Introduction
Most of the seven major industrial countries are now experiencing significant changes in their demographic structure. A persistent pattern of declining fertility and improving life expectancy has created major segments of the population that are already relatively aged or will become so in the near future. This substantial change in the demographic structure is likely to have far-reaching implications. As to the economic ramifications alone, it is likely to affect the size, structure, and dynamics of the labor force and may cause difficulties in accommodating an aging work force; it may significantly challenge the maintenance of sustained and buoyant growth; and it is likely to alter the demand for goods and services. As to implications for the public sector, it is likely to influence the demand not only for pensions, but for other social expenditures as well (on education, medical care, etc.,). Furthermore, such a change could pose serious financial problems as the working-age segment of the population shrinks in proportion to the retired segment heightening the likelihood of intergenerational conflict. In most countries, concern over the aging problem has led to considerable discussion and the enactment of specific policy measures in the areas of pensions and medical care. Nevertheless, despite the potential importance of this issue, none of the seven major industrial countries has undertaken a comprehensive analysis of the combined impact of an aging population on the various components of government expenditure.1
While the Fund has often been accused of dealing only with short-term aspects of the economy, its function of surveillance clearly extends to evaluating whether short- and medium-term policies should be influenced by long-term structural developments. Longterm demographic changes are clearly of this nature, signaling the need for anticipatory policy measures while potentially constraining current policy options. Such concerns led several Executive Directors in early 1984 to suggest that the Fiscal Affairs Department conduct a comparative study of trends in government social expenditure in the industrial countries, with particular attention to the implications of current demographic trends.
This paper examines the impact of prospective demographic trends on the level and structure of social expenditure by the governments of the seven major industrial countries (the Group of Seven) through the year 2025. It attempts to place these demographic factors in perspective with the other factors likely to influence the growth of social expenditure. It also reviews the key policy issues likely to emerge, both at an aggregate and a sectoral level, and the types of policies countries have initiated in trying to cope with the effects of these demographic trends.2
The focus is on government expenditure in the different social sectors. Social expenditure is defined to include, as a central core, medical care, education, pensions, welfare payments, unemployment insurance, and family benefits. A smaller residual category of social expenditure is included which differs across countries (see Chapter II). No change in the existing division of responsibility between the public and private sectors is assumed for any social program, nor is any evaluation made as to whether such changes would be more efficient in meeting given social objectives. Similarly, no attempt is made to provide a detailed examination of the financial viability of particular social programs, nor to speculate on the sources of financing that might be required to fill potential resource gaps. Neither is there an attempt to quantify the shift in the balance of overall demand for resources by the young, the working-age, and the elderly populations, as this will be influenced by the relative importance of the private sector in a given country in providing for the needs of the elderly and those of the young.
This paper is based on the results of detailed studies by the Fund’s Fiscal Affairs Department on the probable evolution of government social expenditure in each of the seven major industrial countries.3 While similar in overall design and methodology, the individual country studies were carried out independently, with the intention of accurately reflecting national views on the development of key demographic and economic parameters, the scale of existing social expenditure programs, and the factors likely to influence their evolution. In those studies, the specific national definitions of expenditure in any given social sector were used.
In this paper, some attempt has been made to eliminate areas of sharp incomparability, both in the definition of social expenditure categories and in specific assumptions on parameters, and to explain the factors underlying major differences in the size and structure of particular expenditure programs across countries. In addition, this paper seeks to incorporate the impact of recent policy developments, especially with respect to pensions, and to adjust the projections for the effect of any significant changes in the social expenditure base between 1980 and the present. Nevertheless, on balance the results largely reflect those of the individual country studies, and more emphasis should be placed on an examination of the trends in expenditure for a given country over time than on an excessive focus on the differences in the ratio of expenditure to gross domestic product (GDP) across countries at any given point in time.4
It should be emphasized that the results presented are projections, not predictions. While projections of the population structure for 20–25 years in the future have a reasonably high degree of credibility, beyond this time frame increasing uncertainty must inevitably be attached to the results. By evaluating the impact of alternative demographic scenarios, this study has attempted to capture the range of possible outcomes. Nevertheless, one should recognize the limits of this type of analysis. Equally relevant, though based on reasonable assumptions, many outcomes appear to be politically unsustainable, particularly when judged by the financing burdens that they would imply. Such results effectively signal the need for, and the likelihood of, changes in the institutions, procedures, real benefit levels, and policies associated with important social services and transfers. Equally likely, the projected demographic shifts could themselves induce reactions in the economy that are at least partially corrective. For example, the increased demands of an enlarged elderly component of the population are likely to create labor market pressures to keep more of the elderly in active employment beyond the (heretofore) “normal retirement age.”
Such changes, only some of which are immediately foreseeable, would almost certainly alleviate some of the financial burdens implied by the projections. Certainly in the area of pensions and medical care, considerable thought is being given in most of the seven countries to ways in which to stem the anticipated growth in expenditures, and this is being translated into concrete policy measures. The recent pension reform measures in Japan, the Federal Republic of Germany, the United Kingdom, and the United States, and the policies instituted to control medical costs in the United States and other countries are manifestations of these efforts. In fact, in some countries, the appearance of sustainability in the projections for some programs may be the result of the implementation of such recent policy measures specifically designed to address the rising financial burden associated with the aging of the population. In cases where the results are not sustainable, it is hoped that this study will help to signal the need for changes in social service and transfer programs.
The remainder of this chapter summarizes the principal results for each of the seven major industrial countries, with emphasis on the probable trend of the ratio of aggregate social expenditure to GDP. It also considers, in a sometimes speculative vein, the larger policy issues that are implied. Subsequent chapters examine the economic and demographic assumptions and the methodology of the study, the results of the projections for the major social expenditure programs—pensions, medical care, education, unemployment insurance, and family benefits—and some of the key policy issues in these programs. Inter alia, these include such issues as the effect on pension expenditure of a later retirement age or of linking pensions to net-of-tax rather than gross wages, and the impact on medical care expenditure of efforts to contain medical costs. The sectoral chapters also indicate the types of policy actions that have been introduced by the different countries to stem the likely growth in expenditure.
Principal Results of the Study
The study required demographic projections for each of the seven major industrial countries. In general, the baseline scenario assumes an increase in fertility rates5 and a modest improvement in life expectancies at birth over their present levels. Such scenarios are in most cases similar to the middle scenarios used by the national authorities. Alternative “greater aging” scenarios are also evaluated in light of the possibility that fertility rates may fail to rise above their current low levels and that medical progress may make it possible to sustain a somewhat longer life expectancy. In effect, the demographic situation implied by current demographic parameters is largely within the range implied by the two alternative scenarios. These projections are discussed in greater depth in Chapter III. In summary, there will be a striking increase in the share of the elderly and a reduced share of the young in the population in all of the countries over the next 40 years. In some countries, such as Canada and the United States, these changes will occur only after the year 2010. In others, they have already occurred and will be accentuated over time, as in the Federal Republic of Germany, Italy, and the United Kingdom. In Japan, these changes will come swiftly.
If it is assumed that governments simply maintain the real 1980 per capita level of social benefits for the different age groups, the increase in the total size and the change in the structure of the population would lead to significantly greater social expenditure by 2025, ranging from a 70 percent increase in Canada to a 361 percent increase in Japan. However, when viewed relative to the growth in GDP that would be implied by projections of the labor force and historically reasonable productivity assumptions, these increased expenditures appear financially manageable. In Canada, France, the United Kingdom, and the United States, the ratio of social expenditure to GDP would actually decrease; in the Federal Republic of Germany, Italy, and Japan, the increases would be moderate, ranging from 1 to 4 percentage points of GDP. In other words, the growth in government social expenditure strictly implied by demographic factors alone can be realistically financed. Unfortunately, to assume that real 1980 benefit levels can be held constant is not realistic.
In particular, in most pension programs real benefits are typically linked to real wage rates, and as the latter increase with productivity so would the former under present policies. Efforts to cut the linkage by significantly lowering the “replacement rate” (i.e., the ratio of a pension to the last period’s wage level) would require a change in the implicit pension contract that might be politically difficult to enact. It is equally difficult to expect that real salaries in the medical and education industries would fail to keep pace with the growth in aggregate productivity in the economy. Given the labor intensity of these industries, such wage increases are likely to be translated into an increase in real per capita expenditure. In fact, the experience of the past two decades suggests that it will be difficult simply to contain the growth of real benefit levels in medical care programs to that of productivity.
As a result, the projections of social expenditure in this study require assumptions about the relationship between the growth in real benefit levels and that of productivity in the economy.6 The specific assumptions underlying the benefit levels in the different programs, as well as the likely obstacles associated with cost containment in education and medical care are discussed in the following chapters. As much as possible, such assumptions attempt to take account of recent policies to contain costs or limit the growth of benefits. The implied growth in overall expenditure is thus less than would have been obtained with a simple extrapolation of the growth in real benefit levels experienced over the past two and a half decades. As will be seen, an important implication of this study is that there will be a need for additional policy measures to contain the real increase in per capita benefits, especially in those countries with high tax rates.
The principal conclusions for each of the countries are provided below. In brief, the projections to the year 2025 indicate a significant increase in the ratio of government social expenditure to GDP in most of the countries except Canada, and particularly sharp increases in most of the European countries and Japan (Chart 1 and Table 14). For illustrative purposes, the magnitude of the implied net additional financing burden can be expressed by relating it to the total wage bill in the economy. In effect, suppose the net increase in expenditure was to be financed exclusively through the payroll tax. Chart 2 provides a rough estimate of the increase over recent payroll tax rates that would be required to fund the additional burden.7 The tax rate would have to increase by 7 percentage points in France and by as much as 21 percentage points in Japan. This higher burden would be associated with a substantial increase in intergenerational financial transfers from the working-age population to the elderly.
Ratio of Expenditure to GDP for Total Government Social Expenditure, 1980–2025
(In percent)
Ratio of Expenditure to GDP for Total Government Social Expenditure, 1980–2025
(In percent)
Ratio of Expenditure to GDP for Total Government Social Expenditure, 1980–2025
(In percent)
Implied Change in Payroll Tax Rate Arising from Changes in the Government’s Social Expenditure Burden, 1980–2025
(In percent)
Source: Organization for Economic Cooperation and Development, The Tax/Benefit Position of Production Workers: 1979–1983 (1984).1 This chart assumes that the net change in government social expenditure since 1980 is financed by an increase in the payroll tax relative to the rate levied in 1983. The 1983 tax rate is calculated by dividing the employee and employer’s contributions by the income of the average production worker (the latter being augmented by the employer’s contribution).Implied Change in Payroll Tax Rate Arising from Changes in the Government’s Social Expenditure Burden, 1980–2025
(In percent)
Source: Organization for Economic Cooperation and Development, The Tax/Benefit Position of Production Workers: 1979–1983 (1984).1 This chart assumes that the net change in government social expenditure since 1980 is financed by an increase in the payroll tax relative to the rate levied in 1983. The 1983 tax rate is calculated by dividing the employee and employer’s contributions by the income of the average production worker (the latter being augmented by the employer’s contribution).Implied Change in Payroll Tax Rate Arising from Changes in the Government’s Social Expenditure Burden, 1980–2025
(In percent)
Source: Organization for Economic Cooperation and Development, The Tax/Benefit Position of Production Workers: 1979–1983 (1984).1 This chart assumes that the net change in government social expenditure since 1980 is financed by an increase in the payroll tax relative to the rate levied in 1983. The 1983 tax rate is calculated by dividing the employee and employer’s contributions by the income of the average production worker (the latter being augmented by the employer’s contribution).The timing of the increased pressure for social expenditure by government differs across countries, emerging before the end of this century in the Federal Republic of Germany, Italy, and Japan. In France and the United Kingdom, projections show only a modest increase in the ratio of social expenditure to GDP through the year 2010, with sharper increases between 2010 and 2025. In Canada and the United States, the expenditure ratio will decrease until 2010. In both countries, the ratio would then increase in the following 15 years, but in Canada the increase would still be very limited. In terms of specific programs, a steady increase is projected for the ratio of government medical expenditure to GDP in all countries. The timing of the change in the pension expenditure ratio will largely mirror the change in the share of the elderly in the population. The ratio of government education expenditure to GDP will fall in the next 15 years, stabilizing thereafter.
In reviewing the individual country results, one should again caution that countries differ both in their definition of what is included in “government social expenditure” and in the relative importance of the public and private sectors in its provision and financing. In countries where the private sector is relatively important in providing support for the elderly, focus on the growth in the government’s social expenditure will significantly understate the overall resource allocation implications of the growth in the share of the elderly in the population (see the section on Broad Policy Issues below).
Canada
While Canada appears to be subject to the same process of demographic change as the other six major industrial countries, reflecting a historically declining fertility rate and an increasing life expectancy, it is far less advanced in the process. Coupled with a relatively lower level of government social benefits, particularly in the area of pensions, and a correspondingly lower overall ratio of government social expenditure to GDP, the more dramatic effects of the aging process on the expenditure ratio are not likely to occur at least through the year 2025. This is particularly the case under current policy provisions, whereby the flat-rate pension (Old Age Security), family allowance benefits, and the Canada Assistance Plan are indexed to prices and not to earnings (and have been for many years).
In 1980, Canada’s government social expenditure ratio was among the lowest of the seven major industrial countries—20 percent of GDP. Equally important, the ratio of pension expenditure to GDP was substantially less than in any of the other six countries, reflecting relatively lower average pension benefits and a low share of the elderly in the population. Though its fertility rate has recently declined to a level comparable with the rates of Japan, the United Kingdom, and the United States, its age structure is much younger, particularly in terms of the share of the population under age 40 (matched only by the United States).
The declining fertility rate will gradually have an impact on the age structure of Canada’s population, even with a baseline scenario calling for a return by 2010 to the level of fertility (2.1) required to maintain a constant population size (from approximately 1.9 in 1980). It will be reflected initially in a decline in the share of the under-15 age group in the population by 2000. The structure of the population aged 20–64 will also shift sharply, with the share of the group aged 20–29 declining from 32 percent to 21 percent between 1980 and 2000, and with a significant increase in the share of the group aged 40–49. The significant increase in the share of the elderly and in the elderly dependency rate8 will principally occur after 2010. Between 2010 and 2025, the share of the elderly in the population will rise from almost 13 percent to almost 19 percent. The population will grow at a rate of over 1 percent annually through the year 2000 but will then markedly decelerate in the first quarter of the century. Assuming that Canada’s productivity growth becomes comparable to that in the United States, its real GDP growth will be close to 3 percent through 2000, and then will decline, reflecting the decelerating growth in the labor force.
The baseline scenario projects a decline in the ratio of government social expenditure to GDP in the next 15 years, with a resurgence in the ratio occurring only after the year 2010, and even then leading to only a modest increase over the 1980 level (equivalent to 0.5 percentage points of GDP) by 2025. The ratios to GDP of expenditure on education, family allowance benefits, and expenses under the Canada Assistance Plan are projected to decline over the whole period, with the education ratio reflecting demographic factors, and the others the continuation of existing policies whereby benefit levels are adjusted only for inflation. Only the ratio of expenditure on medical care to GDP will increase steadily over the period. The ratio of pension expenditure to GDP is expected to decline through 2010, and although it will rise by 40 percent between 2010 and 2025, this increase entails only a one percentage point increase in the ratio to GDP over its 1980 level. In the case of a significant change in benefits policy, this result would no longer hold. In particular, if Old Age Security, family allowance, and Canada Assistance Plan benefits were adjusted in line with earnings rather than prices, the overall ratio of social expenditure to GDP would rise to 24 percent by 2025, rather than almost 21 percent under the baseline scenario. Pension expenditure alone would be 6.6 percent of GDP rather than 4.3 percent. However, the government social expenditure ratio would still decline through 2010 (to 18.5 percent and 19.1 percent of GDP by 2000 and 2010, respectively).
If the fertility rate remained at its current level, there would be a more rapid rate of aging in the population. Coupled with the more pessimistic economic scenario, this would yield an additional increase in the social expenditure ratio of approximately 2 percentage points of GDP by 2025. In both scenarios, these results focus only on government programs; the increase in the overall social expenditure ratio, public and private, would be larger, reflecting the relatively important role of the private sector in the support of the elderly.
France
In 1980, the ratio of government social expenditure to GDP in France was one of the two highest in the seven major industrial countries, equal to that of the Federal Republic of Germany. In demographic terms, the share of the elderly in the population and the elderly dependency rate were both relatively high. The share of the young was comparable to that in the other six countries (though considerably higher than the share in the Federal Republic of Germany). Its fertility rate averaged 1.9 in 1982–83, though it has declined over the past two decades. Even under the relatively optimistic baseline economic scenario and with a demographic scenario showing no increase in life expectancy, the social expenditure ratio to GDP in France will rise through 2025, with the sharpest increase between 2010 and 2025.
There are significant differences between the results of the baseline scenario and those of the “greater aging,” more pessimistic economic scenario, particularly in the period 2010–25. The baseline scenario reflects the most recent official and available demographic assumptions made in 1978 by the Institut National de la Statistique et des Etudes Economiques (INSEE) and assumes a return to a replacement level of fertility by the year 2000. Life expectancy is assumed to show no improvement over its base period level, reflecting the continued lack of improvement over the past several years in mortality rates among the elderly, especially for males. Comparatively, this yields a life expectancy in 2025 that is several years lower in France than in the other countries. In contrast, the “greater aging” scenario holds the fertility rate at current levels (at 1.95, above that of any of the other six countries), but assumes a life expectancy closer to that of the other six countries through the projection period. The baseline and pessimistic economic scenarios differ primarily in the assumed unemployment rate achieved by 2025 (2.5 percent vs. 7 percent), and in the labor force growth rates implied by the alternative demographic scenarios.
The projections suggest a fairly stable population structure through 2010, with a gradual aging of the labor force and a modest decline in the population under age 15, but with the share of the elderly in the population remaining essentially unchanged. After 2010, a sharp increase will occur in both the share of the elderly and the elderly dependency rate, with the magnitude of change depending on the rate of improvement in life expectancy. Population growth will be moderate through the year 2000, and then will decelerate strongly thereafter. Real economic growth in the baseline scenario is projected at slightly over 2 percent annually, with the deceleration in the growth of the labor force offset by the decline in the unemployment rate. In the pessimistic economic scenario, real growth falls to less than 2 percent, between 2010 and 2025.
Both scenarios suggest a fairly similar picture for the evolution of the government social expenditure ratio through the year 2010. Together, expenditures on pensions and medical care are expected to rise gradually by 3–4 percentage points of GDP. The ratio of education expenditure will decline by 0.5 percent of GDP, principally over the next 15 years. The ratio of unemployment benefits will either rise or fall by approximately 0.5 percent of GDP, depending on the economic scenario. Beyond 2010, the results of the scenarios diverge more sharply. Under the baseline scenario, the social expenditure ratio would rise by an additional 2.5 percent of GDP by 2025; under the “greater aging,” more pessimistic economic scenario, it would rise by almost 7 percent of GDP, with these increases primarily arising from higher pension and medical care expenditure.
Finally, much of the rapid growth in the social expenditure ratio in France in the past two decades has stemmed from nondemographic factors. The projection model assumes that these sources of growth will be largely contained, reflecting the almost universal coverage of pension and health programs and recent declines in the relative price effect in the medical care program. If total social expenditure is allowed to rise at the rate experienced in the past decade, the social expenditure ratio would be substantially higher by the year 2000 than in these projections, making the financial pressures accompanying subsequent demographic developments more difficult to cope with.
Federal Republic of Germany
Among the seven major industrial countries, the Federal Republic of Germany had one of the highest ratios of government social expenditure to GDP—31 percent—in 1980. The aging of the population was also the most advanced, with a high share of the elderly in the population (both over 65 and over 75), the smallest share of the young, a high elderly dependency rate, and the lowest youth dependency rate. Its fertility rate of 1.4 was the lowest of the Group of Seven countries—significantly below that required to maintain a constant population size.
The results of the baseline demographic scenario indicate that, even with a significant increase in the fertility rate to almost 1.7, the aging of the population will continue, with the share of the elderly rising steadily to almost a quarter of the population by the year 2025 and the share of the young declining, albeit gradually. The average age of the labor force will sharply increase. The population will decline from 62 million in 1980 to 54 million by 2025 (in part reflecting an assumed emigration of 1 million persons by 1990). The labor force will also shrink, on average declining by 0.5 percent annually over the period. Even with a reduction in the unemployment rate over the period, the economic growth rate is likely to be slower than that experienced in the past decade.
The authorities have recently introduced measures to curb the growth in expenditure on pensions and medical care. However, the baseline scenario suggests a continuous increase in the ratio of government social expenditure to GDP, to 33 percent by the end of the century and to 38 percent by 2025. This is primarily due to an expansion in the pension ratio (from 13 percent of GDP in 1980 to almost 21 percent by 2025) but also reflects an increase in the medical expenditure ratio. Some savings will be realized in expenditures on education, unemployment insurance, and other social programs, but only in the next decade or so; no further significant declines in the expenditure ratios for such programs are expected after 1990. These results will be further exacerbated if the fertility rate does not increase and under a more pessimistic economic scenario, with the increased ratio of government social expenditure to GDP emerging between 2010 and 2025. The results clearly imply that if contribution rates are not to rise to excessive levels, a fundamental restructuring of the retirement system will be necessary, most likely implying some combination of lower pension benefits or a later retirement age.
Italy
Like the other European countries, Italy had a high ratio of government social expenditure to GDP in 1980—reaching 25 percent, with much of this devoted to pensions and medical care. The share of the elderly in the population was 13.5 percent in 1980, one fifth of the potential working-age population. As in the Federal Republic of Germany, the fertility rate (1.6) was quite low.
The baseline demographic scenario suggests that, even with an increase in the fertility rate to 1.9 by 2000, the aging of Italy’s population will continue, with the share of the elderly in the population rising to almost 16 percent by 2000 and almost 20 percent by 2025, more than one quarter of the potential working-age population. At the same time, the share of the young in the population will decline. The share of the higher-savings group aged 40–60 in the population aged 20–64 will initially decline; however, after 2000, the share of this group will rise at the expense of the group under age 40. Italy’s population will peak at 58 million around the year 2000, declining slowly thereafter. These demographic trends will be reflected in a slow growth in the labor force, averaging only 0.1 percent annually. Productivity growth is assumed to increase to 2.2 percent annually, similar to Italy’s European neighbors, implying real economic growth of over 2 percent annually (compared with 1.8 percent between 1974 and 1984).
The baseline scenario projects a steady increase in the ratio of government social expenditure to GDP over the period, rising to almost 28 percent in 2000, 30 percent in 2010, and almost 35 percent in 2025. The ratio of pension expenditure to GDP will alone rise from 12 percent in 1980 to almost 21 percent in 2025. The ratio of education expenditure to GDP will decline, but only through 2000. As in the Federal Republic of Germany, a “greater aging,” more pessimistic economic scenario will show the greatest impact after 2010, with the ratio of government social expenditure to GDP being higher by 1 percentage point by 2010 and 2.5 percentage points by 2025.
Italy is unique among the Group of Seven in that its social expenditure program is not fully on a pay-as-you-go basis and has already required debt financing. Future working generations are likely to be forced to bear not only the burden of a larger retired population, but also the debt-service burden arising from present deficits. Beyond the obvious measure of an increase in the eligibility age for pensions, most other proposals for change involve more fundamental changes in the social security and welfare systems and measures for cost containment in medical care programs.
Japan
The impact of demographic change on the Japanese economy is likely to be the most extreme among the Group of Seven, reflecting both the lateness with which Japan began the demographic transition to a lower fertility rate and a higher life expectancy, and the sharpness of the change that has been experienced (particularly the decline in fertility in the early 1950s). In 1980, Japan had the lowest ratio of government social expenditure of the seven major industrial countries—15.4 percent of GDP (less than half that of the Federal Republic of Germany), reflecting low ratios of expenditure on pensions and medical care. Japan also appeared the least aged, with the lowest elderly dependency rate (13 percent), the lowest share of the elderly in the population (9 percent), and the highest share of the young. Its fertility rate (1.7) was comparable to that of Canada, the United Kingdom, and the United States.
In the next 40 years, Japan will make the transition to a population structure closely approximating that of the European countries. Compared with other countries, what is striking is the rapidity and magnitude of the change, rather than the ratios of expenditure or the actual population structure that will emerge. Specifically, the elderly dependency rate will rise by 65 percent in the next 15 years (from 13 percent in 1980 to 22 percent by 2000) and will be two and a half times its 1980 level by 2025. In contrast, the youth dependency rate will fall from 35 percent in 1980 to 26 percent in 2000. The share of the elderly will rise from 9 percent of the population in 1980 to 15 percent in 2000 and 21 percent in 2025. With an increasing life expectancy, those aged 75 and over will constitute over half of the elderly population by 2025.
The structure of the population aged 20–64 will also change, with a decline in the share of the under–40 age group from 52 percent in 1980 to 45 percent by 2000. The population will grow rapidly through 2000 but then growth will sharply decelerate. The labor force will grow through 2000 but will then decline thereafter. The baseline economic scenario assumes that Japan will maintain its historically high rate of productivity growth of 3 percent. As a result, the real growth rate is projected at almost 4 percent through 2000, falling to under 3 percent thereafter.
Government social expenditure is projected to rise to 27 percent of GDP by 2025—still considerably less than that of most of the European countries in the Group of Seven. However, the rate of change would be considerably greater, with the social expenditure ratio rising by almost 40 percent through 2000 and by 76 percent through 2025. This would imply the need for considerable fiscal adjustment. Pension expenditures will still grow substantially, with the pension expenditure ratio expected to more than double by 2000 and triple by 2025. These increases in pension expenditure are, however, lower than they would have been in the absence of the 1985 Pension Reform Act. Savings in education expenditure are likely to be realized, but by less than 1 percent of GDP. In the event that fertility rates do not rise to replacement levels and under a more pessimistic economic scenario, the social expenditure ratio would be 1 to 2 percentage points of GDP greater throughout the period than in the baseline scenario.
United Kingdom
Government social expenditure in the United Kingdom was approximately 23 percent of GDP in 1980, significantly lower than in some of the other European countries, in part reflecting the relative importance of private pensions. Yet by 1980, the aging of the population had proceeded further in the United Kingdom than in any of the other major industrial countries except the Federal Republic of Germany. The elderly dependency rate of 23 percent was the highest of the Group of Seven, the share of the over-65 age group in the population was approximately 15 percent, and the share of the young was among the lowest. However, in contrast to the Federal Republic of Germany, the fertility rate of 1.8 was well within the range of most of the other major industrial countries, although still lower than that required to ensure a stable population size.
Only modest further aging of the United Kingdom’s population is likely to occur through the year 2010, as reflected by a shift in the relative shares of the young and those over age 65. The share of the very old (i.e., over 75) will rise more rapidly. Thereafter, the aging of the population is projected to accelerate, as the share of the elderly will rise from 16 percent of the population by 2010 to almost 19 percent by 2025, primarily at the expense of the share of the working-age population. The elderly dependency rate will rise from 24 percent to 29 percent over the same period. These developments will occur even with an assumed increase in the fertility rate to replacement levels by the year 2010. Given the existing population structure, only limited growth in the labor force can be expected for most of the period (averaging 0.1 percent annually) and real GDP is projected to grow by 1.5 percent annually, compared with the 1.1 percent annual rate in the period 1974–84.
Through the end of the century, virtually no change in the government social expenditure ratio is expected to occur, remaining at 23 percent of GDP. The ratio is then projected to rise to 24 percent by 2010 and to 26.5 percent by 2025. As in the Federal Republic of Germany and Italy, the expenditure ratio on education will decline through the year 2000 and then stabilize thereafter. The ratios of expenditure on medical care and pensions to GDP are projected to rise continuously through the period. The increases in pension expenditure are, however, considerably less than would have emerged in the absence of the reforms proposed in the December 1985 White Paper. If fertility rates remain unchanged, the effects would be principally felt between 2010 and 2025, with the social expenditure ratio rising by an additional percentage point of GDP by 2025 relative to the baseline scenario.
United States
The combination of a comparatively young population, by Group of Seven standards, and a relatively limited involvement by the government in the provision of medical and pension benefits is likely to limit the impact of demographic trends on the growth of government social expenditure in the United States. In fact, the ratio of government social expenditure to GDP—18 percent in 1980—is projected to decline through the year 2010. As in Canada, the aging of the population is likely to be relatively sharp thereafter, resulting in a steady increase in the social expenditure ratio. The impact on the overall allocation of resources will be considerably greater, reflecting the private sector’s provision of more than half of total medical expenditure and a quarter of total pension disbursements. Moreover, the shift in the distribution of expenditure among education, health, and pensions is likely to necessitate major structural changes in the distribution of the financing burden at the federal, state, and local levels of government.
In 1980, a relatively high share of the population was under age 40, with a particularly large group between ages 20 and 29. A high immigration rate supplemented a fertility rate that was relatively high (though below the replacement rate) in providing a buoyancy to the population growth rate. Under the baseline demographic scenario, the population should experience a relatively slow process of aging through 2010, with a gradual decline in the share of the young and an increase in the share of the age group 65 and over. There will also be increased aging of the labor force. Within the population aged 20–64, the share of those aged 20–29 will fall from 32 percent to 22 percent. After 2010, the process of retirement of the “baby boom” generation will be reflected by a more than 50 percent increase in the elderly dependency rate; the increase in the share of the elderly in the population will be matched by a decline in the share of the working-age population. The total population is projected to grow by over a third through 2025, reflecting the return to a replacement rate of fertility and continued immigration. The baseline scenario assumes that productivity will be sustained at 1.6 percent annually over the period, implying a deceleration in the growth of real GDP, reflecting the diminishing growth of the labor force.
The overall government social expenditure ratio is expected to decline slightly before the year 2010, reflecting an almost 20 percent drop in the ratio of education expenditure to GDP, a corresponding increase in that for medical care, and a decline in the pension expenditure ratio. Between 2010 and 2025, the social expenditure ratio is projected to rise by almost 20 percent, or 3 percent of GDP. However, the growth in the overall resources required to support the elderly should be significantly greater, given the private sector’s role in providing for the needs of the elderly.
Since the bulk of government medical expenditure is focused on the elderly, the sharp aging of the population after 2010 will imply a significant increase in the share of medical care expenditure financed by the government. Education expenditure is expected to fall from almost 30 percent of total government social expenditure to less than 20 percent, suggesting the need to shift some of the burden of revenue absorption from the local to the federal level of government or to shift expenditure responsibilities. Finally, it should be noted that these projections reflect recent measures to limit the future growth of pension expenditure, following the adoption of some of the recommendations of the National Commission on Social Security Reform, known as the Greenspan Commission (1983). This underscores the sensitivity of the results to projections of medical care expenditure, with the ratio of nonmedical social expenditure declining relative to GDP.
Robustness of Projections to Assumptions
For most of the countries in this study, the projections suggest that the prospective growth in social expenditure is likely to be daunting, even in those cases where this growth will not occur for many years. How robust are these estimates? To what assumptions do they prove particularly sensitive? What problems have been excluded from the analysis which could affect the dynamics of the growth in social expenditure? What are the key policy issues that will need to be addressed during the period? This section addresses these questions.
Demographic Assumptions
The study has attempted to delimit a range of alternative patterns of fertility and mortality through its use of two demographic scenarios: one that closely approximates the most recent official projections, the other embodying demographic assumptions implying “greater aging.” In most cases, the baseline assumptions should be regarded as being reasonable, if not optimistic (in terms of “slower aging”), to the extent that they project a significant increase in fertility rates to replacement levels. A continuation of fertility rates at their present low levels is embodied in the “greater aging” scenario. The projections of expenditure in the latter scenario generally suggest only a moderate accentuation of the patterns derived from the baseline assumptions.
The individual country assumptions also implicitly assume that significant differences in unemployment rates can persist among the European countries over a 45–year period but this should not markedly affect the results (see Chapter II). Alternative immigration scenarios have also not been modeled systematically. The range of immigration assumptions across countries is fairly wide. In the Federal Republic of Germany, it is assumed that a net emigration of a million persons, principally returning migrant workers, will occur over the period 1980–90, with no further movement thereafter—a net emigration rate of 1.6 individuals per thousand; in Canada and the United States, a net immigration rate of 1.6 and 4.2 individuals, respectively, per thousand is assumed. Immigration rules are an important, but controversial, policy instrument available to national decision makers that could affect the underlying demographic and economic parameters and have important sociological and economic consequences.
A change in national policies toward immigration, particularly in some of the European countries and possibly Japan, would obviously qualify these projections. It could change both the size and composition of the labor pool, the age composition of the population (and the associated dependency rates), overall fertility rates, the level and structure of demand for different social services, and the wage base for financing social expenditure programs and transfers.
Time Frame of the Study
The projections of this study focus on the period 1980–2025. Does limiting the time frame to 2025 significantly bias the results, by omitting any marked structural changes that occur in the years thereafter? This question is particularly relevant for Canada and the United States—the two countries where the effects of aging appear to emerge only after the year 2010. For the United States, the years 2025–40 are likely to see a further aging of the population, though the pace of aging, as measured by the increase in the elderly dependency rate, will decelerate.9 The composition of the elderly population is projected to shift markedly, however. Between 1980 and 2025, the share of the group aged 75 and over among the elderly is expected to rise from 40 percent to 43 percent; in the next 15 years, it should increase to 56 percent. While such a change will not have dramatic effects on pension expenditure, it does have implications for medical expenditure, as the very elderly have higher medical expenditure per capita. Fund staff estimates suggest that, under the baseline demographic scenario, the social expenditure ratio will rise by an additional percentage point of GDP and, under the “greater aging” scenario, by approximately 2½ percentage points, between 2025 and 2040. In Canada, the projections suggest a further increase in the share of the elderly and in the elderly dependency rate, which is likely to be reflected in a higher social expenditure ratio by 2035.10
Economic Assumptions
A number of important uncertainties are associated with the effects of an aging population on the structure and size of the labor force and on the rate of growth of productivity. Population aging will be accompanied by a slowdown in the growth rate of the working-age population; in the Federal Republic of Germany, the growth rate will actually decline, in some cases, this is assumed to lead to increased labor force participation and lower unemployment rates. At the same time, the labor force will progressively age (see Chapter III), in some countries over the next two decades, a marked shift will be seen in the relative shares of the labor force in the 20–29 age bracket vis-à-vis the 40–49 age group (e.g., Canada, United States) or vis-à-vis the 50–59 age group (e.g., Japan). In Japan, a sharp decline in the size of the age group 30–50 will be seen—from 52 percent of the population aged 20–64 to 43 percent. In other countries, such as the Federal Republic of Germany, the next 15 years will see a sharp decline in the share of the 20–29 age group in the labor force, compared with the share of those over age 50.
A simple life-cycle model could be constructed whereby such changes would influence the rate of technological progress and the rate of productivity growth in the economy, but the direction of causality would be ambiguous. Do the more fundamental innovations in basic knowledge derive from a large and vibrant younger segment of the population? Some argue that an individual’s productivity peaks between the ages of 30 and 45. There is empirical evidence that suggests that productivity increases with aging in the labor force, though the range of historical experience may not capture the changes in aging that are likely to occur in these countries (Wachter (1976), p. 127). On that basis, the sharp increase in the share of the labor force in these groups in Canada and the United States should bolster productivity in the next two decades. On the other hand, there may be limits as to how rapidly the rate of technological progress or the rate of productivity growth among the key industrial countries can diverge, particularly over a 45-year time frame.
Other life-cycle factors will have an impact on the prospective growth of productivity. Of the working-age population, the highest-savings groups are those between ages 40 and 60. In Canada and the United States, the share of these groups in the population aged 20–64 is expected to rise by 10 to 14 percentage points, compared with only 2–4 percentage point increases in France, Japan, and the United Kingdom, and declining shares in the Federal Republic of Germany and Italy. Does this suggest the likelihood of a change in savings and investment rates? As an added uncertainty, will perceptions of an aging population and doubts about the adequacy of government social insurance benefits induce a higher savings rate in the economy? Savings rates might increase, simply reflecting an individual’s awareness that life expectancy is increasing and that the amount of private retirement savings necessary to complement public pension benefits may need to be augmented.
In terms of the projections of the study, the key issue is whether changes in the growth of productivity will influence the growth in real benefits or per capita expenditure for the principal beneficiary groups. In the case of education and medical care, the projections tend to assume either that such expenditures rise at the same rate as productivity (as with education) or by a constant factor in excess of productivity (as with medical care), primarily reflecting the labor intensity of the production function of these services and the assumption that the specialized factors of production in these sectors maintain their relative income position. As a result, the projections for these programs are not particularly sensitive to the productivity assumptions. The driving force behind the growth in the ratios of medical and education expenditures to GDP is, in most cases, the difference between the growth in the number of beneficiaries and that of the active labor force.11
In the case of pensions, the relationship is a bit more complex. Certainly, it would be expected that the starting pension of any retiring person would be linked to the wage or salary level received in the years prior to retirement and thus would be affected by productivity trends. In the 1960s and 1970s, there was also a tendency to link the pension benefits of already retired individuals to the growth in real wages in the economy—in effect, correcting not only for the effects of inflation (e.g., at least maintaining the pensioner’s standard of living in real terms) but also adjusting it to maintain the pensioner’s relative income position as well. In recent years, there has been a withdrawal from the expectation that the latter adjustment would necessarily occur, so that a higher rate of productivity in principle, with no inflation, would not necessarily be associated with an adjustment of the pensions of those already past retirement age.12 This situation would imply that higher productivity growth would reduce the ratio of pension expenditure to GDP. On the other hand, if the economic climate changed and productivity rose substantially above current rates, the pressures on policymakers to adjust pensions could be particularly strong.
There are situations in which the factors determining the growth of productivity in the key productive sectors of the economy, and thus wage rates, may differ from the factors driving total GDP per employee. For example, suppose a nation perceived correctly that it would be necessary to increase its national savings rate and build up its stock of real capital in anticipation of the need to draw down on this capital in order to finance future social expenditure. This could lead to a higher capital-labor ratio but also to increased foreign lending and equity participations abroad, with the return on such capital being reinvested. The growth in income derived from foreign investment and lending would not necessarily be reflected in increased productivity in the domestic economy, nor in increased wage rates or costs in the various social service programs. In this sense, there would be a less direct linkage between the growth in total income per capita and productivity as a factor influencing wage rates. This might be a difficult policy for the group of industrial countries to pursue simultaneously, but in the national interest of individual countries, it might be a rational approach.
Sectoral Assumptions
Subsequent chapters (Chapter IV though Chapter VIII) discuss some of the specific policy measures that might be introduced to modify the linkage between the growth in productivity and that of real benefits or expenditure per recipient. For example, the “replacement rate” (i.e., the formula relating initial pension benefits to previous earnings levels) can be lowered and the initial age of receipt of a full pension can be raised. To control the growth of expenditure for medical care, limits can be placed on government spending, as in the case of the United Kingdom, the copayment rates of individuals can be raised, and individual medical voucher plans can be introduced in connection with prepaid health plans.
It is important to emphasize that such a weakening of the link between the growth in costs or benefits per recipient and productivity can have a fairly dramatic effect in terms of moderating the growth in social expenditure. Over a 45–year period, benefits per capita rising at an annual rate 0.5 percentage points less than assumed could reduce absolute expenditure by approximately 20 percent. That could virtually eliminate any increase in the social expenditure ratio for all countries in the Group of Seven except Italy and Japan, and it would imply a significant decline from the present ratio in Canada, France, and the United States. Although it would be difficult to achieve such a reduction in real benefit levels, it is still important to underscore the sensitivity of the projections to these assumptions.
In fact, it should be stressed that there are other offsetting pressures that will complicate the task of limiting the growth of costs within the bounds assumed in this study. This is particularly the case for medical insurance programs but also, over the medium term, for education as well. In the former, new technologies may lead to a rapid buildup in medical costs. Recently developed procedures to clean out arteries, to disintegrate kidney stones, or to transplant organs are likely to tap a strong, latent demand for new and expensive forms of medical care. In some cases, such technologies may reduce the overall cost of treatment for an illness, but this may be offset by the increased demand for care they may induce. The pressures for an expansion of medical expenditure even in an efficiently organized medical care system, will be great; withstanding these pressures will inevitably require difficult decisions, involving equity and ethics, as to who will receive financing for many of the expensive procedures or whether such procedures should be rationed.
In the education sector, the next 15 years will most likely see a sharp contraction in the demand for educational services, in view of the anticipated decline in the number of students. The projection model in this study effectively assumes the maintenance of constant teacher-pupil ratios in the face of such a contraction in almost all of the countries. Tenure rules in public education systems may be a significant obstacle to reducing the number of teachers employed.
Broad Policy Issues
Structural Financial Adjustments Related to Shifts in Demand
Adjustments are likely to be necessary in real factor and product markets associated with the shift in the structure of demand for social and other services (for example, out of education and other child-oriented services and products and into medical care and other products and services catering to the elderly). These changes are likely to be more significant than would be indicated in the projections. Even in countries where the government’s role is predominant in the provision of social services, there still remains an asymmetry between the young and the aged segments of the population in terms of their relative dependence on government intervention for their overall support. The only element of youth care involving substantial government expenditures appears to be education. A comprehensive accounting for the shifts in absorption of economic resources implied by the prospective demographic changes would have to trace also the relative decline in private expenditures for food, shelter, clothing, entertainment, and in some countries, medical care. This was beyond the scope of this study.
In the next 15 years, a need will also arise to develop alternative arrangements for the financing of social programs. Some countries, notably France, the Federal Republic of Germany, Italy, and Japan, will experience pressures either to increase tax and contribution rates or to reduce benefits in order to meet the increased burden implied by the potential growth of social expenditure. In some countries, increased tax burdens might be feasible, or politically acceptable, only if associated with reforms yielding a more equitable distribution of the burden.
Financing of social programs may require alternative or additional institutional arrangements. Such financing might involve the establishment of private schemes complementary to the existing state pension system, thus allowing for an increasing role of voluntary contributions in countries where private pensions play a limited role. Existing institutional arrangements that imply the emergence of surpluses in one fiscal account and growing deficits in others will also need to be reconciled. This is particularly the case with respect to the withdrawal of resources from education, which is largely the prerogative of state and local governments, and the addition of resources to pension and medical insurance programs, for which the expenditures tend to be managed at the central government level. It also applies in instances where cross-financing presently occurs between various retirement schemes (for example, those of wage earners vs, self-employed) and where the financial viability of cross-financing will become increasingly difficult.
In principle, in the European countries and Japan, shifting resources from the local level to the central government should not be difficult (though such a shift runs counter to recent trends, which have shown an increased role for the local authorities in some countries). Most resource mobilization occurs at the central government level, with the local governments receiving either transfers or designated shares of particular taxes. Inevitably, the intersectoral shift of resources will necessitate a renegotiation of these earmarking formulas, with obvious political ramifications. In Canada and the United States, the state or provincial and local governments are not only responsible for education expenditure but they also finance such expenditure principally from local taxes. In both countries, the need for a substantial increase in expenditure on pensions and medical care at the central government level will not occur until after 2010, so that social expenditure considerations will not be the factor dictating any obvious realignment of revenue bases between the central and local government accounts until that time.
An interesting issue that will arise in these two countries is the response of local governments to the decline in expenditure required in the education sector. Will there be a commensurate reduction in state and local government taxes, or will there be a ratchet effect, as increased funding of other goods and services is provided, with no diminution in local government tax rates? It can be imagined that increasing pressures by the central government will force the local governments either to pick up a larger share of the burden of financing of other social expenditures (for example, by requiring a higher share of financing of Medicaid outlays by the states in the United States) or to lower tax rates. In the United States, both the recent elimination of revenue sharing and the threat of elimination of deductibility from income tax of state and local government taxes may possibly be seen as a reaction to the need for restructuring of fiscal responsibilities.
New Demands and the Public-Private Mix
This study attempts to gauge the impact of demographic trends and other factors on expenditure for existing social expenditure programs. Less predictable, but no less certain, is that new candidates for government expenditure in the social sphere are likely to emerge. The most obvious will relate to the needs of the very elderly (i.e., those over age 80). This group will never be a large share of the population, but its absolute size will triple or quadruple in many countries over the period. As will be discussed in Chapter V, over and above the medical needs of this group, there will be a need for a dramatic expansion in long-term care facilities as well—a need that is largely not covered in most existing government social service programs.13
This need also highlights the important role of the private sector in the financing of social expenditure in some countries. In the United States, a significant component of medical and educational services, as well as the consumption by the elderly, is effectively financed from private sector resources. The projections in this study for the United States are likely to be reasonably comprehensive with respect to the acute medical needs of the elderly population (and some chronic care), given the coverage of Medicare and Medicaid, but private sector wealth and pensions account for a sizable share of the financing of the needs of the elderly. As a result, the increase in the share of total output that will need to be allocated to finance the remaining consumption needs of the elderly (including some chronic medical care) is likely to be considerably greater than that reflected in the projections of government expenditure on pensions. While private sector pensions are less important in the European countries, Canada, and Japan, the role of household savings and intergenerational transfers is still important in some cases. For example, in Japan, public pensions appear at present to be only a basic form of support for most elderly Japanese, with heavy reliance placed on wage income, other sources of income (presumably capital income), and support by their families (in fact, about two thirds of the group aged 65 and over live with their children). While receipts from government pensions are likely to become more significant as such pension programs mature, there will still be a need for additional private sector resources.
However, without understating the important potential role of private sector resources in supporting the elderly, it must be emphasized that the increasing share of the elderly in the population is likely to pose overall macroeconomic problems in financing the associated intergenerational transfer of resources, and these problems transcend the issue of the division of responsibility between the public and private sectors. Simply transferring increased responsibility to the private sector will not significantly diminish the share of resources in the economy that will be needed to support the elderly population, though it may lead to revised incentives in the present working population relating to savings and consumption.
Implications for Macroeconomic Surveillance
An important question that emerges from this study is whether reasonably predictable long-term structural financial problems in the major industrial nations (be they the growth in social expenditure or the implications of rising public debt-service obligations), with possible ramifications for the balance of payments, should either prompt current policy action by the authorities or imply serious constraints on the range of macroeconomic policy options presently available to them. The results of this study suggest that important policy problems confront some of the Group of Seven countries. For Italy and Japan, the problem is more urgent, either because of the rapidity of the likely change, as in the case of Japan, or when viewed in terms of an inability to finance even existing social insurance programs on a pay-as-you-go basis, as in the case of Italy. For others, such as France and the Federal Republic of Germany, a growing pressure on public expenditure may appear manageable in an aggregate financing sense, but will necessitate difficult policy measures in the areas of pensions and medical care and limit the sphere of action for other policies. In particular, actions to provide additional fiscal stimulus to these economies must contend with the fact that such a stimulus may increase the likely ratio of public sector interest expenditure to GDP at a time in the future when tax rates and social benefit regimes are likely to be subject to increased demand pressures. Consideration of the implications of such problems may also affect one’s views on the appropriateness of a country’s current macroeconomic and balance of payments stance.
The fiscal problems arising from social expenditure trends in the major industrial countries are likely to become far more serious after 2010, notwithstanding the great uncertainties surrounding projections so far into the future. While such problems may pose less urgent constraints on today’s policies, failure to take present action will most likely result in more serious policy dilemmas in the future. The problem of financing social expenditure demands in the first half of the next century is likely to engender important intergenerational conflict at that time. Clearly, there are policy actions that countries could be considering now, in anticipation, that would lessen the severity of this problem. Among these, the effectiveness of policies to raise national savings and investment rates should be explored.
There is clearly a question as to the adequacy of the current savings rate in those countries where the problem will become more serious after the year 2010 (see King (1983)). At present, are social insurance schemes effectively viewed by much of the present working population as substitutable for savings? Will the present rate of capital accumulation be sufficient to generate additional income to supplement government pension receipts when the financial pressures on government insurance schemes become intense? Policies to encourage higher savings by the current working-age generation will necessarily clarify and confirm the fact that this generation is likely to have to fund both its own and its parents’ retirement.14 The notion would be the actuarial one of ensuring that a sufficient capital stock is developed to finance some of the burden of supporting the elderly population, thus facilitating a lower overall tax burden on the wage incomes of the working population of that time. An equally interesting question is how to ensure an adequate rate of return from a higher national savings rate, particularly if a number of industrial countries attempt to increase their overall investment rates. Will this lead to an increase in investments in Third World countries? What are the issues associated with the ultimate repatriation of this capital? This gives rise to many other systemic questions on the evolving structure of world capital markets and trade flows.
Other current policy actions may be needed either to dampen the likely long-term financial impact of current trends (e.g., policies to contain medical costs) or to begin the process of adjustment to the fundamental structural changes in the economy that will parallel the aging of the labor force. The possibility of encouraging increased immigration may need to be carefully examined in light of the demographic pressures of an aging population.
