Health and Elderly Care Expenditure in an Aging World Leslie Mayhew RR-00-21 September 2000 International Institute for Applied Systems Analysis, Laxenburg, Austria Tel: +43 2236 807 Fax: +43 2236 71313 E-mail: publications@iiasa.ac.at Web: www.iiasa.ac.at International StandardBook Number 3-7045-0139-5 Research Reports, which record research conducted at IIASA, are independently reviewed before publication. Views or opinions expressed herein do not necessarily represent those of the Institute, its National Member Organizations, or other organizations supporting the work. Copyright c 2000 International Institute for Applied Systems Analysis All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy,recording, or any information storage or retrieval system, without permission in writing from the copyright holder. CoverdesignbyAnkaJames. Printed by Remaprint, Vienna. Contents Abstract iv Acknowledgments iv 1 Introduction 1 2 Health Care Services 6 2.1 MeasuringHealthExpenditure 6 2.2 MethodofAnalysis 10 2.3 More Developed Countries . . 13 2.4 Less Developed Countries . . 16 3 Disability and Welfare Services 19 3.1 Measuring Disability 19 3.2 MethodofAnalysis 22 3.3 More Developed Countries . . 24 3.4 Less Developed Countries . . 27 3.5 Disability and the Provision of Elderly Care Services 27 4 Conclusions 36 Annex: Overview of Method Used to Measure Disability 38 References 41 iii Abstract The world’s population is aging, albeit at different rates in different countries. The International Institute for Applied Systems Analysis (IIASA) is building an economic–demographic model for exploring the consequences of populationaging on the global economy. So far it has concentrated on impacts mediated through public and private pension systems. It now wishes to extend the model to cover other sectors whose provision is also highly age sensitive, including health and elderly care services. This report explores the consequences of population aging for these vital services and considers the basic mechanisms fueling their growth. These mechanisms fall into essentially two categories: The first is related to the biomedical processes of aging, which can lead to chronic illness and disability in old age. The second concerns the costs of treatment and long-term care, which in turn are a function of medical technologyand institutionalfactors, how services are delivered, and who bears the costs. Using simple but explicit projection methodologies, we project health care and disability-related expenditure in two major world regions, corresponding to more developed countries(MDCs) and less developed countries (LDCs). The key policy- related conclusions are as follows: • Aging will overtake population growth as the main demographic driver of health expenditure growth, but its effect will be less than that of technological and institutional factors. • Health expenditure will expand rapidly in LDCs (relative to gross domestic product) to reach levels currently observed in MDCs. • The number of people with disabilities will grow substantially, but will level out in MDCs by 2050 (earlier for all but the oldest age groups), while the number of people with disabilities in all age groups will continue to grow in LDCs. Assuming that most care for the disabled continues to be provided by the family and community, projected increases in disability-relatedexpenditure are modest. Acknowledgments I am grateful to my colleagues at IIASA for the stimulating discussions on the issues raised in this paper, particularly to Landis MacKellar, who heads IIASA’s Social Security Reform Project. iv 1 Introduction The impact of population aging on the global economy is now a major issue. This report, a contribution to the project on global social security reform at the International Institute for Applied Systems Analysis (IIASA), focuses on health and elderly care services (MacKellar and Reisen, 1998; MacKellar and Ermolieva, 1999). While these expenditure areas are less economically significant than pen- sions, the other main area of impact, they still account for over 10% of gross do- mestic product (GDP) in developed countries. They are major consumers of public expenditure; they straddle the public and the formal and informal private sectors, and are sensitive to the size and age distribution of the population and to patterns of morbidity. Their growth and development over the past 30 years or so, however, are only partly explained by aging and population growth. More important are fac- tors such as technological change (new treatments and drugs), higher utilizationper capita, institutional behavior, higher labor costs, etc. Our focus is on population and aging because of the very different population trajectories in developed and developing regions and their different starting posi- tions. It is now firmly established, for example, that older people consume more health services per capita than any other age group except perhaps the newly born. On average, their ability to perform daily tasks slowly erodes until, at some stage, they become dependent on others for home help, or possibly residential care or long-term care in a hospital. The degree of dependency, and sometimes also the need for medication, reaches a maximum in the period just before death (Seale and Cartwright, 1994). The economic consequences are therefore varied, directly or in- directly involving the work place, households, and agencies in the public, private, and voluntary sectors (see, e.g., Jackson, 1998). Not surprisingly, governments are becoming increasingly aware of the need for coordinated policies in the fields of employment, pensions, disability, and health. Some trends, though, will pull in opposite directions. It is expected, for exam- ple, that future generations of older people will be better prepared to live indepen- dent lives into advanced old age, particularly with the aid of modern technology and medical breakthroughs such as body-part replacement, which may improve the quality of life for some. There is some evidence that older people already have 1 2 healthier lifestylesand are better educated and informed than previous generations, with the result that the threshold for frailty and disability is being pushed later into old age in some instances (ONS, 1997). Estimates of future health- and disability- related expenditure depend crucially on whether the longevity revolution is adding healthy life years or years of illness and dependency to the human life span. Despite the uncertainty arising from countervailing forces, and certainly based on experience over the past 50 or so years, it is expected that demand will continue to grow and that health care services will continue to consume a rising share of GDP in all major world regions. To some extent, this merely reflects the changing consumption basket of aging societies (in the case of the more developed coun- tries, or MDCs) and societies undergoing structural economic and social change, including rapid health transition(in the case of less developed countries, or LDCs). A rising health-sector share of GDP is not necessarily an adverse trend (Aaron, 1996). However, the health sector’s increasing claim on resources is not without consequences for the real economy and represents an important index of structural change. While in some countries health systems confer universal coverage, the same is not true of elderly care services, which continue to be dominated by care within the family unit or immediate community, the so-called informal sector. A central issue in this case is the extent to which services provided by third parties (state or private residential and nursing homes, etc.) in the formal sector should be paid for out of personal income, sales of assets, and so forth. Again, the picture varies substantially, even within countries, because of differences in income and social factors such as deprivation and home and family circumstances. The aim of this report is to provide greater clarity and a firmer empirical basis for analysisof these issues in the context of IIASA’s global economic–demographic model, which is aimed at the medium to long term. Using recently available data, we attempt to separate aging effects from other contributors to growth, focusing on aging and disability and the demands older people and the disabled make on health and other services. In IIASA’s model, the world is divided into two regions. One region comprises the MDCs and includes the newly independent countries in the European part of the former Soviet Union. This region accounts for 82% of world GDP, but only 22% of global population. The other region comprises LDCs and includes China, India, and the newly independent Central Asian countries of the former Soviet Union. The differences between the economies and population age profiles of the two regions are telling, providing important clues as to the future impact of population aging on health and elderly care services. Figure 1.1 shows two population pyra- mids based on IIASA’s central population projections at two points in time, 1995 and 2050. The horizontal axes are scaled to show the percentage of population by age group rather than population number in order to emphasize the differences in 3 0-4 10-14 20-24 30-34 40-44 50-54 60-64 70-74 80-84 90-94 1 0 0 + Percentage of population 15 10 500 51015 Age Age Percentage of population 15 10 500 51015 0-4 10-14 20-24 30-34 40-44 50-54 60-64 70-74 80-84 90-94 1 0 0 + ''# # MDCs LDCs MDCs LDCs (a ) (b ) Figure 1.1. Population pyramids in (a) 1995 and (b) 2050. Population in each age group is expressed as a percentage of the total population in a region. Source: IIASA central population projections (Lutz, 1996). shape between regions and between years. In 1995, the MDC pyramid is highly tapered but still quite broad at the base, whereas the LDC pyramid is dominated by youngergenerations,withrelativelysmall percentages of older people. By 2050 the aging process reaches maturity in MDCs, with the majority of the population con- centrated in older age groups. In LDCs the pyramid is substantially transformed, resembling the MDC pyramid for 1995. 4 Sources of Information on Health and Elderly Care Services In considering the scope of health and elderly care services, we are dependent to a significant degree on the availability of suitable data in the private and public sectors. For this report, elderly care services are defined to include personal and social services such as social care in the home or in an institution such as a nursing or residential home. These services may include help with daily living, advice on financial affairs, companionship, and so forth. A key problem with elderly care ser- vices is how to evaluate the relative importance and size of each sector – whether state-funded, private, or informal. Details about the informal sector are especially scarce, and its economicvalueremains an unknownquantity, althoughit iscertainly very large (usually assumed to be over 80% of the total). The size of the formal sec- tor, which provides residential, day, and home (domiciliary) services and benefits in kind (such as meals), is better reported, but there remain many hidden transfers which are categorized elsewhere in national accounts. One example is the cost of residential care, part or all of which may be paid directly by the state or by the individual, or indirectly through social security benefits. These and other hidden transfers, the opportunity costs of unpaid care, and benefits provided in kind from many sources, including voluntary organizations, increase the difficulty of piecing together the elderly care jigsaw puzzle. There is no single or complete source of data on all aspects of these issues. Expenditure in each country is sensitive to cultural, behavioral, and institutional factors; to morbidity and mortality profiles; and to the level of economic develop- ment. Set against this, however, are the similarities in demography within each ma- jor region, the increasingly shared experiences of medical advances, and common outlooks and values, for example, in terms of national and international policies toward the disabled, in which emphasis is on equality within society.[1] It should be noted that while the picture that emerges is coherent and persuasive, it is built up partly from information available in every country and partly from fragmentary information from one or more countries that has been extrapolated to the rest of the region. It follows that the structure of the approach is as important as the results themselves, because the framework, including the IIASA model, can be updated as new and better information becomes available. A key source of information for this report was IIASA’s central scenario for world populationprojectionsfrom 1995 to2100(Lutz, 1996), althoughfor the most part we concentrated on the period to 2050, for which information was the most re- liable. Also invaluable were Organisation for Economic Co-operation and Devel- opment (OECD) databases covering health and social expenditure, including some information on activity levels and unit costs (OECD, 1998a, 1998b); the European System of Social Protection Statistics(Eurostat, 1996, 1998); United Nations (UN) 5 and World Bank data (UN, 1998; World Bank, 1993, 1999), especially macroeco- nomic and some health expenditure data for LDCs and miscellaneous sources and studies drawn from countries as diverse as the UK, USA, Canada, Australia, Fin- land, Japan, and China; and relevant conference proceedings. There were major shortcomings with respect to health and disability data for LDCs; consequently, key issues are only scratched at the surface. In the case of MDCs (comprising OECD countries and countries in Eastern Europe and the former Soviet Union), the analysis prior to 1995 is based on OECD databases only. The results presented are therefore a mixtureof thefirm and not-so-firm, the rel- atively precise and the merely indicative. Therefore, where necessary, appropriate assumptions and qualifications are spelled out. To a significant degree, this report builds on established trends over long periods, relatively stable features of the pop- ulation such as the onset and prevalence of disabilities, and underlying trends in economic growth. No attempt is made to predict technological changes that may have an impact on the delivery of health care and other services, or major break- throughs in medical treatments that may otherwise have an impact on longevity, health service costs, and so forth. These are presumed to be subsumed in the un- derlying growth rate. Part 2 of this report considers health care services. Part 3 looks at disabilityand elderly care services. Conclusions are presented in Part 4. 2 Health Care Services 2.1 Measuring Health Expenditure Medical expenditure is high in the first few years of life and increases again in old age with the onset of chronic illnesses and disability. To determine the contri- bution of population growth and aging to future expenditure, we need to separate the proportion of growth attributable to population trends and aging from growth attributable to other causes. The OECD publishes data on health expenditure per capita in selected older age groups as a ratio of expenditure in the 0 to 64 age group (OECD, 1998a). Although there are many gaps, a coherent picture emerges across countries showing expenditure in older age groups to be significantly greater than that in other age groups apart from the very young (see van der Gaag and Preker, 1998; European Commission, 1997). Data from England and Wales (see Figure 2.1) are consistent with the wider OECD picture and have the advantage of being available in time series over the entire age spectrum. They are also consistent with general examples provided by Cichon et al. (1999). Although the period is relatively short (1982–1993), the data are remarkably stable in most age groups. An exception occurs in the case of the 85+ age group, where the increase and subsequent downturn in the mid-1980s marks a change of policy concerningtheappropriatenessof keepingvery old people in hospitals (we return to this point later). Otherwise, the flatness of the curves is noteworthy, especially given increases in health service utilization, changes in treatments, improvements in quality, decreasing lengths of hospital stay, and the growing use of, for example, day services. The stability evident in Figure 2.1 suggests that relative age-specific expendi- ture indices should be fairly stable over time, at least in MDCs. Table 2.1 presents such indices calculated from the data plotted in Figure 2.1, with the lowest age group (0–4) as the baseline. We will presently apply these indices to project how the changing age structure of the population in MDCs is likely to affect growth in health care expenditure. Note that our assumption is not that levels of age-specific per capita health care expenditure in Englandand Wales are representative of levels of expenditure in MDCs as a whole, but that the age profile of such expenditure is 6 [...]... that rU can be interpreted as the rate of growth of total health care expenditure normalized by an index of population size and structure 11 The underlying rate reflects technological change, changes in per capita utilization, shifts in the care provided, and other factors, whereas the demographic rate combines population trends and aging, and is designed to capture the health needs of a growing population... fallen to around half (Endo and Katayama, 1998), with the rest living either alone (approximately 45%) or in institutions The public, private, and voluntary sectors therefore complement the informal sector rather than dominate it and provide a variety of services ranging from long-term care in hospitals to day services, home help, and other benefits in kind such as transport and home meals The aim of... observed in reality, can be misleading 3.5.2 Projecting services in the formal sector The costs of informal-sector care consist mostly of the opportunity costs of time spent in unpaid caregiving at the expense of remunerated employment The relationship between caregiving and employment is complicated, and it is known that many caregivers choose to continue in paid employment, performing their caregiving... remain at home We used these distributions to provide exploratory indications of demand in MDCs for different types of care, assuming previous population projections and age-specific disability prevalence rates, and assuming the split between formal and informal care remains at 20:80 The results show a growth of between 40% and 45% in the number of clients in each care category However, to obtain the... growth is due to aging In the former, the aging component is only beginning to be felt, but its impact is increasing With an assumed underlying growth rate of 3% pa, total public expenditure on health care as a proportion of GDP is set to increase from 2.7% in 1995 to about 4.2% in 2020 and 6.9% in 2050, with the final percentage being higher when private expenditure is added in Doubling the 4.2% GDP... projected increases in demand 29 3.5.1 Supply of caregivers Only fragmentary information on the supply of caregivers in LDCs exists In these countries, most caregiving takes place within the family or local communities where there are strongly ingrained traditions of looking after people in old age In some communities, having children is still regarded as an alternative to establishing a pension (World Bank,... 8 4 A g e Figure 2.2 Relative health care expenditure by age group (age 0–4 = 1.0), MDCs (1995) and LDCs (1995–2050) compared regions and time periods are scaled so that the index is equal to 1.0 at the 0 to 4 age group, differences in the health care costs in MDCs and LDCs cannot be inferred As can be seen in Figure 2.2, the estimated age profile of health expenditure in LDCs is projected to evolve... from Murray and Lopez, 1996), showing the estimated percentage of deaths by major causes in different world regions in 2000 and 2020 In MDCs the majority of deaths are currently from noncommunicable diseases, whereas in LDCs 8 Table 2.2 Pattern of mortality in MDCs and LDCs, in 2000 and 2020 (projected) Source: Based on data from Murray and Lopez (1996), baseline scenario in tables 12a to 12h and 16a to... out in six countries in 1980 and 1993 (Australia, Japan, Canada, UK, France, and Norway), the OECD observed that this period had hardly changed over that time frame and that it was significantly higher for women than for men, with a combined average of around two years By combining these data with mortality schedules, it is possible to obtain an estimate of the number of severely disabled that is independent... Figure 2.3 Health care expenditure as a percentage of GDP in OECD countries, 1960 to 1997 Source: OECD, 1998a 2.3 More Developed Countries In this and the next section, we consider the application of the growth factor model to health care expenditure in both world regions In OECD countries, health care expenditure increased at a rate of 5.7% per year between 1960 and 1995 in real terms GDP, meanwhile, . transformed, resembling the MDC pyramid for 1995. 4 Sources of Information on Health and Elderly Care Services In considering the scope of health and elderly care services,. sources and studies drawn from countries as diverse as the UK, USA, Canada, Australia, Fin- land, Japan, and China; and relevant conference proceedings.