Introduction
The 2000 revisions in United Nations population projections predict that most of the
developed nations will experience negative population growth rates in the 21st century, mainly due
to declining fertility. Fertility rates in many countries are already below the replacement level of
2.1 births per woman. With falling fertility rates and rising life expectancy, the population of the
world is expected to age more rapidly in the next 50 years than during the past half century. For
example, the ratio of the elderly to the total population in all developed countries is projected to
grow from 14.3 per cent in the year 2000 to 26.8 per cent by 2050. In Japan, the increase in the
ratio of the elderly to the total population will be even more striking: the projected increase over the
next five decades is from 17.2 per cent to 36.4 per cent. The dramatic nature of Japan’s
demographic shift relative to developed countries as a whole can be seen in visual terms in Figure 1,
based on the revised UN projections.
Most developed countries main a public pension system in some form, . Pronounced
population aging will put strong pressures on those systems. As is now widely understood, in most
countries the benefit payments made to an increasing number of elderly retirees will have to be born
by a relatively smaller number of nonelderly workers. A widespread expectation thus exists in
political and administrative circles that population aging will pose significant challenges for
policymakers in the decades to come.
A large literature devoted to demographic change has focused much of its attention on
population aging and its effects on saving, investment and growth. From a policy perspective, the
research has focused primarily on the increasing burdens that rising elderly dependency ratios could place on national budgets and pension systems and on the menu of possible options for reform.
Either for simplicity or a lack of awareness, however, most studies have failed to focus on the full
effects of declining fertility rates. Declining fertility not only raises elderly dependency ratios, but
also reduces youth dependency ratios (Figure 1). The effects of projected demographic changes for
both the short and the long run will thus ultimately depend on whether lower youth dependency
(which lowers total dependency) or higher elderly dependency (which raises total dependency) will
dominate. Which of the two dominates, and the timing and magnitude of the net effects, will
determine whether population aging represents a severe or moderate challenge for current and
future generations.
In a recent paper (Bryant, Faruqee and Velculescu (2001)), we concentrated on the analysis
of youth dependency, explicitly incorporating children and child support into our theoretical
framework and empirical model. We showed there that changes in youth dependency can have
significant implications for the dynamic behavior of macroeconomic variables in both a closed- and
open-economy context. In this paper, we build on our previous work by introducing a public
pension system into our model, thereby accounting for transfers from working adults to both ends of
the age distribution: child support transfers from parents to youths, and public-pension transfers to
the elderly.
The merit of our revised approach in this paper is that it allows us to capture the economic
effects of both the early and the latter stages of the aging process. Declining youth dependency, on the one hand, entails lower aggregate transfers to children for their support, freeing up resources for
other uses, such as higher consumption per adult. Higher elderly dependency, on the other hand,
reduces consumption per adult because the members of a relatively smaller labor force must provide
for a larger number of retirees.
This paper contributes to the existing literature on demographic change in two ways. First,
it analyzes the implications of population aging in an open-economy context, focusing on the cross
border and global effects of lower fertility rates when a public pension system is operative in the
economy. We examine the effects of alternative variants of public pension (“social security”)
systems on world saving, investment, exchange and interest rates, and the external-sector positions
of individual countries. The variants include a government hands-off approach (no public pension
system at all, so that individuals provide for old age entirely from their own savings) and alternative
conventions by which the fiscal authorities set pension taxes, elderly benefits, and the management
of imbalances between the two. Second, as mentioned above, our analytical framework explicitly
incorporates children and child support as well as the elderly, and thus allows us to study the overall
combined effects of changes in both youth and elderly dependency ratios. An analytical framework
that accounts both for openness to the rest of the world and for declining child support due to lower
youth dependency is likely to yield more robust conclusions about the net economic effects of
population aging than models previously used in the literature.
Section 2 of the paper presents our theoretical framework, which is based on the Blanchard
(1985)-Weil(1989)-Yaari(1965) model, modified to incorporate richer demographics, youth and
elderly dependency, a child support mechanism, and a public pension system. Section 3 provides
further details about the way in which the theoretical framework is incorporated into our two-region
empirical model (an abridged version of the IMF staff’s MULTIMOD). Section 4 presents
simulation results with alternative pension arrangements but no child support for an illustrative
population-aging shock, assumed to occur identically in all regions of the global economy (equivalent therefore to closed-economy analysis); this section builds intuition about the effects of
various pension arrangements on domestic macroeconomic variables. Section 5 continues to
exclude child support and explore the alternative pension arrangements but focuses on the crossborder
effects of a demographic shock now assumed to occur asymmetrically in only one region of
the world. Section 6 presents preliminary simulation results for the asymmetric, country-specific
shock using the full capacities of the model to incorporate both child support and pension transfers
to the elderly. Section 7 concludes and sketches future avenues for research.