Although an aging population has profound effects on the economy, a declining savings rate is not a direct consequence of a large elderly population. Typically, the elderly have low savings rates, while the working age population in their prime have higher savings rates. Though it would seem logical that an economy with a large aging population would have a low savings rate, the evidence reveals that the nature of savings rate changes are not as simple as the aging popluation theory would suggest.
Michael D. Hurd in his article, The Effects of Demographic Trends on Consumption, Savings, and Government Expenditures In the US, focuses on the impact of a growing elderly population on the Social Security trust funds, government spending, and medical expenditures in the United States. Hurd explains that the current demographic change is a result of temporarily high fertility rates that produced the “baby boom” and the falling mortality rates that have increased life expectancy. Despite this large aging population, Hurd concludes that medical expenditures and tax increases related to Social Security will be manageable, because through his research he found that the change in the savings rate over the next twenty years would be negligible.
The commonly held belief is that if the elderly are dissaving and selling off their assets, then the working-age population must be buying those assets, and thus, their savings rate will decline in turn. Hurd calculates the saving rates in 1989 and 2020, based the amount of assets possessed by the elderly and the elderly dissavings rate. Using the given amount of assets held by the elderly in 1989, Hurd estimated the assets of the elderly 2020 based on an increase of 1.8% in the elderly population and an increase of 1.01% per year in GDP. The 1984 Survey of Income and Program Participation found that the rate of dissaving among the elderly is 2.9% of bequeathable non-housing wealth. By multiplying the rate of dissaving and the assets held by the elderly, Hurd determines the amount of “asset decumulation”- the amount of assets that are sold by the elderly. Hurd admits that his most important assumption in his research is that the dissaving rate is 2.9%. He justifies this rate because it was calculated in the mid-80’s when the economy was rather stable, thus Hurd asserts that the “rate of wealth change may well represent the desired long-run rate of change.” (Hurd p. 18) From the percentage of asset decumulation after taxes, Hurd determines that in 2020, given the assumptions of the model, that the savings rate wil fall; however, the savings rate will fall by only .8%, from 4.6% to 3.8%. Hurd concludes “ the fall in the household savings rate from 4.6% to 3.8% does not seem like a large change.”(Hurd p.19) Hurd illustrates that there is a lack of empirical evidence to conclusively support the theory that the aging population will have a negative impact on the savings rate.
Matthew Higgins and Jeffrey G. Williamson in their article, “Age Structure Dynamics in Asia and Dependence on Foreign Capital”, approach the issue of falling savings rates from a different angle: falling savings rate are not necessarily the result of aging populations, but of a growing dependency burden caused by youth dependency rates in a financially open economy. This article analyzes the Ansley Coale and Edgar Hoover theory (1958), which simply stated is: “sustained high fertility and falling mortality rates leave households and governments burdened with high youth dependency rates, and therefore unable to save more than a small share of household incomes or tax revenues.” (Higgins and Williamson p.261) Focussing on Asia, Higgins and Williamson research the validity of this theory.
Asia experienced a large surge in youth dependency rates beginning in the 1950’s due to a sharp decline in infant mortality in the 1930’s, while a high fertility rate persisted through the 1960’s and 70’s. When Asia reached its peak in youth dependency rates in the 1960’s and 70’s, the youth dependency burden in Asia was far higher than it had ever been in Europe or North America. Meanwhile elderly dependency rates have been relatively stable in Asia over the last century. In addition to population demographics, Higgins and Williamson analyze how current account balances have changed considerably in Asia. The authors point out that from the statistics, there seems to be an obvious positive relationship between dependency ratios and foreign investment. In the case of East Asia, there has been a large influx of foreign capital that has spurred growth, while dependency ratios have been falling. Higgins and Williamson conclude that, “the countries that have been the most successful in shaking off foreign capital dependency also seem to have undergone the most dramatic decline in dependency rates.”(Higgins and Williamson p.268) The question that arises is how much of this experience can be explained by dependency rates?
Coale and Hoover’s theory has faced much opposition since its inception. At the foundation of Coale and Hoover’s dependency theory hypothesis is the intuition that with a large young population and high dependency ratios, the tendency is to consume at the expense of saving. James Tobin’s life-cycle model in 1967 presented an opposing view that the national savings rate should actually increase with faster population growth. The life-cycle model supposed that faster population growth tilts the age distribution toward young, savings households and away from older dissavings ones. The life-cycle model, however, fails to recognize a youth dependency burden, and assumes that all the new additions to the population go straight into the workforce. (Higgins and Williamson p.269) A third theory by Andrew Mason (1988) integrates the life-cycle theory with the dependency theory, called the “variable-rate-of-growth effect” model. Mason’s model asserts that youth dependency can shift the timing of consumption in the life-cycle model from the child-rearing stage, when there is a high youth dependency ratio, to the non-childrearing rearing stage, when the youth dependency ration is lower. If consumption shifts to later stages of life, then aggregate savings will rise to a certain level based on the growth rate of national income and the dependency rate itself. Mason then analyzed the data from fifty different countries, and found a negative relationship betweens savings rates and youth dependency ratios, taking into account the interactive impact of dependency and income growth. Higgins and Williamson point out that Mason’s variable rate-of growth model has two major flaws. Firstly, the model is only able to describe the steady state relationship between dependency and savings rates, which is a tendency inherited from the life-cycle model. Secondly, the model ignores the factors of investment demand and assumes perfect capital mobility. For Higgins and Williamson’s research, they extend Mason’s model to include both open and closed economies, and in the case of an open economy they distinguish between savings supply and investment demand embodied in net capital flows.
From this model, Higgins and Williamson conclude that investment demand and savings supply each have a “center of gravity.” The center of gravity for investment demand should be earlier than that for savings supply. Investment demand should be most closely related to the size of the youth population, and savings supply should be most closely connected to the size of the mature adult population. The difference between the centers of gravity is determined by the openness of the economy to foreign capital flows. In an open economy, a shit towards a younger population will produce a tendency towards current account deficits, because savings rate will fall due to the increase in dependency burden, even though there is an increase in the investment from a growing labor force. For a closed economy, on the other hand, the younger center of gravity for investment would project a positive relationship between a young population and savings rate. Thus, Higgins and Williamson emphasize that the impact of the dependency burden on savings rate is determined by the openness of the economy.
Is population aging a major cause of falling saving rates? Higgins and Williamson, though their article does not address this question directly, would answer, not necessarily. According to their analysis, savings rates are not only determined by elderly dependants alone, but also by young dependents. In addition, the authors emphasize that in determining the cause of falling savings rates, one must consider the openness of the economy and the role of foreign investment. In the case of the United States, Hurd’s article through the analysis of projected changes in the savings rate further discredits a concrete link between aging populations and falling saving rates. The authors of both articles would agree that a single-causation explanation of falling saving rates fails to recognize the complexity of the factors at work.
Michael D. Hurd, The Effects of Demographic Trends on Consumption, Savings, and Government Expenditure in the U.S., NBER Working Paper, No. 4601, National Bureau of Economic Research, Cambridge, Massachusetts, December 1993.
Matthew Higgins and Jeffrey Williamson, “Age Structure Dynamics in Asia and Dependence on Foreign Capital” in Population and Development Review, Vol. 23, Number 2, June 1997, pp. 261-294.