This is an important book, one that provides a novel and comprehensive way of identifying and portraying concerns of particular relevance for economies with aging populations. The approach focuses attention on how four fundamental activities – working, consuming, sharing, and saving – vary over the life course, creating flows across generations. The pattern and evolution of these intergenerational flows raise areas of genuine concern with current arrangements for the transfer of funds and the provision of services. However, the approach followed in the book suggests ways of addressing concerns about population aging. Among its many contributions, this volume effectively dispels various myths related to possible “solutions” (pp. 27–28) and notes that “reform of transfer systems is … an alternative to pronatalist policies” (p. 102).
The framework for analysis, intergenerational transfer accounts, is largely the work of Ronald Lee, and he and Andrew Mason are the lead authors of this volume. Its 32 chapters describe the analytic framework, detail how estimates of the components are derived, and the 57 contributors provide parallel applications in relation to 23 countries at various stages of economic development.
The approach relies on the most recent available estimates of the average level of individual labour income and consumption at each age for each of the 23 countries. The age-income profiles reflect what is produced through participation in the market economy, as indicated by the before-tax return to employment and self-employment, including employer-provided benefits. Consumption is comprehensively defined to include goods and services that are publicly provided (such as health care and educational services) – and the rental value of owner-occupied housing – as well as those goods and services that are privately purchased. The difference between consumption and labour income, summed over all ages, is the life-cycle deficit. Building on the work of Willis (Reference Willis, Lee, Arthur and Rodgers1988), any difference between the mean ages of consumption and production shows the direction of intergenerational flows. If the mean age of consumption exceeds that of production, the intergenerational flows are upwards, towards those who are older. The converse is true as well. A positive difference that is sustained across generations would have to be made up either by drawing down assets held by the current generation or relying on future generations to make up the shortfall.
When summed across all individuals to obtain national aggregates, the measures feed into what the authors define as National Transfer Accounts (NTAFootnote 1). The NTA are “broadly consistent with and complementary to” the familiar System of National Accounts (SNA) (p. 59), but a major difference, aside from the measure of consumption, is that the NTA introduce age into the SNA framework. That more comprehensive framework can then be applied to estimate the current life-cycle deficit at the economy-wide level. The authors find that the flows generally favour the young in less-developed economies and the old in more highly developed economies. NTA can also be used to assess how the direction of intergenerational flows would change as the age distribution of the population changes if the current age profiles of income and consumption were maintained.
That future could be based on a realistic projection for, say, 2050 as in Figure 4.8 or on a purely hypothetical situation in which the population is growing at a constant rate (e.g., Japanese age profiles with steady-state annual growth at –1%, as in Figure 2.1). Projection for the future could also be based on a population with any age distribution of analytic or policy interest (for example, to compare how the directions and magnitudes of intergenerational flows across the various countries would differ if they all had populations with the same age distribution, as in Figure 4.4). The conclusion, in the case of most countries, is that the current situation is unsustainable. However, the authors identify many avenues on both the consumption and production sides through which the needed adjustments could occur. Of particular interest, they note the importance attaching to how old-age consumption is funded: when “through transfers [rather than assets] population aging imposes a heavier support burden on … the population of working age …” (p. 91).
A feature of this book, and one used to great advantage throughout, is the provision of measures that permit meaningful comparisons across countries and, indeed, even across societies ranging from hunter-gatherers to those that are the most economically developed. By way of example, the consumption of children (those aged 0 through 24) and the cost of human capital (defined as the sum of education and health expenditures on children) are expressed in relation to the average annual labour income of individuals aged 30 through 49. By that measure, the consumption of children in Japan is 22 per cent greater than in Germany even though both have total fertility rates of 1.3, and the net private cost of raising a child to age 21 is six years of prime-age labour income in the United States and nine years in Taiwan.
All analytical frameworks have their limitations, as have studies based on them. A significant concern with the NTA approach follows from the fact that individuals and not households are the basic units of analysis. In consequence, a way must be found to allocate expenditure among members of households in order to obtain measures of average consumption (and well-being) at each age. Although the approach adopted is both pragmatic and reasonable, it raises concerns. For public expenditures, such as those on education and health, which are not paid directly by households, the NTA use age-specific spending when those data are available in published statistics or infer the data from an aggregate level of expenditures. For private expenditures, survey estimates at the household level are allocated to individuals within the household based on relative weights: each child under age 5 is assigned a weight of 0.4, those 20 and older a weight of 1.0, and those in the 5–19 age range are assigned weights by linear interpolation from 0.4 to 1.0. Thus, an individual with a weight of w will be assigned consumption equal to C *w/W where C is total household consumption and W is the sum of all individual weights in the household. For instance, in a family of two adults and two children aged 3 and 7, each adult is assigned a fraction 1.00/2.88 of total consumption, and the children 0.40/2.88 and 0.48/2.88, respectively.
The concern is that by assigning weights to individuals, the economies of household size that result from the public goods nature of some of what is consumed within households (e.g., heating, cooling, and cable services) is overlooked. (The standard equivalence scale for such a household would be 2.00 rather than 2.88.) That is possibly a matter of considerable quantitative importance: drawing on the 2006 Census for Canada, for example, reveals that less than 20 per cent of persons aged 25 to 69 live in one-person households, as compared to 43 per cent of those 70 and older. As a result, the NTA consumption measure for older people may be overstated. This would appear to be one case in which some sensitivity analysis would be informative.
A second concern is that the “age profiles do not reflect the value of time devoted to home production or to care for children or the elderly”. As the authors recognize, “This omission undoubtedly distorts the intergenerational picture that emerges” (p. 84). Moreover, both these issues (equivalence scale and value of time) tend to overstate upward transfers and may thereby exaggerate the unsustainability of current arrangement patterns of NTA. Yet another concern is the treatment of bequests (because a method for estimating them has not been established; p. 52). The matter is of quantitative importance and would represent a downward transfer of resources, offsetting the substantial future increase in upward transfers that is anticipated.
A few small points: Table 1.1 shows that “human capital spending” (i.e., spending on health and education for those aged 0–24) exceeds total consumption for some countries in the table. How could that be? And why do the public and private shares of expenditure on health and education not sum to 100 per cent? Equation 2.1, which defines life-cycle wealth, is not consistent with the interpretation on page 35; the equation should have consumption less labour income, not the reverse.
Such minor points aside, Lee and Mason, together with the other contributors, have produced an important volume that provides us with novel and persuasive perspectives on the economic implications of population aging, perspectives that we should find both reassuring and challenging.
