to bear in mind in the pension reform scenarios discussed in section 5 where a reduction in the generosity of the EU’s PAYG system is shown to have very significant effects on the evolution of the EU’s net foreign asset position, especially in the absence of an improvement in the growth and investment environment in the EU.
For example, according to the UNCTAD World Investment Report (2001), over the period 1998–2000, 75% of global FDI inflows and 85% of outflows were between the EU, US and Japan. A similar picture emerges with regard to stocks of FDI, with the EU, US and Japan accounting for roughly 60% and 80% of inward and outward FDI stocks worldwide respectively.
Net foreign asset (NFA) developments for all areas in the model are sustainable. There are no instabilities, although the model predicts that NFA positions need a considerable period of time over which to stabilise, with this slow stabilisation process especially evident in the case of Japan. This pattern however is not surprising given the extent of the demographic changes being envisaged, with stock variables such as NFA’s invariably taking time to settle down and return to more “historically normal” positions after such a large shock. However, in terms of net flows (i.e. the current account balance), the stability constraint is more clearcut since it is not possible for countries to run large current account surpluses or deficits for long periods of time and corrections are therefore inevitable. In this context, while Japan’s NFA position continues to rise for some decades, in fact its current account as a % of GDP would reach its peak around 2010–2015 and move steadily downwards over the subsequent period to 2050 In addition, these projected changes in current accounts / NFA’s over this period only take account of the age-related effects on growth and on savings and investment in the respective areas, with many other possible influences such as relative changes in debt levels, productivity or short-run cyclical influences not being allowed for. The above discussion is not in contradiction however with the widespread view that the present level of current account imbalances globally are unsustainable. Clearly the cyclical divergences created by the bubble conditions in the US over the second half of the 1990’s have to be corrected. However, in terms of underlying current account developments, there are sound reasons for believing that the divergences in current account positions in the US and Japan which had emerged over the period from the early 1980’s to the mid-1990’s are to a significant extent structural in origin (i.e. ageing related). They are therefore sustainable over the coming years and even decades, with for example the model estimating that of the present Japanese current account surplus of 2 1/2–3% of GDP, perhaps 1–11/22% could be related to ageing.
Labour productivity can be further decomposed into the contribution from capital accumulation (i.e. the amount of capital stock per worker) and that from total factor productivity (i.e. the level of technology in the economy). Over the decades there has been a clear downward movement in rates of capital formation globally and during the 1970’s a structural break in the growth rate of technical progress. Are developments in the second half of the 1990’s in the US suggestive of a turnaround in TFP growth rates ? Other interesting issues for discussion, from the perspective of ageing, include whether ageing has a negative or positive effect on TF and whether TFP is partly dependent on investment (i.e. vintage effects).
While not identical due to dependency ratio changes, the growth rates of labour productivity and GDP per capita are very similar over time.
Most of the present differences between the EU and the US are due to differences on the labour input side, with productivity per hour worked in the EU in the late 1990’s being of the order of 90–95% of US levels.
However, it is important in this context to differentiate between populations which are growing more slowly (i.e. lower fertility rates) and populations which are growing older (i.e. higher life expectancy). It is a key issue for analysts to decipher if smaller, but relatively older, labour forces in the future can be equally good at generating the efficiency gains as previous generations. In addition a further complication is in assessing the influence of a more highly educated labour force over time, with each new cohort of workers tending on average to be better educated than the retiring cohort.
This is a group average for the fast ageing countries and trends for individual members, especially the high income OECD countries, China and the east Asian countries, are very different to those of other constituent members such as Russia and the central and eastern European countries. As explained earlier, the slow ageing group is, on the other hand, broadly homogeneous in terms of population and income levels, with the majority of countries in this group being, low income, developing countries.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Mc Morrow, K., Roeger, W. (2004). Global Ageing Scenario 2000–2050. In: Welfens, P.J.J. (eds) The Economic and Financial Market Consequences of Global Ageing. European and Transatlantic Studies. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-24821-7_4
Download citation
DOI: https://doi.org/10.1007/978-3-540-24821-7_4
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-07355-7
Online ISBN: 978-3-540-24821-7
eBook Packages: Springer Book Archive