An Aging World 2001
Upcoming SlideShare
Loading in...5
×

Like this? Share it with your network

Share
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads

Views

Total Views
6,373
On Slideshare
6,372
From Embeds
1
Number of Embeds
1

Actions

Shares
Downloads
93
Comments
0
Likes
0

Embeds 1

http://www.slideee.com 1

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. An Aging World: 2001 Issued November 2001 P95/01-1 International Population Reports By Kevin Kinsella and Victoria A. Velkoff Demographic Programs U.S. Department of Health and Human Services U.S. Department of Commerce National Institutes of Health Economics and Statistics Administration NATIONAL INSTITUTE ON AGING U.S. CENSUS BUREAU
  • 2. ACKNOWLEDGMENTS This report was prepared by and activities of the Census Bureau (as Kevin Kinsella and Victoria A. well as to many other agencies in the Velkoff under the general direction of United States and worldwide) for the bet- Peter O. Way, Chief, International ter part of two decades. Programs Center (IPC), Population Within the IPC Aging Studies Branch, Division, Census Bureau. Research for Valerie Lawson was instrumental to and production of this report were sup- the completion of myriad tasks involving ported under an interagency agreement data compilation, verification, table and with the Behavioral and Social Research graph production, and general report Program, National Institute on Aging, preparation. Branch members Wan He Agreement No. Y1-AG-9414-01. and Jennifer Zanini also contributed to For their review of and/or suggestions this report. regarding this report, the authors are Frances Scott, Janet Sweeney, grateful to Emily M. Agree, Department Barbara Adams, and Arlene C. Butler of Population Dynamics, Johns Hopkins of the Administrative and Customer University; Nikolai Botev, Population Services Division, Walter C. Odom, Activities Unit, Economic Commission for Chief, provided publication and printing Europe; Richard V. Burkhauser, management, graphics design, and com- Department of Policy Analysis and position and editorial review for print Management, Cornell University; and electronic media. General direction James C. Gibbs, Assistant Center Chief, and production management were pro- IPC; Albert I. Hermalin, Institute for vided by Michael G. Garland, Assistant Social Research, University of Michigan; Chief, and Gary J. Lauffer, Chief, Linda Hooper, Aging Studies Branch, Publications Services Branch. IPC; Rose Maria Li, Behavioral and Social Research Program, National Thanks also are due to Lorraine Wright Institute on Aging; Jean-Marie Robine, of the Bureau’s Information and Equipe INSERM Demographie et Sante, Resources Services Branch, who cheer- Montpellier; Beth J. Soldo, Population fully and efficiently handled the acquisi- Studies Center, University of tion of a multitude of necessary sources Pennsylvania; Richard M. Suzman, and references. Behavioral and Social Research Program, Finally, we would like to thank those National Institute on Aging; Peter O. international organizations and associa- Way, Chief, IPC, Loraine A. West, tions engaged in the collection, tabula- Eurasia Branch, IPC; and tion, analysis and dissemination of data Richard Woodbury, National Bureau relevant to the aging of the world’s pop- of Economic Research. We also thank ulation. In particular, we would like to David Rajnes, Housing and Household acknowledge the efforts of the Economic Economic Statistics Division, U.S. Census Commission for Europe, the International Bureau, who provided useful data and Labour Office, the International Network other information regarding retirement on Healthy Life Expectancy, the and pensions. International Social Security A very special debt of gratitude is owed Administration, the Organization for to George C. Myers, former director of Economic Co-Operation and the Center for Demographic Studies at Development, the Statistical Office of the Duke University, who passed away in European Union, the United Nations 2000. Not only did Dr. Myers offer Population Division, the United Nations instructive comments on this report, but Statistical Office, and the World Health he also provided invaluable advice and Organization. guidance to the aging-related products
  • 3. An Aging World: 2001 Issued November 2001 P95/01-1 U.S. Department of Commerce Donald L. Evans, Secretary Economics and Statistics Administration Kathleen B. Cooper, Under Secretary for Economic Affairs U.S. CENSUS BUREAU William G. Barron, Jr., Acting Director
  • 4. Suggested Citation Kinsella, Kevin and Victoria A. Velkoff, U.S. Census Bureau, Series P95/01-1, An Aging World: 2001, U.S. Government Printing Office, Washington, DC, 2001. ECONOMICS AND STATISTICS ADMINISTRATION Economics and Statistics Administration Kathleen B. Cooper, Under Secretary for Economic Affairs U.S. CENSUS BUREAU William G. Barron, Jr., Acting Director William G. Barron, Jr., Deputy Director John H. Thompson, Principal Associate Director for Programs Nancy M. Gordon, Associate Director for Demographic Programs John F. Long, Chief, Population Division For sale by the Superintendent of Documents, U.S. Government Printing Office Internet: bookstore.gpo.gov Phone: toll free 866-512-1800; DC area 202-512-1800 Fax: 202-512-2250 Mail: Stop SSOP, Washington, DC 20402-0001
  • 5. 20 Questions About Global Aging (to test your knowledge of global population aging at the turn of the century) Answers appear on next page. 1. True or false? In the year 2000, children under the 11. True or false? Today in some countries life expectancy age of 15 still outnumbered elderly people (aged 65 at birth is less than 40 years. and over) in almost all nations of the world. 12. What are the leading killers of elderly women in 2. The world’s elderly population is increasing by approx- Europe and North America? imately how many people each month? a. Cancers b. Circulatory diseases a. 50,000 b. 300,000 c. 500,000 d. 800,000 c. Respiratory diseases d. Accidents 3. Which of the world’s developing regions has the 13. True or false? Elderly women outnumber elderly men highest aggregate percent elderly? in all developing countries. a. Africa b. Latin America 14. There are more older widows than widowers in virtu- c. The Caribbean d. Asia (excluding Japan) ally all countries because: 4. China has the world’s largest total population (more a. Women live longer than men than 1.2 billion people). Which country has the b. Women typically marry men older than themselves world’s largest elderly (65+) population? c. Men are more likely than women to remarry after divorce or the death of a spouse a. Japan b. Germany c. China d. Nigeria d. All of the above 5. True or false? More than half of the world’s elderly 15. In developed countries, recent declines in labor force today live in the industrialized nations of participation rates of older (55 and over) workers are Europe, North America, and Japan. due almost entirely to changing work patterns of a. Men b. Women c. Men and women 6. Of the world’s major countries, which had the highest percentage of elderly people in the year 2000? 16. What proportion of the world’s countries have a public a. Sweden b. Turkey c. Italy d. France old-age security program? a. All b. Three-fourths c. One-half d. One-fourth 7. True or false? Current demographic projections sug- gest that 35 percent of all people in the United States 17. Approximately what percent of the private sector will be at least 65 years of age by the year 2050. labor force in the United States is covered by a private pension plan (as opposed to, or in addition to, public 8. True or false? The number of the world’s “oldest old” Social Security)? (people aged 80 and over) is growing more rapidly than that of the elderly as a whole. a. 10 percent b. 25 percent c. 33 percent d. 60 percent 9. More than one-third of the world’s oldest old live in 18. In which country are elderly people least likely to live which three countries? alone? a. Germany, the United States, and the a. The Philippines b. Hungary c. Canada d. Denmark United Kingdom b. India, China, and the United States 19. True or false? In developing countries, older men are c. Japan, China, and Brazil more likely than older women to be illiterate. d. Russia, India, and Indonesia 20. True or false? In most nations, large cities have 10. Japan has the highest life expectancy at birth among younger populations (i.e., a lower percent elderly) than the major countries of the world. How many years the country as a whole. can the average Japanese baby born in 2000 expect to live? a. 70 years b. 75 years c. 81 years d. 85 years U.S. Census Bureau An Aging World: 2001 iii
  • 6. Answers 1. True. Although the world’s world’s growth rate for the 80+ But because older men work in population is aging, children still population from 1999 to 2000 much greater numbers than do outnumber the elderly in all was 3.5 percent, while that of older women, increases in major nations except six: the world’s elderly (65+) popula- female participation were more Bulgaria, Germany, Greece, Italy, tion as a whole was 2.3 percent than offset by falling male partic- Japan, and Spain. (compared with 1.3 percent for ipation. the total (all ages) population). 2. d. The estimated change in the 16. b. Of the 227 countries/areas total size of the world’s elderly 9. b. India has roughly 6.2 million of the world with populations of population between July 1999 people aged 80 and over, China at least 5,000, 167 (74 percent) and July 2000 was more than has 11.5 million, and the United reported having some form of 9.5 million people, an average of States 9.2 million. Taken togeth- an old age/disability/survivors 795,000 each month. er, these people constitute nearly program circa 1999. 38 percent of the world’s oldest 3. c. The Caribbean, with 7.2 per- 17. d. The share of the private sec- old. cent of all people aged 65 or tor U.S. labor force covered by older. Corresponding figures for 10. c. 81 years, up from about 52 private pension plans was about other regions are: Asia (exclud- in 1947. 60 percent in the mid-1990s. ing Japan), 5.5 percent; Latin However, not all employees who 11. True. In some African countries America, 5.3 percent; and Africa, are covered by such plans actu- (e.g., Malawi, Swaziland, Zambia, 3.1 percent. ally participate in them. and Zimbabwe) where the 4. c. China also has the largest HIV/AIDS epidemic is particularly 18. a. The Philippines. The percent elderly population, numbering devastating, average life of elderly people living alone in nearly 88 million in 2000. expectancy at birth may be as developing countries is usually much as 25 years lower than it much lower than that in devel- 5. False. Although industrialized otherwise would be in the oped countries; levels in the lat- nations have higher percentages absence of HIV/AIDS. ter may exceed 40 percent. of elderly people than do most developing countries, 59 percent 12. b. Circulatory diseases (especial- 19. False. Older women are less of the world’s elderly now live in ly heart disease and stroke) typi- likely to be literate. In China in the developing countries of cally are the leading cause of 1990, for example, only 11 per- Africa, Asia, Latin America, the death as reported by the World cent of women aged 60 and Caribbean, and Oceania. Health Organization. In Canada over could read and write, com- in 1995, for example, 44 percent pared with half of men aged 60 6. c. Italy, with 18.1 percent of all of all deaths occurring to women and over. people aged 65 or over. Monaco, at age 65 or above were attrib- a small principality of about 20. We do not know. Data for uted to circulatory disease. The 32,000 people located on the selected cities/countries are pre- percentage was virtually the Mediterranean, has more than sented in Chapter 5. Some liter- same for elderly men. 22 percent of its residents aged ature from developed countries 65 and over. 13. False. Although there are more suggests that the statement is elderly women than elderly men false; evidence from certain 7. False. Although the United in the vast majority of the developing countries suggests States will age rapidly when the world’s countries, there are that it is true. Both the Census Baby Boomers (people born exceptions such as India, Iran, Bureau’s International Programs between 1946 and 1964) begin and Bangladesh. Center and the National Institute to reach age 65 after the year on Aging’s Behavioral and Social 2010, the percent of population 14. d. All of the above. Research Program would be aged 65 and over in the year 15. a. From the late 1960s until most interested in empirical 2050 is projected to be slightly very recently, labor force partici- input from interested parties. above 20 percent (compared pation rates of older men in Understanding global aging is a with about 13 percent today). developed countries were declin- dialectical process. 8. True. The oldest old are the ing virtually everywhere, where- fastest-growing component of as those for women were often many national populations. The holding steady or increasing. iv An Aging World: 2001 U.S. Census Bureau
  • 7. Contents 20 Questions About Global Aging . . . . . . . . . . . . . .iii 3. Average Annual Growth Rate and Percent Change Over Time for Older Age Groups: Chapter 2000 to 2015 and 2015 to 2030 . . . . . . . . . .130 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .1 4. Median Population Age: 2. The Demographics of Aging . . . . . . . . . . . . . . . .7 2000, 2015, and 2030 . . . . . . . . . . . . . . . . . .132 3. Life Expectancy and Changing Mortality . . . . . .23 5. Total and Elderly Urban Population by Sex: Available Data From 1970 to the Present . . . .133 4. Health and Disability . . . . . . . . . . . . . . . . . . . .37 6. Sex Ratio for Population 25 Years and 5. Urban and Rural Dimensions . . . . . . . . . . . . . .49 Over by Age: 2000 and 2030 . . . . . . . . . . . . .136 6. Sex Ratios and Marital Status . . . . . . . . . . . . . .57 7. Marital Status of Older Persons by Sex: 7. Living Arrangements . . . . . . . . . . . . . . . . . . . .65 Selected Years 1970 to 1995 . . . . . . . . . . . . .137 8. Family and Social Support of Older People . . . .73 8. Support Ratios: 2000, 2015, and 2030 . . . . . .154 9. Educational Attainment and Literacy . . . . . . . . .85 9. Parent Support Ratios: 1950, 2000, and 2030 . . . . . . . . . . . . . . . . . . . . . .155 10. Labor Force Participation and Retirement . . . . .93 10. Labor Force Participation Rates by Age and 11. Pensions and Income Security . . . . . . . . . . . .115 Sex: Selected Years, 1970 to 1999 . . . . . . . . .156 Appendix A. Appendix B. Detailed Tables . . . . . . . . . . . . . . . . . . . . . . .125 Sources and Limitations of the Data . . . . . .163 Table Appendix C. International Comparison of 1. Total Population, Percent Elderly, and Percent Urban and Rural Definitions . . . . . . . . . . . . .167 Oldest Old: 1975, 2000, 2015, and 2030 . . . .126 Appendix D. 2. Population by Age: 2000 and 2030 . . . . . . . . .128 References . . . . . . . . . . . . . . . . . . . . . . . . . . .169 U.S. Census Bureau An Aging World: 2001 v
  • 8. CHAPTER 1. Introduction The United Nations designated 1999 poses myriad challenges to policy- of the total population in each cate- as "The Year of the Older Person,” makers in many societies. After the gory). The developing world, by thereby recognizing and reaffirming year 2010, the numbers and propor- contrast, still had a high proportion what demographers and many others tions of elderly, especially the oldest of children (35 percent of all people have known for decades: our global old, will rise rapidly in most devel- under age 15) and a relatively low population is aging, and aging at an oped and many developing coun- proportion of older people (10 per- unprecedented rate. Fertility decline tries.2 The projected increase is pri- cent aged 55 and over). and urbanization arguably have been marily the result of high fertility after What is less widely appreciated is the dominant global demographic World War II. It is secondarily, but that absolute numbers of elderly in trends during the second half of the increasingly, the result of reduced developing nations often are large twentieth century, much as rapid death rates at all ages; in most and everywhere are increasing. improvements in life expectancy nations of the world, there have been Well over half of the world’s elderly characterized the early 1900s. As we major reductions in the prevalence of (people aged 65 and over) now live begin the twenty-first century, popu- infectious and parasitic diseases, in developing nations (59 percent, lation aging is poised to emerge as a declines in infant and maternal mor- or 249 million people, in 2000). By preeminent worldwide phenomenon. tality, and improved nutrition during 2030, this proportion is projected The confluence of lowered fertility the 1900s. One focus of this report to increase to 71 percent (686 mil- and improved health and longevity is a look at the numbers, propor- lion).3 Many developing countries has generated growing numbers and tions, and growth rates (past, cur- have had or are now experiencing a proportions of older population rent, and projected) of the elderly significant downturn in their rate of throughout most of the world. As population. natural population increase (births education and income levels rise, Most people, for good reason, asso- minus deaths) similar to what previ- increasing numbers of individuals ciate the growth of elderly popula- ously occurred in most industrial- reach "old age” with markedly differ- tions with the developed, industrial- ized nations. As this process accel- ent life expectancies and personal ized countries of Europe and North erates, age structures will change. expectations than their forebears. America. Most developed nations The elderly will be an ever-larger Population aging represents, in one are in fact the demographically old- proportion of each nation’s total sense, a human success story; soci- est in the world today, and some eties now have the luxury of aging. may have more grandparents than 3 Throughout this report, projections of pop- However, the steady, sustained children before the middle of the ulation size and composition come from the growth of elderly1 populations also twenty-first century. In the early International Programs Center, Population Division, U.S. Census Bureau, unless otherwise 1990s, developed nations as a whole indicated. As discussed further in Appendix B, 1 There is a growing awareness that the these projections are based on empirical analy- had about as many children under term “elderly” is an inadequate generalization ses of individual national population age and that conceals the diversity of a broad age 15 years of age as people aged 55 sex structures, components of population group, spanning more than 40 years of life. For change (rates of fertility, mortality, and net cross-national comparative purposes, however, and over (approximately 22 percent migration), and assumptions about the future some chronological demarcation of age cate- trajectories of fertility, mortality, and migration gories is required. This report uses the follow- 2 The “developed” and “developing” country for each country. ing terms for component age groups: the elder- categories used in this report correspond direct- Projections, strictly speaking, are neither ly (65 and over); the young old (65 to 74 ly to the “more developed” and “less developed” forecasts nor predictions. Projections are “cor- years); and the oldest old (80 years and over). classification employed by the United Nations. rect” in the sense that they are actual results of In some contexts (e.g., older people in the labor Developed countries comprise all nations in mathematical calculations based on specified force), it may be most useful or necessary (due Europe (including some nations that formerly assumptions. Forecasts are projections that to data restrictions) to refer to the “older popu- were part of the Soviet Union) and North analysts judge to be the most probable end lation,” those 55 years and older. The term America, plus Japan, Australia, and New results. There can be alternative projections, “frail elderly” refers to people 65 years or older Zealand. The remaining nations of the world but it would be contradictory to make alterna- with significant physical and cognitive health are classified as developing countries. While tive forecasts. It may, however, be appropriate problems. This term is used to emphasize the these categories commonly are used for com- to develop numerical ranges for forecast values. fact that a majority of the elderly, especially the parative purposes, it is increasingly evident that Predictions have no formal statistical meaning; young old, do not have serious health prob- they no longer accurately reflect developmental they are related more to forecasts than to pro- lems. differences between nations. jections. U.S. Census Bureau An Aging World: 2001 1
  • 9. population. Elderly populations responsibilities change considerably the IPC and is funded in part by the also have grown because of world- with increased age. Behavioral and Social Research wide improvements in health servic- Program of the U.S. National An Aging World: 2001 is the sev- es, educational status, and econom- Institute on Aging. IDB contents are enth major cross-national report in ic development. The characteristics readily available from the Census a Census Bureau series on the of the elderly are likely to be Bureau’s Web site; the direct access world’s elderly/older populations. increasingly heterogeneous within address is www.census.gov/ipc/ The first two reports, An Aging nations. Thus, a second focus of www/idbnew.html World (1987) and Aging in the An Aging World: 2001 is to summa- Third World (1988), used data pri- Appendix B provides more informa- rize socioeconomic statistics for marily from the 1970 and 1980 tion about the sources, limitations, both developed and developing rounds of worldwide censuses and availability of IDB files and nations. This report shows such (those taken from 1965 to 1974 report data in general. There are data for 52 nations when available and 1975 to 1984, respectively), vast differences in both the quantity and reasonably comparable. In as well as demographic projections and quality of statistics reported by 2000, these 52 nations (listed in produced by the United Nations various countries. The United Appendix A, Table 1) contained Population Division from its 1984 Nations has provided international 77 percent of the world’s total pop- assessment of global population. recommendations for the standardi- ulation, and are referred to as Subsequent reports — Population zation of concepts and definitions “study countries” at various points and Health Transitions (1992); of data collected in censuses and in the text.4 Aging in Eastern Europe and the surveys. Nevertheless, there are This report focuses primarily on Former Soviet Union (1993); An still wide discrepancies in data col- people aged 65 years old and over. Aging World II (1993); Older lection and tabulation practices As is true of younger age groups, Workers, Retirement and Pensions because of legitimate differences in people aged 65 and over have very (1995); and the current report — the resources and information different economic resources, health include historical data from the needs among countries. As a statuses, living arrangements, and earlier reports, available data from result, any attempt to compile stan- levels of integration into social life. the 1990 and 2000 rounds of cen- dard data across countries requires An Aging World: 2001 acknowl- suses, information from national consideration of whether and how edges this diversity by disaggregat- sample surveys and administrative the reported data should be ana- ing statistics into narrower age records, historical and projected lyzed to achieve comparability. groups where possible. Such exam- data from the United Nations, and The demographic data in this report ination may reveal important demo- data from component population have been judged by Census Bureau graphic, social, and economic differ- projections prepared by the analysts to be as representative as ences that have direct bearing on International Programs Center possible of the situation in a given social policy now and in the future. (IPC), Population Division, U.S. country. The data are internally con- For example, the fastest growing Census Bureau. Differences among sistent and congruent with other portion of the elderly population in reports in projected data may facts known about the nations. many nations are those aged 80 reflect either a change in the These demographic data also have and over, referred to as the oldest source of the projections or, more been checked for external consisten- old. Rapidly expanding numbers of importantly, revised demographic cy, that is, compared with informa- very old people represent a social insights based on the most recent tion on other countries in the same phenomenon without historical information. region or subregion and with those precedent, and one that is bound to Many of the data included in this elsewhere at approximately the alter previously held stereotypes of report are from the Census Bureau’s same level of socioeconomic devel- older people. The growth of the International Data Base (IDB). The opment. The socioeconomic data, oldest old is salient to public policy tabular statistics provided in by contrast, typically are as reported because individual needs and social Appendix A represent only a small by the countries themselves. portion of the total IDB files. The Although Census Bureau analysts 4 In some parts of the text, data from addi- tional countries have been included. IDB is maintained and updated by have not directly evaluated these 2 An Aging World: 2001 U.S. Census Bureau
  • 10. data, analysts have attempted to proportion of the world’s population increasingly, biologic and genetic. resolve discrepancies in reported fig- that is elderly (Figure 1-2). The The IDB and this report are an effort ures and to eliminate international coming growth, especially of the to contribute to a consistent, sys- inconsistencies; data with obvious oldest old, will be stunning. As tematic, quantitative comparison of incongruities are not included. their numbers grow, there is a older populations in various coun- heightened need to understand the tries. Information is the first step We are all part of an increasingly characteristics of older populations, toward a better understanding of interdependent and aging world their strengths, and their require- the effects of population aging with- (Figure 1-1). Current growth of eld- ments. The effects will be felt not in and across national boundaries. erly populations is steady in some just within individual nations but As individuals, as nations, and as an countries and explosive in others. throughout the global economy. international community, we face As the World War II baby-boom Understanding the dynamics of the challenge of anticipating the cohorts, common to many coun- aging requires accurate descriptions changing needs and desires of an tries, begin to reach their elder of the elderly from interrelated per- aging world in a new millennium. years after 2010, there will be a spectives including demographic, significant jump by 2030 in the social, economic, medical, and U.S. Census Bureau An Aging World: 2001 3
  • 11. 4 An Aging World: 2001 Figure 1-1. Percent Aged 65 and Over: 2000 Less than 3.0 3.0 to 7.9 U.S. Census Bureau 8.0 to 12.9 13.0 or more Source: U.S. Census Bureau, 2000a.
  • 12. U.S. Census Bureau Figure 1-2. Percent Aged 65 and Over: 2030 An Aging World: 2001 5 Less than 3.0 3.0 to 7.9 8.0 to 12.9 13.0 or more Source: U.S. Census Bureau, 2000a.
  • 13. CHAPTER 2. The Demographics of Aging The current level and pace of popula- Projections of older populations developed countries during and tion aging vary widely by geographic may be more accurate than projec- after World War I. A second, less region, and usually within regions as tions of total population, which severe, decline in the rate of well. But virtually all nations are now must incorporate assumptions growth began in the mid-1990s experiencing growth in their num- about the future course of human and will be most noticeable in the bers of elderly residents. Developed fertility. Short-term and early 2000s. This decline corre- nations have relatively high propor- medium-term projections of tomor- sponds to lowered fertility during tions of people aged 65 and over, but row’s elderly are not contingent the Great Depression and World the most rapid increases in elderly upon fertility, because anyone who War II. These drops in growth rate population are in the developing will be aged 65 or over in 2030 has highlight the important influence world. Even in nations where the already been born. When projecting that past fertility trends have on elderly percentage of total population the size and composition of the current and projected changes in remains small, absolute numbers world’s future elderly population, the size of elderly populations. may be rising steeply. Everywhere, human mortality is the key demo- The current aggregate growth rate the growth of elderly populations graphic component. As discussed of the elderly population in devel- poses challenges to social institutions in the next chapter, current and oping countries is more than double that must adapt to changing age future uncertainties about changing that in developed countries, and structures. mortality may produce widely diver- also double that of the total world gent projections of the size of WORLD’S ELDERLY population. The rate in developing tomorrow’s elderly population. POPULATION INCREASING countries began to rise in the early 795,000 EACH MONTH ELDERLY POPULATION 1960s, and has generally continued The world’s elderly population has GROWING FASTEST IN to increase until recent years. After DEVELOPING COUNTRIES a brief downturn — again related to been growing for centuries. What is new is the rapid pace of aging. The Population aging has become a lower wartime fertility — the elderly global population aged 65 and over well-publicized phenomenon in the growth rate in developing countries was estimated to be 420 million industrialized nations of Europe is expected to rise beyond and people as of midyear 2000, an and North America. What is not remain above 3.5 percent annually increase of 9.5 million since widely appreciated is the fact that from 2015 through 2030 before midyear 1999. The net balance of developing countries are aging as declining in subsequent decades. the world’s elderly population grew well, often at a much faster rate EUROPE STILL THE “OLDEST” by more than 795,000 people each than in the developed world. WORLD REGION, AFRICA THE month during the year. Projections Seventy-seven percent of the “YOUNGEST” to the year 2010 suggest that the world’s net gain of elderly individu- Europe has had the highest propor- net monthly gain will then be on als from July 1999 to July 2000 — tion of population aged 65 and over the order of 847,000 people. In 615,000 people monthly — among major world regions for 1990, 26 nations had elderly popu- occurred in developing countries. many decades and should remain lations of at least 2 million, and by Figure 2-2 shows the different pat- the global leader well into the 2000, 31 countries had reached the terns of growth in developed ver- twenty-first century (Table 2-1). 2-million mark. Projections to the sus developing countries. Most Until recently, this region also had year 2030 indicate that more than notable in developed countries is the highest proportions of popula- 60 countries will have 2 million or the steep plunge in growth in the tion in the most advanced age cate- more people aged 65 and over early 1980s. The slowing of the gories. But in 2000, the percentage (Figure 2-1). growth rate was the result of low of population aged 80 and over in birth rates that prevailed in many U.S. Census Bureau An Aging World: 2001 7
  • 14. 8 An Aging World: 2001 Figure 2-1. Countries With 2 Million or More Elderly People: 2000 and 2030 2000 U.S. Census Bureau 2030 Source: U.S. Census Bureau, 2000a.
  • 15. North America was equal to that of Figure 2-2. Europe as a whole, probably as a Average Annual Percent Growth of Elderly result of small European birth Population in Developed and Developing Countries cohorts around the time of World War I. By 2015, however, these per- Percent growth 5 centages are again expected to be highest in Europe; in 2030, nearly Developing countries, 65 years and over 12 percent of all Europeans are pro- 4 jected to be over the age of 74 and 7 percent are projected to be over the age of 79. 3 North America and Oceania also have relatively high aggregate per- Developed countries, 65 years and over centages of elderly, and these are 2 projected to increase substantially between 2000 and 2030. Levels for 2000 in Asia and Latin 1 America/Caribbean are expected to more than double by 2030, while Total world, all ages aggregate proportions of elderly 0 population in Sub-Saharan Africa 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 will grow rather modestly as a Source: United Nations, 1999. result of continued high fertility in many nations. Two important factors bear mention Table 2-1. when considering aggregate elderly Percent Elderly by Age: 2000 to 2030 proportions of regional populations. 65 years 75 years 80 years The first is that regional averages Region Year and over and over and over often hide great diversity. Bangladesh and Thailand may be Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 15.5 6.6 3.3 2015 18.7 8.8 5.2 close geographically, but these 2030 24.3 11.8 7.1 countries have divergent paths of expected population aging. North America . . . . . . . . . . . . . . . . . . . . . . 2000 12.6 6.0 3.3 2015 14.9 6.4 3.9 Likewise, many Caribbean nations 2030 20.3 9.4 5.4 have high proportions of elderly Oceania . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 10.2 4.4 2.3 population (the Caribbean is the 2015 12.4 5.2 3.1 “oldest” of all developing world 2030 16.3 7.5 4.4 regions) in relation to their Central Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 6.0 1.9 0.8 American neighbors. Secondly and 2015 7.8 2.8 1.4 more importantly, percentages by 2030 12.0 4.6 2.2 themselves may not give a sense of Latin America/Caribbean . . . . . . . . . . . . . 2000 5.5 1.9 0.9 population momentum. Although 2015 7.5 2.8 1.5 2030 11.6 4.6 2.4 the change in percent elderly in Sub-Saharan Africa from 2000 to Near East/North Africa . . . . . . . . . . . . . . . 2000 4.3 1.4 0.6 2015 is barely perceptible, the size 2015 5.3 1.9 0.9 2030 8.1 2.8 1.3 of the elderly population is expect- ed to jump by 50 percent, from Sub-Saharan Africa . . . . . . . . . . . . . . . . . . 2000 2.9 0.8 0.3 2015 3.2 1.0 0.4 19.3 million to 28.9 million people. 2030 3.7 1.3 0.6 Source: U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 9
  • 16. ITALY NOW THE WORLD’S “OLDEST” MAJOR COUNTRY Figure 2-3. The percent of population aged 65 The World's 25 Oldest Countries: 2000 and over ranged from 12 to 16 per- (Percent of population 65 years and over) cent in 2000 in most developed countries. For many years Sweden Italy 18.1 had the highest such proportion, Greece 17.3 but recently Italy became the demo- Sweden 17.3 graphically oldest of the world’s Japan 17.0 major1 nations. Over 18 percent of Spain 16.9 all Italians are aged 65 or over, with Belgium 16.8 levels approaching or exceeding Bulgaria 16.5 17 percent in Greece, Sweden, Germany 16.2 Japan, Spain, and Belgium. With France 16.0 the exception of Japan, the world’s United Kingdom 15.7 25 oldest countries are all in Europe Portugal 15.4 (Figure 2-3). The United States, Austria 15.4 with an elderly proportion of less Norway 15.2 than 13 percent in 2000, is rather Switzerland 15.1 young by developed-country stan- Croatia 15.0 dards, and its proportion elderly Latvia 15.0 will increase only slightly during the Finland 14.9 next decade. However, as the large Denmark 14.9 birth cohorts of the baby boom Serbia 14.8 (people born from 1946 through Hungary 14.6 1964) begin to reach age 65 after Estonia 14.5 2010, the percent elderly in the Slovenia 14.1 United States will rise markedly, Luxembourg 14.0 likely reaching 20 percent by the Ukraine 13.9 year 2030. Still, this figure will be Czech Republic 13.9 lower than in most countries of Western Europe. Source: U.S. Census Bureau, 2000a. 1 Some small areas/jurisdictions have very high proportions of elderly population. In 2000, three of the world’s seven highest esti- mated percentages of elders were in the European principality of Monaco (more than 22 percent), Guernsey (17 percent) and the Isle of Man (more than 17 percent). 10 An Aging World: 2001 U.S. Census Bureau
  • 17. SOME ELDERLY POPULATIONS Figure 2-4. TO MORE THAN TRIPLE BY 2030 Percent Increase in Elderly Population: 2000 to 2030 During the period 2000-2030, the projected increase in elderly popu- Developed countries Developing countries lation in the 52 study countries ranges from 14 percent in Bulgaria Singapore 372 to 372 percent in Singapore (Figure Malaysia 277 2-4). Today’s “older” nations will Colombia 258 Costa Rica experience relatively little change 250 Philippines 240 compared with many developing Indonesia 240 nations; in countries as diverse as Mexico 227 Malaysia and Colombia, elderly South Korea 216 Egypt populations are expected to 210 Bangladesh 207 expand to more than three times Peru 206 their size in 2000. Thailand 197 Guatemala 196 Morocco 193 Brazil 192 Chile 183 Sri Lanka 178 Turkey 177 India 174 Tunisia 171 China 170 Liberia 160 Pakistan 153 Jamaica 134 Canada 126 Kenya 117 Australia 108 Israel 102 United States 102 New Zealand 92 Luxembourg 87 Argentina 81 Poland 75 Austria 67 Denmark 64 Czech Republic 63 Germany 63 Norway 62 France 56 United Kingdom 55 Japan 54 Belgium 50 Russia 48 Uruguay 48 Sweden 45 Zimbabwe 45 Italy 43 Greece 43 Malawi 41 Hungary 37 Ukraine 21 Bulgaria 14 Source: U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 11
  • 18. THE LEGACY OF FERTILITY DECLINE Figure 2-5. The most prominent historical fac- Total Fertility Rate: 2000 tor in population aging has been (Births per woman) fertility decline. The generally sus- Western Europe tained decrease in total fertility Austria 1.4 rates (TFRs) in industrialized nations Belgium 1.6 since at least 1900 has resulted in Denmark 1.7 France 1.8 current levels below the population Germany 1.4 replacement rate of 2.1 live births Greece 1.3 per woman in most such nations Italy 1.2 Luxembourg 1.7 (Figure 2-5). Persistent low fertility Norway 1.8 since the late 1970s has led to a Sweden 1.5 decline in the size of successive United Kingdom 1.7 birth cohorts and a corresponding Eastern Europe Bulgaria 1.1 increase in the proportion of older 1.2 Czech Republic relative to younger population. Hungary 1.3 Poland 1.4 Fertility change in the developing Russia 1.3 world has been more recent and Ukraine 1.3 Other Developed more rapid, with most regions hav- 1.8 Australia ing achieved major reductions in Canada 1.6 fertility rates over the last 30 years. Japan 1.4 New Zealand 1.8 Although the aggregate TFR 2.1 United States remains in excess of 4.5 children Asia per woman in Africa and many Bangladesh 2.9 China 1.8 countries of the Near East, overall India 3.1 levels in Asia and Latin America Indonesia 2.6 decreased by about 50 percent Israel 2.6 Malaysia 3.3 (from 6 to 3 children per woman) Pakistan 4.6 during the period 1965 to 1995. Philippines 3.5 Total fertility in many developing Singapore 1.2 countries — notably China, South South Korea 1.7 Sri Lanka 2.0 Korea, Thailand, and at least a 1.9 Thailand dozen Caribbean nations — is now Turkey 2.2 at or below replacement level. Latin America/Caribbean Argentina 2.5 Brazil 2.1 Chile 2.2 Colombia 2.7 Costa Rica 2.5 Guatemala 4.7 Jamaica 2.1 Mexico 2.7 Peru 3.0 Uruguay 2.4 Africa Egypt 3.2 Kenya 3.7 Liberia 6.4 Malawi 5.3 Morocco 3.1 Tunisia 2.0 Zimbabwe 3.3 Source: U.S. Census Bureau, 2000a. 12 An Aging World: 2001 U.S. Census Bureau
  • 19. fractious debates over health care Figure 2-6. costs, social security, and intergen- Speed of Population Aging erational equity that have emerged (Number of years required or expected for percent of population aged 65 in Europe and North America. and over to rise from 7 percent to 14 percent) Developed countries AN AGING INDEX France (1865-1980) 115 An easily understood indicator of Sweden (1890-1975) 85 age structure is the aging index, Australia (1938-2011) 73 defined here as the number of peo- United States (1944-2013) 69 ple aged 65 and over per 100 Canada (1944-2009) 65 youths under age 15. Among the Hungary (1941-1994) 53 52 study countries in 2000, 5 coun- Poland (1966-2013) 47 tries (Germany, Greece, Italy, United Kingdom (1930-1975) 45 Bulgaria, and Japan) had more elder- Spain (1947-1992) 45 ly than youth aged 0 to 14. By Japan (1970-1996) 26 2030, however, all developed coun- Developing countries tries in Figure 2-7 have a projected Azerbaijan (2000-2041) 41 aging index of at least 100, and China (2000-2027) 27 several European countries and Singapore (2001-2028) 27 Japan are in excess of 200. Today’s Chile (2000-2025) 25 aging index typically is much lower Jamaica (2009-2033) 24 in developing countries than in the Sri Lanka (2004-2027) 23 Tunisia (2009-2032) 23 developed world, and the pattern of Thailand (2003-2025) 22 future change is likely to be more Brazil (2011-2032) 21 varied. If future fertility rates Colombia (2017-2037) 20 remain relatively high, the absolute change in the aging index will be Sources: Kinsella and Gist, 1995; U.S. Census Bureau, 2000a. small. Generally, however, the pro- portional rise in the aging index in developing countries is expected to EAST AND SOUTHEAST ASIA Southeast Asia (South Korea, be greater than in developed AGING THE FASTEST Taiwan, and Thailand), fueled by countries. In only one-quarter of a century — dramatic drops in fertility levels. from 1970 to 1996 — the percent The rapidity of change in this The aging index also is useful in of population aged 65 and over in region stands in stark contrast to examining within-country differ- Japan increased from 7 to 14 per- some European countries, where ences in the level of population cent (Figure 2-6). Similarly swift the comparable change occurred aging. As noted in Chapter 5, there increases are expected in China, over a period of up to 115 years. can be significant differences in the beginning around the turn of the Such rapidly aging societies are extent of aging between urban and century, and elsewhere in East and soon likely to face the often- rural areas. There may also be U.S. Census Bureau An Aging World: 2001 13
  • 20. Figure 2-7. DEVELOPING COUNTRIES Asia Aging Index: 2000 and 2030 2000 9 2030 (People aged 65 and over per 100 people aged 0-14) Bangladesh 32 27 China DEVELOPED COUNTRIES 91 14 Western Europe India 38 92 15 Austria Indonesia 186 52 36 96 Israel Belgium 73 172 12 80 Malaysia Denmark 36 143 10 Pakistan 85 25 France 155 10 Philippines 103 30 Germany 38 187 Singapore 114 96 Greece 32 206 South Korea 125 127 Italy 25 261 Sri Lanka 85 74 27 Luxembourg Thailand 117 96 76 21 Norway Turkey 131 71 94 Latin America/Caribbean Sweden 169 39 Argentina 82 74 United Kingdom 152 18 Brazil 69 Eastern Europe 26 Chile 106 90 Bulgaria 14 244 Colombia 50 84 16 Czech Republic Costa Rica 216 64 87 9 Hungary Guatemala 179 18 64 22 Poland Jamaica 160 67 13 70 Mexico Russia 45 141 13 78 Peru Ukraine 43 133 53 Uruguay 73 Other Developed Countries Africa 60 Australia 11 126 Egypt 34 66 6 Canada Kenya 148 19 115 8 Japan Liberia 231 11 51 6 New Zealand Malawi 103 10 13 59 Morocco United States 40 102 20 Tunisia 71 9 Zimbabwe Source: U.S. Census Bureau, 2000a. 23 14 An Aging World: 2001 U.S. Census Bureau
  • 21. Figure 2-8. Aging Index in Brazil by State: 1991 (Population aged 65 and over per 100 population aged 0-14) Roraima Amapa Rio Grande Amazonas Para do Norte Maranhao Ceara Paraiba Piaui Pernambuco Acre Alagoas Tocantins Rondonia Sergipe Bahia Mato Grosso Goias Minas Gerais Espirito Mato Grosso Santo do Sul Sao Paulo Rio de Janeiro Parana Santa Catarina Aging index 5.4 to 8.9 Rio Grande 9.0 to 11.9 do Sul 12.0 to 13.9 14.0 to 21.0 Source: 1991 Census of Brazil. U.S. Census Bureau An Aging World: 2001 15
  • 22. broader regional differences, espe- cially in large nations such as Brazil Figure 2-9. (Figure 2-8). Based on 1991 census Median Age in 12 Countries: 2000, 2015, and 2030 data, the overall aging index in (In years) Brazil was 14. However, this meas- 2000 ure ranged from less than 6 in sev- 2015 Developed countries 2030 eral northern states of the country to 21 in the state of Rio de Janeiro. 37 Canada 41 MEDIAN AGE TO RISE IN 44 ALL COUNTRIES 40 Population aging refers most simply Germany 45 to increasing proportions of older 47 people within an overall population 41 age structure. Another way to think Japan 45 of population aging is to consider a 50 society’s median age, the age that 40 divides a population into numerical- Italy 46 ly equal parts of younger and older 52 people. For example, the 2000 36 median age in the United States was United States 38 36 years, indicating that the num- 39 ber of people under age 36 equals the number who have already cele- brated their 36th birthday. Developing countries 26 The 2000 median ages of the 52 Brazil 31 study countries ranged from 17 in 37 Malawi to 41 in Japan. Developed 30 countries are all above the 32-year China 36 level, while a majority of develop- 41 ing nations have median ages under 25. During the next three decades, 18 Liberia 18 the median age will increase in all 20 52 countries, though at very differ- 17 ent rates. By 2030, Italy is project- Malawi 20 ed to have the highest median age, 23 with half its population aged 52 or 23 over (Figure 2-9), reflecting in large Mexico 28 part the extremely low level of fer- 34 tility now occurring. By way of con- 19 trast, persistently high birth rates Pakistan 24 are likely to preclude a large change 29 in median age in some developing 29 countries (e.g., Liberia and Malawi). Thailand 35 40 Source: U.S. Census Bureau, 2000a. 16 An Aging World: 2001 U.S. Census Bureau
  • 23. THE DYNAMICS OF retirees from other countries, and aged 25 to 44, and younger cohorts POPULATION AGING return migration of former emi- were becoming successively small- The process of population aging is, grants who are above the average er. If fertility rates continue as pro- as noted earlier, primarily deter- population age; all three factors jected through 2030, the aggregate mined by fertility (birth) rates and contribute to population aging. pyramid will start to invert, with secondarily by mortality (death) Some demographers expect interna- more weight on the top than on the rates, so that populations with high tional migration to assume a more bottom. The size of the oldest-old fertility tend to have low propor- prominent role in the aging process, population (especially women) will tions of older people and vice particularly in graying countries increase, and people aged 80 and versa. Demographers use the term where persistently low fertility has over may eventually outnumber any “demographic transition” to refer to led to stable or even declining total younger 5-year group. Although a gradual process2 wherein a socie- population size (see Box 2-1). the effect of fertility decline usually ty moves from a situation of high Eventual shortages of workers may has been the driving force in chang- rates of fertility and mortality to generate demands for immigrant ing population age structures, cur- one of low rates of fertility and labor (Peterson, 1999) and may rent and future changes in mortality mortality. This transition is charac- force nations to choose between will assume much greater weight, terized first by declines in infant relaxed immigration policies and particularly in relatively “aged” and childhood mortality as infec- pronatalist strategies to raise birth countries (Caselli and Vallin, 1990), tious and parasitic diseases are rates (Kojima, 1996). and are discussed further in the reduced. The resulting improve- next chapter. Figure 2-10 illustrates the historical ment in life expectancy at birth and projected aggregate population ELDERLY POPULATIONS occurs while fertility tends to age structure transition in develop- THEMSELVES OFTEN ARE AGING remain high, thereby producing ing and developed countries. At large birth cohorts and an expand- An increasingly important feature of one time, most if not all countries ing proportion of children relative societal aging is the progressive had a youthful age structure similar to adults. Other things being equal, aging of the elderly population to that of developing countries as a this initial decline in mortality gen- itself. Over time, a nation’s elderly whole in 1950, with a large percent- erates a younger population age population may grow older on aver- age of the entire population under structure (Lee, 1994). age as a larger proportion survives the age of 15. Given the relatively to 80 years and beyond. In many Generally, populations begin to age high rates of fertility that prevailed countries, the “oldest old” (people when fertility declines and adult in most developing countries from aged 80 and over) are now the mortality rates improve. Successive 1950 through the early 1970s, the fastest growing portion of the total birth cohorts may eventually overall pyramid shape had not population. In the mid-1990s, the become smaller and smaller, changed greatly by 1990. However, global growth rate of the oldest old although countries may experience the effects of fertility and mortality was somewhat lower than that of a “baby boom echo” as women of decline can be seen in the projected the world’s elderly, a result of low prior large birth cohorts reach child- pyramid for 2030, which loses its fertility that prevailed in many bearing age. International migration strictly triangular shape as the size countries around the time of World usually does not play a major role of younger 5-year cohorts stabilizes War I. In other words, people who in the aging process, but can be and the elderly portion of the total were reaching age 80 in 1996, for important in smaller populations. population increases. example, were part of a relatively Certain Caribbean nations, for The picture in developed countries small birth cohort. The growth rate example, have experienced a com- has been and will be quite different. of the world’s oldest-old population bination of emigration of working- In 1950, there was relatively little from 1996 to 1997 was only age adults, immigration of elderly variation in the size of 5-year 1.3 percent. Just a few years later, groups between the ages of 5 and however, the fertility effects of 2 The concept of demographic transition 24. The beginnings of the post- World War I had dissipated; from admittedly is a broad one, and some would World War II baby boom can be 1999 to 2000, the growth rate of argue that it has many permutations or that there is more than one form of demographic seen in the 0-4 age group. By the world’s 80-and-over population transition; see, for example, the discussion in had jumped to 3.5 percent, 1990, the baby-boom cohorts were Coale and Watkins (1986). U.S. Census Bureau An Aging World: 2001 17
  • 24. Box 2-1. Population Aging in the Context of Overall Population Decline European demographers have sounded warning bells means to affect fertility, including direct financial for at least the last 30 years with regard to the pos- incentives for additional births; indirect pension (i.e., sibility of declining population size in industrialized early retirement) or in-kind benefits such as preferen- nations. Until very recently, however, this idea had tial access for mothers with many children to subsi- not permeated public discourse. Societies were dized housing; or measures to reduce the opportunity aware that they were aging, but the equation of costs of additional childbearing. These policies have aging with population decline was uncommon. In had modest impacts in authoritarian states, but only the last 2 years, the visibility of likely population minimal impacts in liberal democracies such as France decline has increased dramatically, in large part due and Sweden (Teitelbaum, 2000). Industrial societies to United Nations (2000a; 2000b) reports suggesting already provide various rewards, but using them to that populations in most of Europe and Japan will deliberately manipulate fertility is a sensitive issue, decrease in size over the next 50 years, and to pub- potentially involving substantial economic transfers. licity accorded to actual declines in Spain, Italy, The United Nations (2000a) undertook an examination Russia, and other nations. of the likely impact of migration as a counterbalance This trend raises several contentious issues: Is per- to aging, building on earlier work by Lesthaeghe, sistent below-replacement fertility a threat to Page, and Surkyn (1988), the Organization for European and other societies, and if so, can it be Economic Co-Operation and Development (1991), and altered? To what extent, if any, should so-called others. The conclusion was that inflows of migrants “replacement migration” be encouraged as a mecha- will not be able to prevent European population nism to offset population aging? Are there important declines in the future, nor rejuvenate national popula- macroeconomic (e.g., transnational capital flows; tions, unless the migration flows are of very large changes in national savings rates) and national securi- magnitude (i.e., millions annually). On the heels of ty issues that should be considered? this report, the United Nations convened an Expert Group Meeting on Policy Responses to Population There are no hard and fast answers to these ques- Aging and Population Decline in October 2000. The tions. One study (Bongaarts and Bulatao, 2000) exam- consensus of the experts was that replacement migra- ined the experience of the diverse set of countries tion was not a viable “solution” in and of itself, but that have made the transition to low fertility. In very could buffer the likely impact of future aging if used few of these countries has fertility stabilized at rates by governments in conjunction with other policies above two children per woman. Such an occurrence (e.g., increased labor force participation, especially would be dependent on substantial proportions of among women; fertility inducements as noted above). higher-order births, but higher-order births are largely With regard to global financial and security issues, lit- “anachronistic” in industrial-country settings. The ten- tle systematic work has been done on overall tative conclusion was that fertility is unlikely to impacts, though researchers are beginning to explore rebound significantly, but it has been noted that few and model various scenarios (see, e.g., CSIS and demographers anticipated the post-World War II baby Watson Wyatt, 2000; Eberstadt, 2000; MacKellar and boom that will soon have a major impact on popula- Ermolieva, 2001; Mason et al., 2001). tion aging. Governments have employed various 18 An Aging World: 2001 U.S. Census Bureau
  • 25. Figure 2-10. Population by Age and Sex: 1950, 1990, and 2030 Developing countries Developed countries Age 1950 80+ Male Female 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5- 9 0- 4 400 300 200 100 0 100 200 300 400 1990 80+ 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5- 9 0- 4 400 300 200 100 0 100 200 300 400 2030 80+ 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5- 9 0- 4 400 300 200 100 0 100 200 300 400 Millions Sources: United Nations, 1999 and U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 19
  • 26. considerably higher than that of the world’s elderly as a whole (2.3 per- Figure 2-11. cent). In the future, we expect to Percent Distribution of World Population see sustained high growth of the Aged 80 and Over: 2000 oldest old. In the first decade of the twenty-first century, the projected United States 13.1 average annual growth rate of the 80-and-over population is 3.9 per- China 16.3 cent (versus 2.0 percent for the India 8.7 65-and-over population), and is Japan 6.6 expected to remain above 3 percent Other Asia 8.9 during the period 2010-2020. Russia 4.2 The oldest old constituted 17 per- Germany 4.1 cent of the world’s elderly in 2000: United Kingdom 3.4 22 percent in developed countries Italy 3.3 and 13 percent in developing coun- France 3.1 tries. More than half (53 percent) of Spain 2.1 the world’s oldest old in 2000 lived Other Europe 10.5 in just six countries: China, the United States, India, Japan, Latin America/Caribbean 6.9 Germany, and Russia (Figure 2-11). Africa/Near East 5.8 About an additional one-fifth (22 All remaining countries 3.1 percent) lived elsewhere in Europe, while 7 percent lived in Latin Note: Data represent the share of the world's total oldest old in each country or region. Individual America/Caribbean and about 6 countries with more than 2.0 percent of the total are shown separately. Source: U.S. Census Bureau, 2000a. percent lived in Africa/Near East regions, and another 9 percent in Asian countries other than China, Sweden, and 4 percent or more of countries (e.g., Barbados, Cuba, India, and Japan. the total in several other European Puerto Rico, and Uruguay) have countries (Denmark, Italy, Norway, higher levels than many Eastern Among the 52 study countries, the and the United Kingdom). In gener- European nations. percentage of oldest old in the total al, Western European nations are population in 2000 was less than Countries vary considerably in the above 3 percent, while other half a percent in several developing projected age components of elder- developed countries are between 2 countries (e.g., Egypt, Guatemala, ly populations. In the United States, and 3 percent. In most developing Indonesia, Kenya, and Malawi). In the oldest old were 26 percent of all nations, less than 1 percent of the contrast, the oldest old constituted elderly in 2000, and are expected population is aged 80 and over, 5 percent of the total population of to continue to be 26 percent in although some developing 20 An Aging World: 2001 U.S. Census Bureau
  • 27. boom may produce a great-grand- Figure 2-12. parent boom in many countries. Oldest Old as a Percent of All Elderly: 2000 and 2030 The numerical growth and increas- (People aged 80 and over as a percent of people aged 65 and over) ing heterogeneity of the oldest old 2000 compel social planners to seek fur- 2030 ther health and socioeconomic 21.7 information about this group, Argentina 27.3 because the oldest old consume 13.2 disproportionate amounts of health Bulgaria 27.8 and long-term care services 21.7 Japan (Suzman, Willis, and Manton, 1992). 39.3 Past population projections often 13.3 Pakistan have underestimated the improve- 14.4 ment in mortality rates among the 15.9 Russia oldest old, and as the next chapter 20.0 points out, actual numbers of 26.5 United States tomorrow’s oldest old could be 26.4 much higher than presently antici- Source: U.S. Census Bureau, 2000a. pated. Because of the sustained increases in longevity in many nations, greater age detail is needed 2030 (Figure 2-12). Some European Stability in the proportion of oldest for the oldest old. In the past, com- nations will experience a sustained old in the elderly population should parable population projections for rise in this ratio, while others will not deflect attention from burgeon- the world’s countries often grouped see an increase during the next two ing absolute numbers. In the everyone aged 80 and over into a decades and then a subsequent United States, the oldest old single, open-ended component. decline. The most striking global increased from 374 thousand in Today, for the first time, agencies increase is likely to occur in Japan; 1900 to more than 9 million today. (e.g., the United Nations Population by 2030, nearly 40 percent of all The small percentage decline for Division; the U.S. Census Bureau’s elderly Japanese are expected to be the United States in Figure 2-12 International Programs Center) are at least 80 years old. Most masks a projected absolute increase producing sets of international pop- developing countries should experi- of over 9 million oldest-old people. ulation projections that expand the ence modest long-term increases in Four-generation families are becom- range of older age groups up to an this ratio. ing increasingly common (Soldo, open-ended category of age 100 1996), and the aging of the baby and over. U.S. Census Bureau An Aging World: 2001 21
  • 28. Box 2-2. The Growth of Centenarians As average length of life increases, the concept of There are several problems with obtaining accurate “oldest old” will change. While people of extreme old age data on very old people (Kestenbaum, 1992; Elo age constitute a tiny portion of total population in et al., 1996), and estimates of centenarians from cen- most of the world, their numbers are of growing suses and other data sources should be scrutinized importance, especially in more-developed nations. carefully. For example, the 1990 United States census Thanks to improvements in nutrition, health, and recorded some 37,000 centenarians, although due to health care, we now have for the first time in history age misreporting, the actual figure is thought to be the opportunity to consider significant numeric closer to 28,000, (Krach and Velkoff, 1999). Still, this growth of the population aged 100 and over. represents a doubling of the population aged 100 and According to researchers in Europe, the number of over from 1980 to 1990, similar to estimates for centenarians has doubled each decade since 1950 in European nations. The potentially spectacular industrialized countries. Using reliable statistics from increase in numbers of centenarians is illustrated by ten Western European countries and Japan, Vaupel and data and projections for France. Dinh (1995) has esti- Jeune (1995) estimated that some 8,800 centenarians mated that there were about 200 centenarians in lived in these countries as of 1990, and that the num- France in 1950, and that by the year 2000 the num- ber of centenarians grew at an average annual rate of ber would be 8,500. His 50-year projections suggest approximately 7 percent between the early 1950s and 41,000 people aged 100 and over by 2025, increas- the late 1980s. They also estimate that, over the ing to 150,000 in 2050. If these projections are real- course of human history, the odds of living from birth ized, the number of centenarians in France will have to age 100 may have risen from 1 in 20 million to 1 multiplied by a factor of 750 in one century. in 50 for females in low-mortality nations such as Japan and Sweden. 22 An Aging World: 2001 U.S. Census Bureau
  • 29. CHAPTER 3. Life Expectancy and Changing Mortality The spectacular increases in human European nations, the normal the world have been fairly uniform. life expectancy that began in the lifetime in many African countries Practically all nations have shown mid-1800s and continued during spans fewer than 45 years. On continued improvement, with some the following century are often average, an individual born in a exceptions in Latin America and ascribed primarily to improvements more-developed country can now more recently in Africa, the latter in medicine. However, the major expect to outlive his/her counter- due to the impact of the HIV/AIDS impact of improvements both in part in the less-developed world by epidemic. The most dramatic gains medicine and sanitation did not 13 years. in the developing world have been occur until the late nineteenth cen- in East Asia, where life expectancy tury. Earlier and more important TWENTIETH CENTURY LIFE at birth increased from less than EXPECTANCY HAS DOUBLED IN factors in lowering mortality were 45 years in 1950 to more than SOME DEVELOPED COUNTRIES innovations in industrial and agri- 72 years today. cultural production and distribution, Table 3-1 shows the enormous which improved nutrition for large strides that countries have made in TREND IN RISING LIFE extending life expectancy since EXPECTANCY MAY BE numbers of people (Thomlinson, CHALLENGED 1976). A growing research consen- 1900. In developed countries, the sus attributes the gain in human average national gain in life While global gains in life expectancy longevity since the early 1800s to a expectancy at birth was 66 percent at birth have been the norm, complex interplay of advancements for males and 71 percent for unforeseen changes and epidemics in medicine and sanitation coupled females during the period 1900-90. may reverse the usual historical pat- with new modes of familial, social, In Italy, life expectancy at birth for tern. Beginning in the 1950s, the economic, and political organization women increased from 43 years in typical sustained increase in life (Moore, 1993). 1900 to over 82 years in 2000. In expectancy at birth in developed some cases, life expectancy has countries began to take different LIFE EXPECTANCY AT BIRTH more than doubled during the cen- paths. While female life expectancy EXCEEDS 78 YEARS IN 28 tury (e.g., Spain). continued to rise virtually every- COUNTRIES where, male gains slowed signifi- Increases in life expectancy were Life expectancy at birth in Japan cantly and in some cases leveled more rapid in the first half than in and Singapore has reached off. From the early 1950s to the the second half of the century. 80 years, the highest level of all the early 1970s, for example, male life Expansion of public health services world’s major countries, and has expectancy changed little in and facilities and disease eradica- reached 79 years in several other Australia, the Netherlands, Norway, tion programs greatly reduced developed nations (e.g., Australia, and the United States. After this death rates, particularly among Canada, Italy, Iceland, Sweden, and period of stagnation, male life infants and children. From 1900 to Switzerland). Levels for the United expectancy again began to rise. 1950, people in many Western States and most other developed nations were able to add 20 years In Eastern Europe and the former countries fall in the 76-78 year or more to their life expectancies. Soviet Union, the pace of improve- range (Figure 3-1). Throughout the ment in the 1950s and early 1960s developing world, there are Reliable estimates of life expectancy was extraordinary. Advances in extreme variations in life expectan- for many developing countries prior living conditions and public health cy at birth (Figure 3-2). While the to 1950 are unavailable. Since policies combined to produce large levels in some developing nations World War II, changes in life declines in mortality by reducing match or exceed those in many expectancy in developing regions of some major causes of death U.S. Census Bureau An Aging World: 2001 23
  • 30. (e.g., tuberculosis) to minimal lev- els (Vishnevsky, Shkolnikov, and Figure 3-1. Vassin, 1991). Resultant gains in Life Expectancy at Birth: 2000 Developed countries life expectancy in excess of (In years) Developing countries 5 years per decade were common. By the mid-1960s, however, the United States 77.1 Western Europe rate of increase had decelerated Sweden 79.6 sharply. In the 1970s and 1980s, Italy 79.0 changes in female life expectancy France 78.8 Norway 78.7 at birth were erratic, while male Greece 78.4 life expectancy fell throughout the Belgium 77.8 region (Bobadilla and Costello, Austria 77.7 United Kingdom 77.7 1997). Following the demise of Germany 77.4 the former Soviet Union, the Luxembourg 77.1 decline has continued into the Denmark 76.5 Eastern Europe 1990s in some countries. The Czech Republic 74.5 decline has been particularly Poland 73.2 severe for Russian men; between Hungary 71.4 Bulgaria 70.9 1987 and 1994, male life Russia 67.2 expectancy at birth plummeted Ukraine 66.0 7.3 years to a level of 57.6, before Other Developed Japan 80.7 beginning to rise again in recent Australia 79.8 years (Figure 3-3). The large Canada 79.4 increases in adult male mortality New Zealand 77.8 Asia usually are attributed to a combi- Singapore 80.1 nation of factors including Israel 78.6 increased homicide and accident South Korea 74.4 Sri Lanka 71.8 rates, excessive alcohol consump- China 71.4 tion, poor diet, and environmen- Turkey 71.0 tal/workplace degradation Malaysia 70.8 Thailand 68.6 (Virganskaya and Dmitriev, 1992; Indonesia 68.0 Murray and Bobadilla, 1997), Philippines 67.5 although most researchers take India 62.5 pains to point out that clear causal Pakistan 61.1 Bangladesh 60.2 mechanisms remain poorly Latin America/Caribbean understood. Costa Rica 75.8 Chile 75.7 Uruguay 75.2 Jamaica 75.2 Argentina 75.1 Mexico 71.5 Colombia 70.3 Peru 70.0 Guatemala 66.2 Brazil 62.9 Africa Tunisia 73.7 Morocco 69.1 Egypt 63.3 Liberia 51.0 Kenya 48.0 Zimbabwe 37.8 Malawi 37.6 Source: U.S. Census Bureau, 2000a. 24 An Aging World: 2001 U.S. Census Bureau
  • 31. U.S. Census Bureau Figure 3-2. Life Expectancy at Birth: 2000 An Aging World: 2001 25 Under 50 years 50 to 64 years 65 to 74 years 75 years and over Source: U.S. Census Bureau, 2000a.
  • 32. Table 3-1. Life Expectancy at Birth in 34 Countries: 1900 to 2000 (In years) Circa 1900 Circa 1950 2000 Region/country Male Female Male Female Male Female DEVELOPED COUNTRIES Western Europe Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37.8 39.9 62.0 67.0 74.5 81.0 Belgium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45.4 48.9 62.1 67.4 74.5 81.3 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51.6 54.8 68.9 71.5 74.0 79.3 France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45.3 48.7 63.7 69.4 74.9 82.9 Germany1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43.8 46.6 64.6 68.5 74.3 80.8 Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.3 55.8 70.3 73.8 75.7 81.8 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.8 55.3 69.9 72.6 77.0 82.4 United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . 46.4 50.1 66.2 71.1 75.0 80.5 Southern and Eastern Europe Czech Republic1 . . . . . . . . . . . . . . . . . . . . . . . . . . 38.9 41.7 60.9 65.5 71.0 78.2 Greece. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38.1 39.7 63.4 66.7 75.9 81.2 Hungary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36.6 38.2 59.3 63.4 67.0 76.1 Italy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.9 43.2 63.7 67.2 75.9 82.4 Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33.9 35.7 59.8 64.3 75.3 82.5 Other Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53.2 56.8 66.7 71.8 76.9 82.7 Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.8 44.3 59.6 63.1 77.5 84.1 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48.3 51.1 66.0 71.7 74.2 79.9 Circa 1950 2000 Male Female Male Female DEVELOPING COUNTRIES Africa Egypt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.2 43.6 61.3 65.5 Ghana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40.4 43.6 56.1 58.8 Mali. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.1 34.0 45.5 47.9 South Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44.0 46.0 50.4 51.8 Uganda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38.5 41.6 42.2 43.7 Congo (Brazzaville) . . . . . . . . . . . . . . . . . . . . . . . . 37.5 40.6 44.5 50.5 Asia China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39.3 42.3 69.6 73.3 India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39.4 38.0 61.9 63.1 Kazakhstan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51.6 61.9 57.7 68.9 South Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46.0 49.0 70.8 78.5 Syria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44.8 47.2 67.4 69.6 Thailand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45.0 49.1 65.3 72.0 Latin America Argentina. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60.4 65.1 71.7 78.6 Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49.3 52.8 58.5 67.6 Costa Rica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56.0 58.6 73.3 78.5 Chile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57.8 61.3 72.4 79.2 Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49.2 52.4 68.5 74.7 Venezuela . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53.8 56.6 70.1 76.3 1 Figures for Germany and Czech Republic prior to 1999 refer to the former West Germany and Czechoslovakia, respectively. Note: Reliable estimates for 1900 for most developing countries are not available. Source: UNDIESA 1988; Siampos 1990; and U.S. Census Bureau, 2000a. 26 An Aging World: 2001 U.S. Census Bureau
  • 33. Elsewhere, the HIV/AIDS epidemic Figure 3-3. has had a devastating impact on life Life Expectancy at Birth in Four expectancy, particularly in parts of Countries: 1950 to 1998 Africa. The effect of the epidemic on life expectancy at birth may be Male considerable, given that AIDS Life expectancy in years deaths often are concentrated in the 90 childhood and middle adult (30 to 80 United States 45) ages. Projections to the year 2010 suggest that AIDS may reduce 70 life expectancy at birth by more than 30 years from otherwise- 60 South Korea expected levels in countries such as Russia Botswana, Namibia, South Africa, 50 Zimbabwe and Zimbabwe. And while the com- 40 mon perception of AIDS mortality usually associates AIDS deaths with 30 children and younger adults, the epidemic may have a direct and 20 growing effect on older popula- 10 tions. In the United States in 1992, nearly three times as many people 0 aged 60 and over died of AIDS as 1950 1960 1970 1980 1990 1998 did people under age 20. Between 1987 and 1992, the annual number Female of U.S. children who died of AIDS Life expectancy in years 90 remained relatively stable, whereas United States the number of deaths to people 80 aged 60 and over nearly doubled (Hobbs and Damon, 1996). 70 Russia South Korea 60 50 Zimbabwe 40 30 20 10 0 1950 1960 1970 1980 1990 1998 Sources: United Nations, 1999; U.S. Census Bureau, 2000a; and country sources. U.S. Census Bureau An Aging World: 2001 27
  • 34. FEMALE ADVANTAGE IN LIFE EXPECTANCY NEARLY Figure 3-4. UNIVERSAL Female Advantage in Life Expectancy at Birth: 2000 The widening of the sex differential (Difference in years between females and males) Developed countries in life expectancy has been a central Developing countries feature of mortality trends in devel- oped countries in the twentieth cen- Belarus 12.7 tury. In 1900, in Europe and North France 8.0 America, women typically outlived Japan 6.5 Hungary 9.1 men by 2 or 3 years. Today, the Russia 10.7 average gap between the sexes is United States 5.7 roughly 7 years, but exceeds Spain 7.2 12 years in parts of the former Bangladesh -0.5 Soviet Union as a result of the Syria 2.3 unusually high levels of male mor- Mexico 6.2 tality discussed above (Figure 3-4). Egypt 4.2 This differential reflects the fact that China 3.7 in most nations females have lower Chile 6.8 mortality than males in every age Source: U.S. Census Bureau, 2000a. group and for most causes of death. Female life expectancy now exceeds 80 years in over 30 countries and is approaching this level in many other Figure 3-5. nations. The gender differential Mortality Rates at Older Ages: 2000 (Per 1,000 population in age/sex group) usually is smaller in developing Male countries, commonly in the 3-6 year Female Canada range, and even is reversed in some 23 South Asian and Middle East soci- 65-69 12 eties where cultural factors (such as 36 70-74 low female social status and prefer- 19 57 ence for male rather than female off- 75-79 32 spring) are thought to contribute to 80+ 117 higher male than female life 87 expectancy at birth. Uruguay 30 65-69 MALE MORTALITY 16 SUBSTANTIALLY HIGHER THAN 46 70-74 25 FEMALE MORTALITY AT 68 OLDER AGES 75-79 45 138 The data in Figure 3-5 illustrate the 80+ 108 usual gender pattern of mortality at Germany older ages, wherein male rates are 26 seen to be consistently higher than 65-69 12 female rates. In Canada and 40 70-74 Germany, for instance, male mortali- 21 65 ty rates for ages 65 to 74 are 75-79 38 roughly twice as great as correspon- 147 80+ ding female rates. Among coun- 116 tries, though, age-specific mortality Source: Estimated by the U.S. Census Bureau based on individual country sources. rates can differ widely even where overall mortality appears similar. 28 An Aging World: 2001 U.S. Census Bureau
  • 35. nations, most demographers expect Figure 3-6. to see a widening of the Life Expectancy at Age 65 in Japan and the female/male difference in upcoming United States: 1970, 1980, and 1998 decades, along the lines of the his- (Years of life remaining for those who reach age 65) torical trend in industrialized nations. Evidence suggests that Japan, Male many developing countries are 1970 12.5 experiencing increases in alcohol 1980 14.6 and tobacco consumption and 1998 17.1 vehicular as well as industrial acci- Japan, Female dents, all of which tend, at least ini- 1970 15.3 tially, to adversely affect men more 1980 17.7 than women. Another factor that 1998 22.0 may promote a widening gender United States, Male gap is education, which is positively 1970 13.1 related to survival. As women 1980 14.1 “catch up” to men in terms of edu- 1998 16.0 cational attainment, female survival United States, Female and health status may improve (Liu, 1970 17.0 Hermalin, and Chuang, 1998). 1980 18.3 1998 19.2 OLD-AGE MORTALITY RATES DECLINING OVER TIME Source: U.S. Census Bureau, 2000a. In countries where infant mortality rates are still relatively high but declining, most of the improvement For example, total life expectancy cited as a source of higher male in life expectancy at birth results at birth in 1995 was about the mortality rates (Statistics Canada, from helping infants survive the same in Cuba (75.4 years) and 1997), suggesting that the gap in high-risk initial years of life. But Portugal (74.7 years). However, life expectancy might decrease if when a nation’s infant and child- World Health Organization data for women increased their use of hood mortality reach low levels, 1995 show that the female mortali- tobacco and alcohol and their par- longevity gains in older segments ty rate for ages 55 to 64 was ticipation in the labor force. of the population begin to assume 30 percent lower in Portugal than in However, data from industrialized greater weight (Caselli and Vallin, Cuba, and the female mortality rate countries still show no clear pattern 1990; Gjonca, Brockmann, and at ages 65 to 74 was about 20 per- of change in the gender gap; the Maier, 1999). Most countries are cent lower. For older men, on the gap is widening in most of Eastern experiencing a rise in life expectan- other hand, rates were 15 percent Europe and the former Soviet Union cy at age 65, as exemplified by higher in each age group in and tends to be narrowing in other Japanese and U.S. data in Figure 3-6. Portugal than in Cuba. developed countries. In the United The average Japanese woman States, for instance, life expectancy reaching age 65 in 1998 could WILL THE GENDER GAP IN LIFE at birth increased 3.0 years for men EXPECTANCY NARROW? expect to live an additional and 1.6 years for women between 22 years, and the average man Precise explanations of the gender 1980 and 1996. But in some more than 17 years. Overall (i.e., difference in life expectancy still nations with very high overall life both sexes combined) Japanese life elude scientists because of the expectancy (e.g., France, Germany, expectancy at age 65 increased apparent complex interplay of bio- Japan), gains in female longevity 40 percent from 1970 to 1998, logical, social, and behavioral condi- continue to outpace those of males. compared with an overall increase tions. Greater exposure to risk fac- Given the small average gender gap in life expectancy at birth of 9 per- tors such as tobacco and alcohol in life expectancy in developing cent. Comparative figures for the use and occupational hazards is countries relative to developed United States are 17 and 8 percent, U.S. Census Bureau An Aging World: 2001 29
  • 36. respectively. Although greater rela- tive improvement in life expectancy Figure 3-7. at older ages may not yet be wide- Percent Change in Death Rates for Two Older spread in developing regions of the Age Groups: Circa 1985 to Circa 1995 world, the proportional increase in (Based on 3-year average of mortality rates centered on 1985 and 1995) life expectancy at older ages is 70-74 years approaching or has surpassed the 80-84 years Female relative increase in life expectancy -6 at birth in some developing coun- France -20 tries, notably in Latin America and -7 Netherlands -8 the Caribbean (Kinsella, 1994). -27 Mauritius 9 The rise in life expectancy at age 65 -6 that is characteristic of most soci- United States -8 eties means that the chances of -24 Chile -10 dying for particular older age -20 Israel groups are declining. Figure 3-7 -16 -27 shows across-the-board declines in Japan -28 mortality in two older age groups -23 Austria (with the exception of women aged -22 80 to 84 in Mauritius) during a fair- Male ly recent 10-year period. In gener- -3 France -14 al, mortality improvements for -13 people aged 70 to 74 have been Netherlands -3 greater than for people aged 80 to -16 Mauritius -12 84, reflecting the growing robust- -14 ness of younger elderly cohorts. United States -10 -20 Chile -3 MORTALITY RATE INCREASE APPEARS TO LESSEN AT VERY -16 Israel -13 OLD AGES -16 Japan -17 As long ago as the early 1800s, -19 research demonstrated that the Austria -20 human death rate increases with Source: United Nations, various issues of the annual Demographic Yearbook. age in an exponential manner, at least to the upper ranges of the age distribution. Recently, researchers deceleration occurs is rising (Lynch may slow down at very old ages, have documented that, at very old and Brown, 2001). and/or that certain genes that are ages, the rate of increase in the detrimental to survival may be sup- mortality rate tends to slow down. Findings such as these have gener- pressed (see Horiuchi and Wilmoth, In a study of 28 countries with rea- ated at least two potential explana- 1998, for a discussion and examina- sonably-reliable data for the period tions. The “heterogeneity” hypothe- tion of these hypotheses). The 1950-90, Kannisto (1994) noted not sis, an extension of the notion of observed deceleration in mortality, only a decline in mortality rates at “survival of the fittest,” posits that combined with the fact that human ages 80 and over, but a tendency the deceleration in old-age mortality mortality at older ages has declined toward greater decline in more- is a result of frailer elderly dying at substantially, has led to the ques- recent time periods. Other work younger ages, thus creating a very tioning of many of the theoretical has confirmed this tendency (e.g., old population with exceptionally tenets of aging (Vaupel et al., 1998). Wilmoth et al., 2000), and a recent healthy attributes resulting from Important insights are being gar- study in the United States suggests genetic endowment and/or lifestyle. nered from “biodemographic” that the age at which mortality A second, “individual-risk” hypothe- research which attempts to sis, suggests that the rate of aging 30 An Aging World: 2001 U.S. Census Bureau
  • 37. series projection suggests that Figure 3-8. there will be 14.3 million people Projections of the United States Population aged 85 and over in the year 2040, Aged 85 and Over while the low-mortality (i.e., high (In millions) life expectancy) series implies 16.8 million. As those who will be 1960-2000 Actual 2040 Census Bureau middle mortality 85 years old and over in the year 53.9 2040 Census Bureau low mortality 2040 are already at least 40 years 2040 Guralnik et al. old, the differences in these projec- 2040 Manton et al. tions result almost exclusively from assumptions about adult mortality rates and are not affected by future birth or infant mortality rates. In the middle-mortality series, the 23.5 Census Bureau assumes that life expectancy at birth will reach 16.8 84.0 years in 2050, while in the 14.3 low-mortality series life expectancy is assumed to reach 86.1 years in 3.1 4.2 2050 (Hollmann, Mulder, and 1.5 2.2 0.9 Kallen, 2000). 1960 1970 1980 1990 2000 2040 Alternative projections (Figure 3-8), Note: U.S. Census Bureau projections for 2040 do not reflect the results of the 2000 census. using assumptions of lower death Sources: Guralnik, Yanagishita, and Schneider, 1988; Manton, Stallard, and Liu, 1993; and U.S. Census Bureau, 1983, 1992, and 2000b. rates and higher life expectancies, have produced even larger esti- mates of the future population of cross-fertilize the biologic and England and Wales is about the United States aged 85 and over. demographic perspectives of aging 50 percent higher today than in Simply assuming that death rates and senescence. While a clearer pic- 1950. Consequently, the number of will continue falling at about the ture of the causes of mortality decel- female octogenarians is about 50 recent 2 percent rate results in a eration at very old ages awaits fur- percent higher than it would have projection of 23.5 million aged 85 ther investigation (and will benefit been had oldest-old mortality and over in 2040 (Guralnik, from the study of evolutionary biol- remained at 1950 levels; in absolute Yanagishita, and Schneider, 1988). ogy and aging in nonhuman terms, this means that there are Even more optimistic forecasts of species; see Olshansky, 1998; more than one-half million oldest-old future reductions in death rates Wachter and Finch, 1997; Le Bourg, British women alive today who oth- have been made from mathematical 2001), its recognition at a time erwise would have been dead in the simulations of potential reductions when numbers of the very old are absence of mortality improvement. in known risk factors for chronic growing rapidly has important poli- disease, morbidity, and mortality. An example from the United States cy implications. Manton, Stallard, and Liu (1993) illustrates the range of future uncer- used such a method to generate an PACE OF MORTALITY CHANGE tainty about the size of tomorrow’s extreme “upper bound” projection DIFFICULT TO PROJECT oldest-old population. The U.S. for the United States of 54 million Census Bureau estimated the num- The pace at which death rates at people aged 85 and over in 2040. ber of people aged 85 and over in advanced ages decline will play a While such projections are not nec- the United States to be about major role in determining future essarily the most likely, they do 3.6 million in 1995 and has made numbers of elderly and especially of illustrate the potential impact of several projections of the future very old population. Vaupel (1997) changes in adult mortality on the size of this age group (Day, 1996; has noted that the remaining life future size of the extremely old U.S. Census Bureau, 2000b). The expectancy of 80-year-old women in population. Census Bureau’s middle-mortality U.S. Census Bureau An Aging World: 2001 31
  • 38. As noted earlier, researchers are increasingly concerned with overall Box 3-1. patterns of mortality decline in Data on Causes of Death addition to studying the pace of Statistics on causes of death are prone to many biases and errors in all decline. One recent study of countries. Underreporting of deaths, lack of precise causal informa- mortality in the G71 countries dur- tion, inaccurate diagnoses, and cultural differences complicate both ing the second half of the twentieth national and international studies of mortality. Also, by attributing century reached a provocative con- death to one specific cause, comorbidities and underlying conditions clusion: mortality at each age has such as anemia and nutritional deficiencies are often masked. declined exponentially at a fairly Nevertheless, World Health Organization efforts to revise and extend constant rate in each country coverage of the International Classification of Diseases produce ongo- (Tuljapurkar, Li, and Boe, 2000). ing improvements in the quality and comparability of data such as The possibility of a “universal pat- those referred to in this report. Although mortality data are imperfect, tern” of mortality decline raises they can be used to illuminate general patterns and orders of magni- important questions about the rela- tude, and to focus the attention of research, planning, and practice. tionship between social expendi- While decisionmaking should be skeptical of small differences between tures on health and their effect on groups, major differences are likely to indicate underlying disparities death rates, and about the likeli- and trends. hood that the mortality decline will be sustained in the future. increase with age but in recent years cardiovascular disease is not just a THE IMPORTANCE OF these rates have declined at older developed-country phenomenon; a CARDIOVASCULAR DISEASE ages in many developed countries. recent U.S. Institute of Medicine study A major disadvantage of summary Nevertheless, cardiovascular disease (Howson et al., 1998) identifies car- mortality indexes such as life remains the primary killer among eld- diovascular disease as the primary expectancy is that they mask erly populations, more so than for noncommunicable health problem changes in mortality by age and/or adults in general. For example, more throughout the developing world. cause of death. While one can than two-thirds of all deaths to elderly In the United States, the heart- examine life expectancy at different people in Bulgaria and nearly half of disease component of cardiovascular ages, it may be more useful to con- all deaths to elderly people in mortality is the leading cause of sider cause-specific changes in mor- Argentina are attributed to cardiovas- death within the elderly population. tality, particularly if the intention is cular disease, and in most countries Among people aged 65 to 74, heart to devise medical or nutritional the proportion increases with age disease and cancers were equally interventions to affect overall (Figure 3-9). One comprehensive likely to be reported as the major longevity and the quality of years analysis of developed nations cause of death in 1997, each lived at older ages. Worldwide, (Murray and Lopez, 1996) attributes accounting for about one-third of all data on cause-specific mortality are nearly 60 percent of all deaths to deaths in that age group. But as age far from ideal for policy making in a women aged 60 and over in the early advances, heart disease claims an majority of countries (See Box 3-1). 1990s to cardiovascular disease; the increasing share, about 49 percent Many developed countries, howev- corresponding figure for older men of deaths to people aged 75 and er, have had reasonably good cause- was 50 percent. Cancer deaths at over (versus 18 percent for cancers). specific mortality data since at least older ages usually rank a distant sec- This pattern also occurs in other (but 1950. ond, but may be more worrisome in not all) developed countries. the public eye; in the 1995 Canadian Death rates due to cardiovascular dis- Women’s Health Test, most women LUNG CANCER RAMPANT eases — a broad category that believed breast cancer to be the num- SINCE THE 1950s includes heart, cerebrovascular ber one killer of women (all ages). (stroke) and hypertensive diseases — Although deaths from cardiovascular Only 16 percent of those surveyed disease are expected to remain most correctly stated that heart disease is 1 The G7 countries include Canada, France, prominent in the future, a major Germany, Italy, Japan, the United Kingdom, and the number one killer (Wiesenberg, concern of health practitioners in the United States. 1996). The prominence of 32 An Aging World: 2001 U.S. Census Bureau
  • 39. responsible for 30 percent of all Figure 3-9. cancer deaths to males in devel- Percent of All Deaths From Cardiovascular Diseases and oped countries, and 12 percent of Malignant Neoplasms at Older Ages in Four Countries all cancer deaths to females. Cardiovascular diseases Proportions for the 60-and-over Malignant neoplasms population are virtually identical. Bulgaria, 1998 Male death rates from lung cancer 66 Male, 65 to 74 years appear to have peaked and are now 16 74 falling in some countries and stabi- 75 years and over 8 lizing in many others, perhaps por- 71 Female, 65 to 74 years tending future declines. 15 75 years and over 79 Conversely, female death rates from 6 lung cancer are rising rapidly, in Argentina, 1996 proportion to the large increases in Male, 65 to 74 years 43 cigarette consumption which began 24 47 several decades ago (Bonita, 1996). 75 years and over 17 Cigarette smoking has been labeled 41 Female, 65 to 74 years 27 as the single most important pre- 52 ventable cause of premature death 75 years and over 13 in women aged 35 to 69 in both France, 1996 the United States and the United 28 Kingdom (Amos, 1996). Still, breast Male, 65 to 74 years 42 37 cancer remains the principal neo- 75 years and over 24 plasm site among females, and 27 Female, 65 to 74 years 40 mortality rates from this source 42 have increased or remained con- 75 years and over 15 stant in most countries since the United States, 1997 post-war period. The rise has been 40 most pronounced in Southern and Male, 65 to 74 years 33 Eastern Europe, which is consistent 75 years and over 46 21 with the hypothesis that a diet high 36 Female, 65 to 74 years in saturated fat is a leading risk fac- 35 52 tor for breast cancer. 75 years and over 15 SUICIDE RATES MUCH HIGHER Sources: World Health Organization, various issues of World Health Statistics Annual, and www-nt.who.int/whosis/statistics/whsa/whsa_table4.cfm?path=statistics,whsa,whsa_table4&language=english AMONG ELDERLY MEN THAN WOMEN Suicide rates in 21 countries with the industrialized world is the rise in cancer has been declining steadily relatively reliable data (Table 3-2) lung cancer among older women as since the 1930s, a decline clearly are consistently higher among men a result of increased tobacco use attributed to nutritional change, i.e., than women in all age groups, since World War II. With regard to a reduction in salt content of food, including the elderly. This is a uni- cancers in general, overall age- especially preserved food (Lopez, versal trend even in societies as dis- standardized death rates for cancer 1990). parate as Singapore, the United rose 30-50 percent among men dur- On the other hand, prevalence of States, Israel, and Bulgaria. Suicide ing the period 1950-85, and fell by lung cancer has increased since rates generally increase with age about 10 percent among women. World War II, initially among men among men, and are highest at However, such broad trends often but now increasingly among ages 75 and over. Suicide rates for are the net result of quite different women. Estimates for the early women also tend to rise with age, changes in mortality for the leading 1990s (Murray and Lopez, 1996) although the peak rate for women sites of disease. In the United suggest that lung cancer is occurs before age 75 in about one States and Western Europe, stomach U.S. Census Bureau An Aging World: 2001 33
  • 40. Table 3-2. Suicide Rates for Selected Age Groups in 23 Countries: Circa 1997 (Rate per 100,000 population in each age group) Male Female Country 15 to 24 45 to 54 65 to 74 75 years 15 to 24 45 to 54 65 to 74 75 years years years years and over years years years and over Europe Belgium, 1994. . . . . . . . . . . . . . . . . . . . . 23 37 43 98 4 17 17 20 Bulgaria, 1998. . . . . . . . . . . . . . . . . . . . . 15 27 48 116 6 9 20 50 Denmark, 1996 . . . . . . . . . . . . . . . . . . . . 13 31 35 71 2 17 16 20 Finland, 1996 . . . . . . . . . . . . . . . . . . . . . 34 55 50 48 7 18 14 7 France, 1996. . . . . . . . . . . . . . . . . . . . . . 13 38 41 87 4 16 15 20 Germany, 1997 . . . . . . . . . . . . . . . . . . . . 13 30 32 71 3 10 13 21 Hungary, 1998 . . . . . . . . . . . . . . . . . . . . 18 93 80 131 4 20 24 50 Ireland, 1996 . . . . . . . . . . . . . . . . . . . . . . 25 19 12 22 5 6 1 2 Italy, 1995 . . . . . . . . . . . . . . . . . . . . . . . . 7 14 23 43 2 5 7 8 Netherlands, 1997 . . . . . . . . . . . . . . . . . 11 16 16 34 4 8 9 12 Norway, 1995 . . . . . . . . . . . . . . . . . . . . . 23 24 32 24 6 10 10 5 Poland, 1996. . . . . . . . . . . . . . . . . . . . . . 17 42 31 31 3 8 7 6 Portugal, 1998 . . . . . . . . . . . . . . . . . . . . 4 7 20 39 1 3 7 9 Russia, 1997 . . . . . . . . . . . . . . . . . . . . . . 53 100 97 97 9 15 20 33 Switzerland, 1996. . . . . . . . . . . . . . . . . . 25 37 47 80 6 16 17 24 United Kingdom, 1997. . . . . . . . . . . . . . 10 14 9 17 2 4 4 4 North America Canada, 1997 . . . . . . . . . . . . . . . . . . . . . 22 27 21 27 5 9 5 4 United States, 1997 . . . . . . . . . . . . . . . . 19 23 26 45 4 7 5 5 Other Australia, 1995 . . . . . . . . . . . . . . . . . . . . 23 22 19 27 6 8 5 5 Chile, 1994 . . . . . . . . . . . . . . . . . . . . . . . 11 14 20 27 2 2 2 1 Israel, 1996 . . . . . . . . . . . . . . . . . . . . . . . 9 9 18 41 2 4 4 15 Japan, 1997 . . . . . . . . . . . . . . . . . . . . . . 11 40 35 53 6 14 19 33 Singapore, 1997 . . . . . . . . . . . . . . . . . . . 9 18 30 74 9 9 14 33 Sources: World Health Organization, various issues of World Health Statistics Annual, and www-nt.who.int/whosis/statistics/whsa/whsa_table4.cfm?path=statistics,whsa,whsa_table4&language=english third of the countries in Table 3-2. their elderly male populations, elderly suicide rates in the world’s The fact that the average woman while Japan, Switzerland, and more developed countries. Few outlives her spouse — coupled Russia have comparatively high nations have experienced the very with studies that show that married rates among elderly women. Levels gradual rise seen in France until the elders are happier than nonmarried for elderly men in the United States mid-1980s, or the downward ten- elders — might lead one to predict are average when compared with dency observed in England and that older women would have other countries, whereas the U.S. Wales (Figure 3-10). More often, higher rates of suicide than older rate for women aged 65 and over is national rates have fluctuated with men, but this is clearly not the case. relatively low. Although some of no perceptible pattern. these differentials may be artificial Among the 23 countries examined, There was a downward trend in eld- due to differences in the reporting Hungary and Russia had the highest erly suicide in the United States for and/or diagnosis of suicide, their suicide rates for elderly men, and nearly half a century prior to 1980. sheer magnitude suggests that real Hungary and Bulgaria had the high- From 1981 to 1988, however, the international differences do exist est rates for elderly women. The suicide rate of older Americans rose and deserve closer study. reported Hungarian rate for men about 25 percent before beginning aged 75 and over is five to seven NO CLEAR TIME TREND IN to decline in the 1990s. Such an times higher than similar rates in ELDERLY SUICIDE WORLDWIDE increase raised questions at a time Ireland, the United Kingdom, and when older people were living Data from the World Health Norway. Belgium and France have longer and were supposedly healthi- Organization for the past 30 years comparatively high rates among er and more financially secure do not show any clear trend in 34 An Aging World: 2001 U.S. Census Bureau
  • 41. (Robinson 1990). Likewise in Japan, Figure 3-10. a 30-year decline in elderly suicide Suicide Rates by Age in England and Wales appeared to level off in the 1980s, and in France: 1965 to 1995 though more recent data show a significant decline in the 1990s. England and Wales However, social scientists remain Rate per 100,000 population in each age group puzzled by the relatively high rates 140 of suicide among elderly Japanese women. The unpredictability of sui- 120 cide trends is perhaps best illustrat- ed by the case of the Netherlands. 100 Dutch society is widely recognized as being more tolerant of voluntary 80 euthanasia than are other Western societies, and one might think it 60 would also have higher rates of Male 75+ recorded suicide. However, the Male 65-74 Female 65-74 country’s rates are lower than the 40 Female 75+ industrialized-country average for most age groups, including the eld- 20 erly, and have not changed greatly during the past 30 years. 0 1965 1970 1975 1980 1985 1990 1995 France Rate per 100,000 population in each age group 140 120 Male 75+ 100 80 60 Male 65-74 40 Female 75+ 20 Female 65-74 0 1965 1970 1975 1980 1985 1990 1995 Source: World Health Organization, various issues of World Health Statistics Annual. U.S. Census Bureau An Aging World: 2001 35
  • 42. CHAPTER 4. Health and Disability Many societies worldwide have expectancy at birth was approxi- DEVELOPING-COUNTRY experienced a change from condi- mately 50 years, and the median TRANSITION APPEARS tions of high fertility and high mor- age at death (the age to which GREATEST IN LATIN AMERICA tality to low fertility and low mortal- 50 percent of females subject to Developing countries are in various ity, a process commonly dubbed the the mortality risks of 1901 could stages of epidemiologic transition. “demographic transition.” Related expect to survive) was about Change is most evident in Latin to this trend is the “epidemiologic 60 years. By 1998, the survivorship America and the Caribbean, where transition,” a phrase first used in curve had shifted dramatically. cardiovascular diseases were the the early 1970s (Omran, 1971) to Female life expectancy had risen to leading cause of death in 29 of 33 refer to a long-term change in lead- 79 years, and the median age at countries examined in 1990 (PAHO, ing causes of death, from infectious death was above 80 years. The 1994). Most deaths from chronic and acute to chronic and degenera- proportion surviving is now quite and degenerative ailments occur at tive. In the classic demographic high at all ages up to age 50, and relatively old ages. Comparative transition, initial mortality declines the survival curve at older ages is data from the mid-1990s (Figure 4-2) result primarily from the control of approaching a much more rectangu- show that half or more of all deaths infectious and parasitic diseases at lar shape as a result of relatively in numerous nations of the Western very young ages. As children sur- more chronic-disease mortality at Hemisphere now occur at ages 65 vive and grow, they are increasingly older ages. and over. exposed to risk factors associated with chronic diseases and acci- dents. As fertility declines and pop- ulations begin to age, the preemi- nent causes of death shift from Figure 4-1. those associated with childhood Survival Curve for U.S. Females: 1901 and 1998 mortality to those associated with Percent surviving older age (Kalache, 1996). 100 Eventually, growing numbers of adults shift national morbidity pro- 90 1998 files toward a greater incidence of 80 chronic and degenerative diseases (Frenk et al., 1989). 70 1901 60 EPIDEMIOLOGIC TRANSITION SHIFTS SURVIVAL CURVE 50 Figure 4-1, which shows survival 40 curves for U.S. females in 1901 and 1998, illustrates a general pattern 30 seen in developed countries. The 20 curve for 1901 represents the early stages of the epidemiologic transi- 10 tion in which the level of infant 0 mortality was high, there was con- 0 10 20 30 40 50 60 70 80 90 100 siderable mortality through the mid- Age dle years, and a gradual increase at Note: Data for 1901 refer to White females. Sources: U.S. Census Bureau, 1936; U.S. Centers for Disease Control, 2001. the later ages. Female life U.S. Census Bureau An Aging World: 2001 37
  • 43. The pace of epidemiologic change are we spending an increasing profound impact on national health, in East and Southeast Asian nations portion of our older years with dis- retirement, and family systems, and is now especially rapid. In the case abilities, mental disorders, and in ill particularly on the demand for of Singapore, where life expectancy health? In aging societies, the long-term care. In the future, health at birth rose 30 years in little over a answer to this question will have a expectancy will come to be as generation (from 40 years in 1948 to 70 years in the late-1970s), the share of cardiovascular deaths rose Figure 4-2. from 5 to 32 percent of all deaths, Proportion of All Deaths Occurring at Ages 65 and while deaths due to infectious dis- Above in 20 Countries: Circa 1995 eases declined from 40 to 12 per- (Percent) cent. Data from Taiwan (Table 4-1) exemplify the typical shift in causes Barbados 75 of death; the infectious and para- Canada 75 sitic diseases that dominated United States 73 Taiwanese mortality in the mid- Cuba 67 1950s have given way to chronic Puerto Rico 62 Argentina 62 and degenerative diseases. By Chile 61 1976, cerebrovascular disease and Saint Lucia 59 cancers had become the leading Trinidad & Tobago 57 causes of death. The situation in Costa Rica 56 the 1996 was similar to that in Paraguay 49 1976, except that the relative Mexico 45 importance of diabetes had risen Suriname 45 substantially while tuberculosis was Brazil 43 no longer a major killer. Although Ecuador 43 reliable data for much of the Bahamas 42 remainder of Asia and for Africa are Colombia 42 lacking, scattered evidence sug- Venezuela 41 gests the increasing importance of Guyana 37 chronic disease patterns in adult Nicaragua 34 populations. Source: Pan American Health Organization, 1998. DOES LONGER LIFE EQUAL BETTER LIFE? Chapter 3 pointed out that continu- Table 4-1. al increases in life expectancy, espe- Rank Order of the Ten Leading Causes of Death in Taiwan: 1956, 1976, and 1996 cially at older ages, have been the norm in most countries worldwide. Order 1956 1976 1996 As individuals live longer, the quali- 1 ....... GDEC1 Cerebrovascular disease Malignant neoplasms ty of that longer life becomes a cen- 2 ....... Pneumonia Malignant neoplasms Cerebrovascular disease tral issue for both personal and 3 ....... Tuberculosis Accidents Accidents 4 ....... Perinatal conditions Heart disease Heart disease social well-being. Are we living 5 ....... Vascular lesions of CNS2 Pneumonia Diabetes mellitus healthier1 as well as longer lives, or 6 ....... Heart disease Tuberculosis Cirrhosis/chronic liver disease 7 ....... Malignant neoplasms Cirrhosis of the liver Nephritis/nephrosis 8 ....... Nephritis/nephrosis Bronchitis3 Pneumonia 1 “Health” is a relative and continually devel- 9 ....... Bronchitis Hypertensive disease Hypertensive disease oping concept which “reflects on the one hand 10 . . . . . . Stomach/duodenum ulcer Nephritis/nephrosis ulcer Bronchitis3 the progress made within the health sciences and on the other hand the meanings, values 1 Includes gastritis, duodenitis, enteritis, and colitis (except diarrhoea of newborns). and prejudices related to health in different 2 CNS refers to the central nervous system. sociocultural contexts.” (Heikkinen 1998; see 3 Includes emphysema and asthma. this source for a useful discussion of the differ- ent orientations toward health that have been Source: Taiwan Department of Health, 1997. adopted in the health sciences). 38 An Aging World: 2001 U.S. Census Bureau
  • 44. An application of this model to data Figure 4-3. from France (Robine, Mormiche, and Survival Without Disease and Survival Without Cambois, 1996) shows that the Disability for French Females: 1981 and 1991 increase in total survival between 1981 and 1991 was generally con- Percent 100 sistent with the increase in disability- free life expectancy, whereas sur- 90 vival without chronic disease 80 showed little change (Figure 4-3). In this case, the differing trajecto- 70 ries of disability and morbidity may be interpreted as support for the 60 theory of dynamic equilibrium, 50 which says that increases in overall life expectancy are driven in part by 40 a reduction in the rate of progres- Total survival in 1981 30 Total survival in 1991 sion of chronic diseases. Survival without disability in 1981 20 Survival without disability in 1991 HEALTHY LIFE EXPECTANCY Survival without disease in 1981 10 Survival without disease in 1991 Since the early 1970s, research has been moving toward the develop- 0 0 10 20 30 40 50 60 70 80 90 100 ment of health indexes that take Age into account not only mortality but Source: Robine, Mormiche, and Cambois, 1996. also various gradations of ill health. As of 1998, 49 nations had esti- mates of healthy life expectancy,3 important a measure as life generated competing theories of an indicator that attempts to inte- expectancy is today. health change, several of which may grate into a single index the mortal- be characterized as: a pandemic of Research into patterns of change in ity and morbidity conditions of a chronic disease and disability mortality, sickness, and disability population. Most estimates of (Gruenberg, 1977; Kramer, 1980); a has suggested that these three fac- healthy life expectancy are derived compression of morbidity (Fries, tors do not necessarily evolve in a from calculations of disability-free 1990); dynamic equilibrium similar fashion. A four-country life expectancy using a methodolo- (Manton, 1982); and the postpone- study (Riley, 1990) notes that in gy pioneered by Sullivan (1971), ment of all morbid events to older Japan, the United States, and Britain, which employs cross-sectional ages (Strehler, 1975). The World mortality decreased and sickness prevalence data but may produce Health Organization has proposed a (morbidity) increased, while in results that underestimate temporal general model of health transition Hungary, mortality increased and trends in a given population. that distinguishes between total sur- sickness decreased.2 Discrepancies Recognizing that these earlier com- vival, disability-free survival, and between the trends in mortality, putational approaches could not survival without disabling chronic morbidity, and disability have capture the full dynamic nature of disease. In other words, it is desir- disability, multistate models have able to quantitatively disaggregate been developed to incorporate 2 The author’s broader review of historical life expectancy into different health data concludes that the relationship between processes such as recovery and falling sick and dying from sickness has shifted states to better understand the rela- over time, and that the link between death and tive health of populations. Thus, a health risks has been unstable rather than sta- ble across time. The risk of being sick has general survival curve such as that 3 The concept of healthy life expectancy as increased as a result of various obvious and typically used refers to expectancy without limi- in Figure 4-1 can be partitioned into tations of function that may be the conse- non-obvious factors — among them earlier and better detection of sickness, declining mortality, different categories that indicate quence of one or more chronic disease condi- and rising real income — which themselves tions. The concept is sometimes called “active constitute valuable human achievements. The overall survival, survival without dis- life expectancy” or “disability-free life expectan- implication is that protracted sickness is a ability, and survival without disease. cy,” to avoid the implication that “healthy” byproduct of such achievements. means “absence of disease.” U.S. Census Bureau An Aging World: 2001 39
  • 45. rehabilitation into the calculations (see, e.g., Manton and Stallard, Box 4-1. 1988; Rogers, Rogers, and Belanger, Network on Health Expectancy 1990). These latter models, howev- To facilitate and promote analyses of health expectancy, an international er, require longitudinal data which network (REVES, the French acronym for Network on Health Expectancy currently are unavailable in most and Disability Process) was formed in 1989 to bring together researchers nations. To date, chronological concerned with the measurement of changes and inequalities in health series are available only for some status, not only within but among nations. REVES has produced numer- developed nations. ous documents and bibliographies of relevant materials, including a sta- In recent years, researchers have tistical yearbook that includes existing estimates of health expectancy in been working toward developing various countries. Further information may be obtained from Jean-Marie integrated, comparable measures Robine, Network Coordinator, INSERM Equipe Demographie et Sante, of healthy life expectancy Centre Val d’Aurelle, 34298 Montpellier Cedex 5, France. (Verbrugge, 1997). Presently, how- ever, it remains impossible to strictly compare estimates among Figure 4-4. nations, due both to different com- Portion of Old Age Lived Without Severe putational methods and, more Disability: Data From the Early 1990s importantly, to differences in con- (Percent) Male Female cepts and definitions that define the basic data. There are impor- tant but not-widely-appreciated 85 Australia distinctions between impairments, 79 disabilities, and handicaps that can 85 Canada lead to different measures of 78 health status (Chamie, 1989). 94 Because “disability” is defined in France 90 many ways, national estimates of disability vary enormously. For 92 Japan example, a compilation by the 87 United Nations (1990a) showed 94 United Kingdom national crude disability rates for 93 the total population ranging from less than half a percent in several Note: Figures refer to the percent of a person's life, after reaching age 65, that she or he might expect to live without needing significant help (personal care) with at least one major activity of daily living. developing countries (Peru, Egypt, Source: Jacobzone, 1999. Pakistan, Sri Lanka) to nearly 21 percent in Austria. Perhaps the and individuals’ ability to function FEMALE ADVANTAGE IN LIFE most commonly-used measure- in everyday life.4 EXPECTANCY PARTIALLY ment tools are scales which assess OFFSET BY DISABILITY the ability of individuals to per- In spite of cross-national compara- form activities of daily living bility problems, several general (ADLs) such as eating, toiletting, observations seem warranted. For and ambulation, as well as instru- 4 ADL measures vary along several dimen- individuals reaching age 65, health sions, including the number of activities consid- mental activities of daily living ered and the degree of independence in per- expectancy varies more than (IADLs) such as shopping and forming physical activities. ADLs do not cover all aspects of disability, however, and are not remaining life expectancy. One using transportation. These meas- sufficient by themselves to estimate the need examination (Kinsella and Taeuber, ures originated in industrialized for long-term care. Some older people have cognitive impairments not measured by ADL 1993) of REVES data (See Box 4-1) societies where debate has cen- limitations, which may or may not be captured showed that the range of life tered on long-term care systems by IADL measures. And, of course, there are many questions regarding the validity and expectancy at age 65 varied by applicability of such measures in different cul- about 12 percent among developed tural settings. 40 An Aging World: 2001 U.S. Census Bureau
  • 46. (Freedman and Soldo, 1994) found Figure 4-5. declines in less-severe disability Number of Chronically Disabled People Aged 65 (i.e., in IADLs) during the 1980s. and Over in the United States: 1982 to 1996 Data from Australia, on the other hand, revealed that the increase in Millions 10 years of disability between 1981 and 1988 was greater than the If disability rate did not change since 1982 overall increase in life expectancy. 8 Data for Canada and Finland sug- gest that changes in disability-free life expectancy have been and Based on declines in chronic disability remain stagnant (Robine and 6 rate occurring since 1982 Romieu, 1998). Researchers posited a number of potential factors — 4 other than actual increases in chronic disease incidence and meas- urement error — which might have contributed to stagnation or 2 declines in healthy life expectancy, including increased survival of chronically ill individuals due to 0 1982 1989 1994 1996 improvements in medical care, earli- er diagnosis or treatment of chronic Source: Manton, Corder, and Stallard, 1997. diseases, greater social awareness of disease and disability, earlier countries for both men and women. spend a slightly greater proportion adjustment to chronic conditions However, the percent of this time of their remaining years in a severe- due to improved pension and health spent in “good health” (free of prob- ly disabled state relative to elderly care/delivery systems, and rising lems with personal and instrumen- men, thus negating some of the expectations of what constitutes tal activities of daily living) had a potential benefit of their higher life good health or normal functioning much wider range, from 45 to more expectancy (Figure 4-4). Other (Mathers, 1991; Verbrugge, 1989). than 80 for males, and from 37 to studies of gender differences in the More recent data and rigorous 76 for females (these figures incidence of disabling conditions at analyses, however, now strongly exclude apparent outlier estimates older ages support this contention suggest that rates of disability in a which would widen the ranges even (Heikkinen, Jokela, and Jylha, 1996; number of developed countries are further). For men in developing Dunlop, Hughes, and Manheim, declining. In the United States, countries, the estimated range of 1997; Robine and Romieu, 1998). researchers (Manton, Corder, and remaining life expectancy at age 65 Stallard, 1997) used data from the was 12 to 15 years. However, the DEVELOPED-COUNTRY DISABILITY RATES AT OLDER 1982, 1984, 1989 and 1994 estimated percent of remaining rounds of the U.S. National Long AGE SEEN TO BE DECLINING years spent in relative health varied Term Care Survey to demonstrate from less than 60 to 88 percent. In nations where time series esti- that the disability rate among peo- For elderly women, the variation in mates of health expectancy are ple aged 65 and over declined over life expectancy was 28 percent, available (e.g., Canada, the United the 12-year period, such that there while the range of healthy remain- States, Australia, England/Wales), were 1.2 million fewer disabled ing life was 50 to 87 percent. the general view in the 1980s was older people in 1994 than would one of uncertainty regarding the Available but geographically limited have been the case if the 1982 relationship between rising life data from the early 1990s suggest a rate had not changed (Figure 4-5). expectancies and trends in health further tentative statement: women Five other U.S. surveys, while vary- expectancies. One comprehensive reaching age 65 can expect to ing in content and nature review of data in the United States U.S. Census Bureau An Aging World: 2001 41
  • 47. (e.g., both longitudinal and cross- obviously have substantial implica- accidents and the consequences of sectional), have yielded findings tions for public and private health drug and alcohol abuse. As that support a temporal decline. programs and expenditures, and economies in developing countries Likewise, a U.S. study of changes possibly for the conceptualization expand and the demographic and between 1982 and 1993 in self- and definition of disability itself.5 epidemiologic pictures change, we reported ability to work found sig- might expect to see related Many countries with aging popula- nificant improvement among both changes in the nature and preva- tions now recognize the need for men and women who were in their lence of various disabilities. longitudinal surveys as a means of sixties (Crimmins, Reynolds, and understanding adult health pat- Numbers of disabled people are Saito, 1999). Another U.S. study terns, transitions to and from differ- almost certain to increase as a cor- (Freedman and Martin, 1998) of ent health statuses, and how to dif- relate of sheer population growth. the effect of changes in living envi- ferentiate between morbidity and Figure 4-6 illustrates the applica- ronments, device use, and survey aging per se (Svanborg, 1996). tion of empirical gross disability design on functional ability among While such survey efforts involve rates to the projected population noninstitutionalized people aged substantial economic investment, of the Philippines. This simplistic 50 and over concludes that these the potential cost savings in policy example assumes that disability changes alone could not account design and implementation would rates for men and women as meas- completely for improved function- seem to dwarf the initial expense. ured in 1980 will remain constant ing, and that there has indeed And as various national longitudinal in the future. Even with no provi- been some improvement in under- analyses expand, both geographi- sion for higher rates of disability lying physiological capability. cally and in terms of specific dis- as the population ages, the project- Increased levels of education have abling conditions, all health sys- ed absolute increases are alarming been identified as a potentially tems stand to benefit from more in terms of future service and care powerful factor influencing disabil- comprehensive comparisons and requirements. ity decline in the United States. the resultant implications for pro- A review of trend data from nine gram priorities. ESTIMATING THE BURDEN OF DISEASE other developed countries plus Taiwan concludes that, with a few DEVELOPING-COUNTRY A major ongoing effort to under- DISABILITY BURDEN LIKELY TO exceptions by gender, disability is stand and predict the effect of epi- INCREASE AS POPULATIONS declining among the elderly else- AGE demiologic change is the Global where as well (Waidmann and Burden of Disease Project under- Manton, 1998), as indicated by the Two decades ago, the World Health taken jointly by the World Health French data in Figure 4-3. Organization noted a distinction in Organization, Harvard University, Researchers increasingly have dis- prominent causes of disability and the World Bank. Using a com- aggregated disability into more- between developed and developing putational (and controversial; see, severe versus less-severe cate- countries. In the latter, disability e.g., Black and McLarty, 1996; gories, and the current consensus was said to stem primarily from Cohen, 2000) concept known as in developed countries is that the malnutrition, communicable dis- Disability-Adjusted Life Years overall decline in disability is prima- eases, accidents, and congenital (DALYs), this study attempts to rily the result of decreases in the conditions. In industrialized coun- measure global, regional, and more-severe forms (Robine and tries disability resulted largely country-specific disease burdens in Romieu, 1998). Freedman, Aykan, from the chronic diseases dis- a baseline year, and to project such and Martin (2001), for example, cussed earlier — cardiovascular burdens into the future. Figure 4-7 have recently demonstrated a disease, arthritis, mental illness, highlights the change in the esti- decline between 1993 and 1998 in and metabolic disorders, as well as mated (1990) and projected (2020) severe cognitive impairment among rank order of disease burden for the noninstitutionalized population 5 For a description of changes and paradigm the ten leading disease categories aged 80 and over in the United shifts in disability policy in certain developed on a global basis. In addition to countries, see Kalisch, Aman, and Buchele, States. Such trends, if sustained, 1998. underscoring the expected shift 42 An Aging World: 2001 U.S. Census Bureau
  • 48. from communicable to non- Figure 4-6. communicable disease patterns Projected Numbers of Disabled Males in the (largely driven by the changing sit- Philippines by Age: 1991 and 2020 uation in developing countries; 1991 Murray and Lopez, 1997), this 2020 study raises specific warning flags Thousands for many countries in terms of a 700 likely increase in disease burden due to neuro-psychiatric conditions 600 and accidents. Among the many results, several stand out: (1) the 500 projected prominence of tobacco use vis-a-vis mortality (as men- 400 tioned in Chapter 3) and through various disease vectors, with the 300 expectation that tobacco will kill more people than any single dis- 200 ease by 2020; (2) the concentra- tion of disease burden in certain 100 developing regions. For example, the inhabitants of India and Sub- 0 Saharan Africa were estimated to 0-4 5-14 15-24 25-34 35-44 45-54 55-64 65-74 75+ bear more than 40 percent of the Sources: United Nations, 1990 and U.S. Census Bureau, 2000a. world’s disease burden in 1990, while constituting only one-fourth of the world’s population. The Figure 4-7. study also points out the fallacy Change in Rank Order of Disease Burden for Top Ten that noncommunicable diseases Leading Causes in the World: 1990 and 2020 are necessarily related to affluence; (Disease burden measured in disability-adjusted life years) the likelihood of dying from a non- communicable disease among 1990 2020 adults under age 70 in both India Rank Disease or injury Disease or injury and Sub-Saharan Africa is greater 1 Lower respiratory infections Ischemic heart disease than in Western Europe; and (3) the 2 Diarrhoeal diseases Unipolar major depression vastly underestimated importance 3 Conditions arising during Road traffic accidents of mental illnesses as major and the perinatal period increasing sources of disease bur- 4 Unipolar major depression Cerebrovascular disease den, which implies significant 5 Ischemic heart disease Chronic obstructive pulmonary disease 6 Cerebrovascular disease Lower respiratory infections long-term care challenges in aging 7 Tuberculosis Tuberculosis societies (Murray and Lopez, 8 Measles War 1996). 9 Road traffic accidents Diarrhoeal diseases 10 Congenital anomalies HIV Source: Murray and Lopez, 1996. U.S. Census Bureau An Aging World: 2001 43
  • 49. Figure 4-8. Share of Population Versus Health Expenditure Percent of total by Age in Nine Countries: 1993 Population Health expenditure Australia Sweden 88 82 <65 <65 66 62 12 18 65+ 65+ 34 38 4 8 75+ 75+ 20 21 Finland Switzerland 86 86 <65 <65 62 60 14 14 65+ 65+ 38 40 6 6 75+ 75+ 22 26 Germany United Kingdom (England) 85 84 <65 <65 68 58 15 16 65+ 65+ 32 42 6 7 75+ 75+ 16 27 New Zealand United States 89 87 <65 <65 67 63 11 13 65+ 65+ 33 37 5 5 75+ 75+ 21 21 Portugal 86 <65 64 14 65+ 36 5 75+ 19 Source: OECD, 1997. 44 An Aging World: 2001 U.S. Census Bureau
  • 50. AGING INCREASES HEALTH is included in public health schemes among noninstitutionalized elderly, CARE COSTS (e.g., Germany), per capita expendi- per capita health expenditure often Population aging might be expected tures for people aged 65 and over peaks at ages 75 to 79 and declines to increase costs of health care in may be two to three times higher thereafter. Costs per service (such most societies because health than those for younger people; in as hospital stays and medicine pre- expenditures by and for older age countries with more inclusive long- scriptions) for older people are less groups tend to be proportionally term care coverage (e.g., Australia than for the population as a whole, greater than their population share and Finland), the ratio is four times although usage rates for older peo- (Figure 4-8). This expectation higher (OECD, 1997). ple are much higher and hence the applies especially to nations where per capita costs are higher (OECD, Although the picture of rising health acute care and institutional (long- 1997). Governments and interna- care costs with age is accurate in a term care) services are widely avail- tional organizations are now recog- general sense, disaggregation of able. Cross-national comparative nizing the need for cost-of-illness data by age shows some surprising data on health care expenditures by studies on age-related diseases, in facets. A large fraction of health age are relatively uncommon, but part to anticipate the likely burden care costs associated with advanc- ongoing work by the Organization of increasingly-prevalent and expen- ing age is incurred in the period for Economic Co-Operation and sive chronic conditions (of which just prior to death, and since an Development has begun to docu- Alzheimer’s disease may be the increasing proportion of people are ment age-specific differentials. most costly), and in part to under- living to very old age, overall health Table 4-2 shows that per capita stand the potentially salubrious care costs rise with age. health expenditures for people aged effects that may accrue to future Treatments to prolong life have 65 and over are uniformly higher generations due to higher levels of made once-certain death much less than for the nonelderly, and that education and access to information certain, but there is some indication this difference varies by country. about healthier lifestyle behaviors. that health care costs taper off at Much of the between-country differ- very old ages (OECD, 1998b), sug- EARLY-LIFE CONDITIONS ence may be attributed to variations gesting that life may be prolonged AFFECT ADULT HEALTH in program coverage. In nations up to a point but that treatment is where relatively little long-term care The last decade has seen a rapidly not desired indefinitely. Likewise, growing interest in examining adult health outcomes from a life-course Table 4-2. Relative Per Capita Health Expenditure by Age Group in 12 perspective. Researchers increas- Countries: Circa 1993 ingly suggest that many negative (0-64 = 1.0) health conditions in adulthood stem from risks established early in life Country 65-74 65+ 75+ (Elo and Preston, 1992). Some Australia . . . . . . . . . . . . . . . . . . 2.8 4.0 6.0 (notably, Barker, 1995) argue that Finland. . . . . . . . . . . . . . . . . . . . 2.8 4.0 5.5 France . . . . . . . . . . . . . . . . . . . . 1 2.2 2 3.0 3 3.7 adult health has a fetal origin, Germany . . . . . . . . . . . . . . . . . . 2.3 2.7 3.2 wherein nourishment in utero and 4 3 Japan . . . . . . . . . . . . . . . . . . . . . 3.1 4.8 5.7 during infancy has a direct bearing Netherlands . . . . . . . . . . . . . . . (NA) 4.4 (NA) New Zealand . . . . . . . . . . . . . . 2.3 3.9 6.2 on the development of risk factors Portugal . . . . . . . . . . . . . . . . . . . 1.4 1.7 2.1 for adulthood diseases (especially Sweden . . . . . . . . . . . . . . . . . . . 2.3 2.8 3.4 Switzerland . . . . . . . . . . . . . . . . 2.6 4.0 5.7 cardiovascular diseases). Childhood United Kingdom (England). . . 2.5 3.9 5.6 infections may have long-term United States . . . . . . . . . . . . . . 3.1 4.2 5.2 effects on adult mortality. A World NA Not available. Health Organization report states 1 Refers to ages 60-69. unequivocally that slow growth and 2 Refers to ages 60+. lack of emotional support in prena- 3 Refers to ages 70+. 4 Refers to ages 65-69. tal life and early childhood reduce Note: Data are relative to the level of per capita spending for people aged 0-64, which physical, cognitive, and emotional is set at 1.0. functioning in later years (Wilkinson Source: OECD, 1997. and Marmot, 1998), as do certain U.S. Census Bureau An Aging World: 2001 45
  • 51. parental behaviors (particularly industrialized world6 has identified example, if marriage equates with smoking and alcohol consumption) a number of socioeconomic factors better health among older and socioeconomic circumstances that affect health and longevity: individuals, do rising rates of (e.g., poverty). Parental divorce has people with higher education tend divorce and increased proportions been linked to decreased longevity to live longer (Kitigawa and Hauser, of never-married individuals por- of children (Schwartz et al., 1995). 1973); being married encourages tend poorer average health? And healthier behaviors in U.S. adults, what of other life dimensions? A While it seems intuitive that child- including people in old age (relative considerable amount of current hood conditions should affect adult to other marital statuses), and the research is focused on not only development and health outcomes, effects may be greater for older social but also psychological and separating cohort effects from peri- men than for older women (Schone biological pathways by which od effects (e.g., from changing liv- and Weinick, 1998); there are clear socioeconomic status affects health ing conditions) is empirically diffi- gradients in the United Kingdom in (see, e.g., Adler et al., 1999). cult. Indeed, some evidence both mortality and health when bro- suggests that current conditions While the weight of existing studies ken down by social class, i.e., high- may be more important than early- clearly supports a strong relation- er social/occupational class is relat- life conditions; Kannisto (1996) has ship between social and economic ed to better health and lowered found period effects to be much factors on the one hand and health mortality risks (Devis, 1993; more significant than cohort effects and mortality outcomes on the Wilkinson and Marmot, 1998); and on oldest-old mortality (i.e., after other, this relationship may not be among the oldest old in Sweden, age 80). And in a study of cohorts as strictly predictive as some have former white-collar workers had born just before, during, and just suggested. Research on marital sta- better physical functioning than for- after a severe famine in Finland in tus and health among the elderly in mer blue-collar workers (Parker, the mid-1860s, the researchers the United States, for instance, has Thorslund, and Lundberg, 1994). In found no major differences in later- shown that while widowhood is in the United States (Crimmins, life survival; the extreme nutritional fact associated with poorer health, Hayward, and Saito, 1996) and in deprivation in utero and during single women are likely to have bet- several European countries (Robine infancy appeared not to translate ter health outcomes than married and Romieu, 1998), socioeconomic into higher adult mortality risks women (Goldman, Korenman, and conditions more strongly affect (Kannisto, Christensen, and Vaupel, Weinstein, 1994). One Japanese functional change than mortality, 1997). Such findings are likely to study (Sugisawa, Liang, and Liu, which means that socioeconomic stimulate considerable future 1994) found no significant effect of differences in active or healthy life research to explore the linkages marital status on the risk of dying; expectancy are greater than those between lingering effects of early- however, higher levels of social par- in total life expectancy. Such find- life and survival at advanced ages ticipation of older people were ings might be expected to have (Vaupel, 1997). strongly linked to lowered mortality implications for the future health risks. Research in Florence, Italy SOCIOECONOMIC CORRELATES status of elderly populations. For and Tampere, Finland uncovered no OF MORTALITY AND HEALTH systematic association between If early-life factors affect future 6 Research on socioeconomic correlates of functional ability levels and educa- mortality and health in developing countries is health and survival, then socioeco- still sparse. Wu and Rudkin (2000) found that tion/previous occupation nomic differences in childhood and lower socioeconomic status was associated (Heikkinen, Jokela, and Jylha, 1996). with poorer health among Malaysians aged 50 throughout life are likely to play an and over, and that this association held for all In a study of nine industrialized intrinsic role. A diverse and long- three major ethnic groups (Malay, Chinese, and nations, differences in mortality by Indian). Liang et al. (2000) have found evi- standing literature from the dence of a socioeconomic gradient in old-age educational level were found to be mortality in China, but they note that there are fairly small in the Netherlands and major differences between developed and developing countries in terms of major health three Scandinavian nations, but parameters, and that much more work needs to be done in validating gradients in Third World settings. 46 An Aging World: 2001 U.S. Census Bureau
  • 52. much more substantial in larger higher mortality in Northern results point to the importance of countries such as the United States, European countries, but no such understanding population diversity France, and Italy. The authors sug- relationship in France, Switzerland, within countries, and suggest that gest that between-country differ- and Mediterranean nations (Kunst et policy planners pay particular atten- ences may be related to different al., 1999). Another multicountry tion to socioeconomic differences social and economic policies. A study of income and mortality by gender and among subgroups 12-country study of occupational implies that the effect of income on when developing intervention class and ischemic heart disease mortality is largely determined by strategies (see Sacker et al., 2001 mortality found the expected rela- the distribution of income within a and National Research Council, tionship between lower class and given nation (Duleep, 1995). Such 2001 for further discussion). U.S. Census Bureau An Aging World: 2001 47
  • 53. CHAPTER 5. Urban and Rural Dimensions Urbanization is one of the most economic and political changes. both sending and receiving areas. significant population trends of the Closed national economies and Definitions of urban and rural resi- last 50 years. Though we may trading blocs have given way to dence often differ greatly from one think of cities as synonymous with open economies that increasingly country to the next, which compli- historical development, not until the are international in scope. In gener- cates global and regional discus- nineteenth century did substantial al, the most rapidly-growing sions of urbanization. The view of portions of national populations live economies since 1950 also are the United Nations has been that and work in large cities, and only in those with the most rapid increase “differences in definition may reflect certain parts of the world. In 1900, in their levels of urbanization. The differences in the characteristic fea- about 14 percent of the world’s world’s largest cities tend to be con- tures of urban and rural settlements population lived in cities (United centrated in the world’s largest considered most relevant in individ- Nations, 1991), and this percentage economies (United Nations Centre ual countries” (United Nations, was still below 20 by 1950. for Human Settlements, 1996). 1973). In spite of definitional incon- However, the global population of Urbanization is linked to changes in sistencies, the basic questions con- all ages living in urban areas (as the socioeconomic profile of a cerning aging are similar in all soci- defined by each country) more than workforce as workers shift from eties: are the elderly increasingly doubled between 1950 and 1975, predominantly agricultural pursuits concentrated in particular areas? If and increased another 55 percent to industrial employment and then so, what are the implications for from 1975 to 1990. By 1995, to services. Clark and Anker (1990) social support and delivery of serv- about 46 percent (2.6 billion) of the have shown that urbanization is ices? For individual cities, do earth’s people lived in urban areas related to decreased participation of changes in age structure bring (United Nations, 1998). Soon after older people in the labor force. In about demands to reorder budget the year 2000, the world likely will developed countries, this decrease priorities? have more urban dwellers than accompanies a decline in manufac- rural dwellers. About three-fourths turing employment, an increased DEVELOPED-COUNTRY ELDERLY ABOUT THREE- of the population in developed prevalence of early retirement FOURTHS URBAN; ONE-THIRD countries is urban, compared with schemes, and lower levels of educa- IN DEVELOPING COUNTRIES slightly more than one-third in tion and job flexibility among older developing countries as a whole. workers relative to younger work- In keeping with the worldwide pat- However, the pace of urbanization ers. In urban areas of developing tern of increased urbanization, the is much faster in the developing nations, the increased importance elderly population has become world. And while the urban growth of the formal sector tends to more concentrated in urban areas rate in most world regions has exclude older workers who find it during the past 50 years. In devel- begun to decline, some parts of the difficult to compete with better- oped countries as a whole, an esti- globe (notably Africa and South educated younger workers. With mated 73 percent of people aged Asia) are now experiencing peak urbanization come changes in the 65 and over lived in urban areas in rates of urban growth. In spite of family unit and kinship networks 1990, and this figure is projected to declining growth rates, the world’s that have both beneficial and reach 80 percent by the year 2015. urban population is projected to adverse consequences for the well- In developing nations, which still virtually double between 1995 and being of elderly members. are predominantly rural, just over 2030, reaching a projected level of one-third (34 percent) of people Urban growth affects all age groups aged 65 and over were estimated to 5.1 billion people (United Nations, of a population. Since urbanization live in urban areas in 1990. This 1998). often is driven by youthful migra- proportion is expected to exceed Twentieth-century trends in urban- tion from rural areas to cities, it one-half by the year 2015 (United ization have stemmed from broad influences the age distribution in Nations, 1991). U.S. Census Bureau An Aging World: 2001 49
  • 54. The elderly of Africa are more likely to live in rural areas than are the Figure 5-1. elderly of other regions, even Percent of Total and Elderly Population Total though the African population over- Living in Rural Areas in 39 Countries Aged 65+ all is slightly more urbanized than Africa that of the Asia/Oceania region 54 Botswana, 1991 70 (excluding Japan). The aggregate South Africa, 1991 43 37 trend toward urbanization is Tunisia, 1994 39 42 stronger in Asia than in Africa, how- Zambia, 1990 61 85 ever. Half of the Asia/Oceania eld- Zimbabwe, 1992 69 86 erly are projected to live in cities by Asia 2015, compared with 42 percent Bangladesh, 1991 80 85 in Africa. 74 China, 1990 76 69 As a region, Latin America and the Indonesia, 1990 74 22 Caribbean is already highly urban- Japan, 1995 28 49 ized. The proportion of elderly in Malaysia, 1991 54 51 urban locales is very similar to that Philippines, 1990 54 22 of the developed-country average. South Korea, 1995 43 81 Unlike in other developing areas, Thailand, 1990 84 41 the elderly in Latin America and the Turkey, 1990 51 Caribbean have been somewhat Europe more likely than the general popula- Austria, 1991 35 31 tion to live in cities (Heligman, France, 1990 26 31 Chen, and Babakol, 1993). Greece, 1991 41 49 28 Norway, 1990 32 ELDERLY MORE LIKELY THAN 17 Sweden, 1990 NONELDERLY TO LIVE IN 17 RURAL AREAS Former Soviet Union 32 Armenia, 1992 32 Despite the increasingly urban 46 Azerbaijan, 1989 45 nature of today’s elderly popula- 31 Belarus, 1997 53 tions, rural areas remain dispropor- 30 Estonia, 1996 31 tionately elderly in a majority of 44 Georgia, 1993 countries. In most nations, this is 51 45 Kazakhstan, 1996 primarily the result of the migration 42 66 of young adults to urban areas, and Krygyzstan, 1997 60 31 to some extent of return migration Latvia, 1996 33 32 of older adults from urban areas Lithuania, 1996 44 53 back to rural homes. Data for 39 Moldova, 1989 65 27 countries from the period 1989 to Russia, 1995 32 71 1997 show that the percent of all Tajikistan, 1993 64 32 elderly living in rural areas was Ukraine, 1995 44 59 higher than the percent of total Uzbekistan, 1989 52 population in rural areas in 27 of Latin America/Other 43 the 39 nations, with no difference Bolivia, 1992 59 in 4 nations (Figure 5-1). Five of 24 Brazil, 1991 24 the eight countries where the elder- 15 Chile, 1997 17 ly were less likely than the total 45 Ecuador, 1990 49 population to live in rural areas are New Zealand, 1991 15 10 predominantly-Muslim nations that United States, 1990 25 25 were formerly part of the Soviet Source: U.S. Census Bureau, 2000a. Union. Differences in the share of 50 An Aging World: 2001 U.S. Census Bureau
  • 55. total versus elderly population Figure 5-2. residing in rural areas are most Urban Sex Ratios for Persons Aged 65 striking in Belarus, Bolivia, and Over: 1986 or Later Botswana, Zambia, and Zimbabwe. (Males per 100 females) Africa SEX RATIOS OF URBAN Botswana, 1991 69 ELDERLY TYPICALLY LESS Cote d'Ivoire, 1988 106 THAN 80 Egypt, 1986 123 Kenya, 1989 106 An examination of national data for Malawi, 1987 109 54 countries from the period 1986 South Africa, 1991 79 to 1997 shows that more elderly Tanzania, 1988 89 Tunisia, 1994 99 women than elderly men have been Zambia, 1990 134 recorded in urban areas in all Zimbabwe, 1992 108 except eight nations, six of which Asia Bangladesh, 1991 137 are in Africa. Sex ratios (number of China, 1990 86 men per 100 women) for the urban Indonesia, 1990 82 elderly usually are well below 100 Israel, 1994 76 Japan, 1995 70 (Figure 5-2), and are below 50 in Malaysia, 1991 80 parts of the former Soviet Union. Pakistan, 1990 113 Philippines, 1990 77 ELDERLY MEN MORE LIKELY Singapore, 1990 81 THAN ELDERLY WOMEN TO South Korea, 1995 55 Thailand, 1990 75 RESIDE IN RURAL AREAS Turkey, 1990 77 Since women live longer than men Europe Austria, 1991 51 virtually everywhere, we might Bulgaria, 1994 77 expect to see sex ratios of less than Czech Republic, 1994 61 100 for the elderly throughout a France, 1990 62 Greece, 1991 73 given population. Although older Hungary, 1995 59 women outnumber older men in Norway, 1990 66 almost every nation, the ratio of Poland, 1994 57 Sweden, 1990 69 older men to older women general- Former Soviet Union ly is higher in rural areas than in Armenia, 1992 66 cities. In the rural areas of some Belarus, 1997 49 Estonia, 1996 48 countries — e.g., New Zealand, Georgia, 1993 54 Paraguay, and Sweden — older men Latvia, 1996 46 52 actually outnumber older women. Lithuania, 1996 Moldova, 1989 54 This rural male surplus is seen most Russia, 1995 44 prominently in many countries of Ukraine, 1995 50 Latin America/Caribbean/Other Latin America and the Caribbean, Argentina, 1995 67 which suggests region-specific pat- Australia, 1986 70 terns of gender-specific migration Bolivia, 1992 77 that have implications for health Brazil, 1991 76 Canada, 1991 67 and social security systems in both Chile, 1997 64 rural and urban areas. Colombia, 1993 80 Costa Rica, 1995 77 Ecuador, 1990 82 Mexico, 1995 81 New Zealand, 1991 70 Peru, 1993 87 United States, 1990 64 Uruguay, 1990 67 Source: U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 51
  • 56. Conversely, elderly women tend to be somewhat more likely than eld- Figure 5-3. erly men to live in urban areas Percent of All Elderly Living in Urban Areas (Figure 5-3).1 The gender difference by Sex in Ten Countries Male in residential concentration proba- Female bly is related to marital status and Developed countries health. As discussed in Chapter 6, 68 elderly women are much more like- France, 1990 72 ly than elderly men to be widowed, 87 and also are more likely to have New Zealand, 1991 92 chronic illnesses. One study of the 54 Poland, 1994 elderly in developed countries 57 (Kinsella and Taeuber, 1993) noted 69 Russia, 1995 an inverse relationship between 67 widowhood rates and sex ratios in 73 United States, 1990 77 urban areas. Urban residence may provide elderly women, especially Developing countries widows, the benefits of closer prox- 24 imity to children and/or to social China, 1990 24 and health services. 67 Colombia, 1993 75 SUBNATIONAL URBAN/RURAL 55 DIFFERENCES IN AGE South Korea, 1995 59 STRUCTURE MAY BE STRIKING 56 Tunisia, 1994 The profile of aging in subnational 61 areas may be very different when 16 Zimbabwe, 1992 examined in view of urban/rural dif- 13 ferences. In Russia’s 73 oblasts Source: United Nations Demographic Yearbook, 1996. (administrative areas), for example, the rural population of many oblasts is skewed in favor of older the same as the overall national population. One reason that rural people.2 Figure 5-4 displays the age average. The majority of Kursk’s areas have such high proportions in and sex distribution of urban ver- urban population is concentrated in older age groups is out-migration of sus rural populations in the Kursk the working ages, which is not true younger people to urban areas in Oblast, located in the Central in the rural areas. These pyramids search of work. Out-migration of Chernozem Region bordering also show disparities in sex compo- young people may leave older Ukraine. The pyramid for the rural sition. Nearly 31 percent of Kursk’s women and men without the direct population has a particularly odd rural females are aged 65 and older, support of their family. Harsh living shape, with people aged 65 and compared with just 15 percent of conditions and lack of amenities in older accounting for nearly one- rural males. The sex ratio for the many rural areas pose additional fourth of the total population. In rural elderly population is 41 men difficulties for the elderly. In 1996, contrast, the urban population of per 100 women. for instance, official statistics indi- Kursk is 12 percent elderly, about cate that only 23 percent of Russia’s Kursk is not the only Russian oblast rural population had running water to have a large proportion of its 1 Grewe and Becker (2000), in a preliminary in their homes, and only 3 percent examination of within-country migration rates rural population, particularly had indoor toilet facilities. for 65 countries with 128 observations from women, in older age groups. In the period 1952-1996, also have noted a likely trend toward disproportionate shares of elderly several other oblasts, more than Skewed age structures may present women in urban areas, and of elderly men in one-fourth of the rural female popu- problems for certain localities in rural areas. 2 Although Russia is highly urbanized (over lation is aged 65 or over. In seven terms of the provision of services 70 percent of the population lives in urban oblasts, elderly women account for and aid to older people. Skeldon areas), its rural population remains substantial, numbering approximately 40 million in 1996. 30 percent of the entire rural female (1999) has noted, based on the 52 An Aging World: 2001 U.S. Census Bureau
  • 57. Figure 5-4. Urban and Rural Population in Kursk Oblast: 1996 Age Urban 85+ Male Female 80-84 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4 40 30 20 10 0 10 20 30 40 Rural 85+ 80-84 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4 40 30 20 10 0 10 20 30 40 Thousands Source: U.S. Census Bureau, 2000a. experiences of Japan and Korea, the important emerging issues will residents. In countries where the confluence of overall population revolve around the social conditions youthful influx of rural-to-urban aging with rural depopulation and of elderly individuals in relatively migrants has slowed in recent stagnation of small and medium- isolated rural areas. decades, many cities may now have sized towns, and suggests that this aging populations (Chesnais, 1991). pattern will be seen increasingly NO CLEAR TREND TOWARD Conversely, in countries where DISPROPORTIONATE AGING OF throughout Asia in the first half of urbanization rates remain high and LARGE CITIES this century. A similar situation has younger residents continue to gravi- been noted by Golini (2000) with Although rural areas tend to be dis- tate toward cities, one would regard to Italy. While there are few proportionately elderly compared expect the proportion of elderly in if any negative national economic with urban areas in general, data cities to be lower than for the coun- consequences associated with this for some large cities reveal a rela- try as a whole. Data for 13 major development, it seems clear that tively high proportion of elderly cities (Table 5-1), however, do not U.S. Census Bureau An Aging World: 2001 53
  • 58. lend themselves to a clear interpre- Table 5-1. tation of trends. The populations of Percent of Older Population in 13 Cities Compared With Respective National Average Budapest, London, and Moscow are in fact older than their respective Age City Country City national averages, but this is not Year group percent percent the case in Berlin, Paris, and Tokyo. Bangkok and Harare are younger Bangkok, Thailand. . . . . . . . . . . . . . . . . . . . . . . . . . 1995 65+ 4.2 5.4 than Thailand and Zimbabwe as a Beijing, whole, but a similar relationship China . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1990 65+ 6.2 5.6 does not hold in the Chinese cities Berlin, of Beijing and Shanghai, nor in Germany . . . . . . . . . . . . . . . . . . . . . . . . . . 1993 65+ 13.7 15.1 Mexico City. In the United States in Budapest, general, census data from 1960 Hungary . . . . . . . . . . . . . . . . . . . . . . . . . . 1990 60+ 21.5 18.9 through 1990 indicate that the cen- Buenos Aires, tral cities of metropolitan areas Argentina . . . . . . . . . . . . . . . . . . . . . . . . . 1991 65+ 8.2 8.9 have, on balance, consistently lost Harare, elderly migrants to nonmetropolitan Zimbabwe . . . . . . . . . . . . . . . . . . . . . . . . 1992 65+ 1.6 3.3 areas (Fuguitt and Beale, 1993). Greater London, United Kingdom . . . . . . . . . . . . . . . . . . . 1991 65+ 14.4 10.0 ARE RURAL ELDERLY Mexico City, DISADVANTAGED? Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . 1995 65+ 5.2 4.4 The exodus of younger people from Moscow, Russia . . . . . . . . . . . . . . . . . . . . . . . . . . . 1989 65+ 12.0 9.6 the countryside to cities raises the New York, rural proportion of elderly residents United States . . . . . . . . . . . . . . . . . . . . . 1990 65+ 13.0 12.6 in many countries. As a result, tra- Paris, ditional family support systems for France . . . . . . . . . . . . . . . . . . . . . . . . . . . 1990 60+ 15.7 19.9 the frail elderly may change. Shanghai, Younger family members living in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1990 65+ 10.1 5.6 urban areas are unlikely to provide Tokyo, direct care for distant elders remain- Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1990 65+ 9.4 12.0 ing in rural areas. At the same Note: Data for Mexico City refer to the Federal District. time, younger family members who Source: Compiled by the U.S. Census Bureau from national statistical volumes. move to cities may have improved financial resources that can be used Wales (Wenger, 1998) suggest that MIGRATION PATTERNS OF to help elderly relatives still living in rural dwellers are more likely than OLDER PEOPLE NOT WELL the rural birthplace. DOCUMENTED their urban counterparts to be Quality-of-life issues for older popu- involved in community and volun- International migration of elderly lations in rural versus urban areas tary activities. Nevertheless, the people, as noted in Chapter 2, is are beginning to receive additional provision of health and other sup- not a significant demographic factor attention as migration streams portive services to ill and disabled in many countries. Within-country increase and the costs of health older people in rural areas contin- migration, however, may be sub- care and public benefits escalate. ues to present special challenges. stantial. One common perception Whereas graying rural communities Perhaps because of these difficul- of older people is that they tend to were once associated with negative ties, the percentage of older dis- be much less mobile than younger socioeconomic consequences, more abled people remaining in the people, typically “aging in place” in recent research in developed coun- community without being institu- communities that have been home tries has considered positive results tionalized is lower in predominantly for many years. While this may be that may stem from increased pro- rural areas than in urban areas true in a general sense, various portions of increasingly affluent eld- (Suzman, Kinsella, and Myers, national studies suggest that geo- erly (Bean et al., 1994). Data from 1992). graphical mobility among older 54 An Aging World: 2001 U.S. Census Bureau
  • 59. people is increasing. In the 1970s, more prevalent as levels of adults, there is mounting evidence the mobility of the Japanese popula- education and retirement income that the movement of older people tion declined in all age groups with increase. In the United States, the to urban areas is becoming signifi- the exception of the elderly (Otomo oldest old have been seen to move cant (Myers and Clark, 1991). and Itoh, 1989). Census data for more frequently than younger elder- There is much less empirical infor- Canada show that 23 percent of all ly, suggesting that the moves of the mation on the topic of return migra- people aged 60 and over changed oldest old are related to health tion of older people to rural places their principal residence at least problems and the need for different of origin. One review of available once during the 5-year period 1986 living arrangements (Hobbs and data for Africa (Becker, 1991) con- to 1991. Thirteen of the fifty states Damon, 1996). Recent Canadian cludes that, while return migration in the United States had net elderly survey data find that older people of older people to their ancestral inmigration rates of more than 10 moved most often because of the homes is not uncommon, several per 1,000 elderly population during size of their home (usually opting factors — growing land pressure, the 1985-90 period, and one out of for a smaller residence), a desire to formalization of rural property five residents of Florida is now aged live in a better neighborhood, or to rights, the increasing viability of 65 or older. A “retirement effect,” build/purchase a home (Che-Alford family support for elders in urban i.e., an increase in mobility rates at and Stevenson, 1998). Among areas — will dampen the likelihood ages 60 to 64, has been noted in older Canadians with daily activity of future return migration. Skeldon the United States and the United limitations in 1991, the percent (1999) notes that, while return Kingdom (Long, 1992). Refugee who moved in the previous 5 years migration may increase the size of movements in Bosnia, Mozambique, was roughly the same as among the older rural cohorts and aggravate and elsewhere have involved vast overall older population (22 per- social support issues, return numbers of older people (Kalache, cent), but about one-third of these migrants also may bring with them wealth, knowledge, and other 1995) who are often overlooked in people relocated to homes with resources, particularly in the form relief operations that focus on chil- special health features. A German of pension income earned during dren and young adults. study of motivating factors for eld- years spent in the urban labor mar- erly mobility in the city of One of the few cross-national stud- ket. Migration from and within Heidelberg (Oswald, Wahl, and ies of elderly migration (Rogers, developing countries in general has Gang, 1997) suggests that basic 1988) identified two basic patterns. come to be seen as a strategic fami- needs (e.g., health) were roughly as One is characterized by amenity- ly decision rather than as an indi- important as “higher-order” needs motivated, long-distance relocation, vidual decision on the part of young such as privacy. the other by intracommunity, leavers (Vatuk, 1995). To the extent assistance-motivated short-distance Information on migration patterns that migration raises family moves. Available studies in devel- of older people in developing coun- incomes and the ability to reunite oped countries suggest that the lat- tries is fragmented at best. members, increased movement of ter are much more common, Although rural-to-urban migration older people may be expected in although the former may become usually is associated with younger the future. U.S. Census Bureau An Aging World: 2001 55
  • 60. CHAPTER 6. Sex Ratios and Marital Status One common characteristic of pop- caregiving that is readily available male mortality rates is that between ulations throughout the world is the in case of illness or disability. age 30 and 40, women usually preponderance of women at older begin to outnumber men. In most ages. Women are the majority of HIGHER MALE MORTALITY countries, the relative female advan- RESULTS IN GENDER the elderly population in the vast tage increases with successively IMBALANCE AT majority of countries, and their OLDER AGES older age (Figure 6-1). share of the population increases with age. This gender imbalance at The primary reason for the numeri- WORLD WAR II IS STILL cal female advantage at older ages EVIDENT IN SEX COMPOSITION older ages has many implications is the sex differential in mortality AT OLDER AGES for population and individual aging, perhaps the most important of discussed in Chapter 3. Although Historical events can play a major which involve marital status and liv- more boys than girls are born, role in shaping the gender composi- ing arrangements. As discussed males typically have higher mortali- tion at older ages. For instance, the further in Chapter 8, family mem- ty rates than females. The sex dif- lingering effects of heavy war mor- bers are the main source of emo- ferential in mortality begins at birth tality during World War II still can be tional and economic support for the and continues throughout the life seen in the proportion female at elderly, although in some developed course. One outcome of higher older ages in certain countries. In countries the state has assumed a larger share of the economic responsibilities. Figure 6-1. Age Percent Female for Older Marital status strongly affects many 65-69 Age Groups: 2000 70-74 aspects of one’s life. Studies in 75-79 developed countries show that mar- 80+ ried people, particularly married men, are healthier and live longer 62 66 than their nonmarried counterparts Russia 75 (Goldman, 1993; Hadju, McKee, and 80 Bojan, 1995; Waite, 1995; Schone 53 and Weinick, 1998). Older married 58 couples tend to be more financially Germany 67 secure than nonmarried people. 74 Changes in marital status at older 56 ages can affect pension potential, 59 Brazil retirement income, and an individ- 61 ual’s social support network; many 67 older widowed men, in particular, 49 may lose contact with much of their 50 India support network after their wife 49 dies (O’Bryant and Hansson, 1996). 48 In contrast, widowed women tend 54 to maintain their support network 56 United States 58 after the death of a spouse (Scott 66 and Wenger, 1995). Marital status also influences one’s living arrange- Source: U.S. Census Bureau, 2000a. ments and affects the nature of U.S. Census Bureau An Aging World: 2001 57
  • 61. Russia, women account for 80 per- cent of the oldest old (aged 80 and Figure 6-2. older) and in Germany they repre- Sex Ratio for Population 65 Years 2000 sent nearly 74 percent. In con- and Over: 2000 and 2030 2030 trast, women account for only two- (Males per 100 females) thirds of the oldest-old population Developed countries in Brazil and the United States. 78 Australia THERE IS GREAT NATIONAL 81 DIVERSITY IN SEX RATIOS 74 Canada AMONG ELDERLY 79 68 A sex ratio is a common measure France 74 used to portray a population’s gen- 70 der composition. A sex ratio is con- Italy 76 ventionally defined as the number 77 of men per 100 women in a given New Zealand 79 population or age category. Sex 46 ratios greater than 100 indicate Russia 58 more men than women, and sex 73 Sweden ratios under 100 indicate the 80 reverse (i.e., more women than 49 Ukraine men). In most countries of the 53 world, sex ratios at older ages are 71 United Kingdom below 100, in some cases quite a 80 bit below (e.g., Russia’s sex ratio is 71 United States 46 men per 100 women aged 65 80 and older). Developed countries Developing countries tend to have lower sex ratios at 71 older ages than do developing Argentina countries (Figure 6-2), although 75 119 there are many exceptions to this Bangladesh 98 generalization. The typical differ- 68 ence between developed and devel- Brazil 70 oping countries is explained by sex 88 differentials in life expectancies at Guatemala 81 birth. As shown in Chapter 3, 103 developed countries tend to have India 91 larger sex differentials in life 60 expectancy at birth than do devel- Kyrgyzstan 63 oping countries, which results in 81 Mexico greater numbers of women than 71 men at older ages. 86 Turkey 86 PROJECTED TREND IN SEX 62 RATIO DIFFERS BY Turkmenistan 65 DEVELOPMENT STATUS 103 Zimbabwe In the future, sex ratios at older 58 ages are projected to move in oppo- Source: U.S. Census Bureau, 2000a. site directions in the aggregate 58 An Aging World: 2001 U.S. Census Bureau
  • 62. developed and developing worlds. Figure 6-3. Between today and 2030, sex ratios Aggregate Sex Ratios for Older Age Groups: for the elderly are expected to 2000 and 2030 increase in many developed coun- 2000 2030 tries because the gender gap in life expectancy is projected to narrow 91 86 (Figure 6-3). In other words, most 81 demographers expect that male life 72 expectancies in developed countries 70 66 are likely to improve at a faster pace than female life expectancies. 57 The opposite is anticipated in many 45 developing countries. In view of the historical pattern in developed nations, sex differentials in develop- ing-country life expectancies are projected to widen, which will in turn lead to lower future sex ratios. Regardless of the projected trends, women are expected to make up 65-79 80+ 65-79 80+ the majority of the world’s elderly Developed countries Developing countries population (particularly at the old- Source: U.S. Census Bureau, 2000a. est ages) well into the next century. Continuing or growing disparities in sex ratios mean that many of the Table 6-1. challenges and problems faced by Sex Ratios for Population Aged 65 and Over for Countries With More Elderly Men Than Women: 2000 the elderly of today and tomorrow are, in essence, challenges and Country Sex ratio problems faced by older women. Qatar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 United Arab Emirates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 Sex ratios at younger ages in some Kuwait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 countries of the world already are Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 quite skewed in favor of women. Bangladesh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Saudi Arabia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 This imbalance may be due to Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 excess male mortality at young Taiwan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Afghanistan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 ages because of wars and other Niger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 forms of violence, disease, or to Oman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 disproportionate out-migration of The Gambia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Eritrea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 young men seeking work in other Yemen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 countries. If mortality is the cause Bahrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Bhutan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 of such severe sex ratios at younger India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 ages, then the implications for the Tunisia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 eventual aging of these cohorts is Source: U.S. Census Bureau, 2000a. different than if the cause is U.S. Census Bureau An Aging World: 2001 59
  • 63. Figure 6-4. Percent Married at Older Ages Male Female DEVELOPED COUNTRIES DEVELOPING COUNTRIES Canada, 1991 Argentina, 1991 83 79 55-64 55-64 71 61 80 76 65-74 65-74 54 42 67 63 75+ 75+ 24 17 Belgium, 1998 Indonesia, 1990 82 90 55-64 55-64 74 55 80 84 65-74 65-74 56 33 65 70 75+ 75+ 23 18 United States, 1995 Chile, 1992 79 74 55-64 55-64 66 57 78 69 65-74 65-74 53 42 68 59 75+ 75+ 25 22 Japan, 1990 South Korea, 1995 92 94 55-64 55-64 77 66 89 88 65-74 65-74 54 34 74 72 75+ 75+ 21 11 Czech Republic, 1991 Zimbabwe, 1992 84 90 55-64 55-64 65 60 80 85 65-74 65-74 42 39 60 78 75+ 75+ 14 21 Source: U.S. Census Bureau, 2000a. 60 An Aging World: 2001 U.S. Census Bureau
  • 64. Figure 6-5. Percent Widowed at Older Ages Male Female DEVELOPED COUNTRIES DEVELOPING COUNTRIES Canada, 1991 Argentina, 1991 3 5 55-64 55-64 14 22 8 11 65-74 65-74 33 44 22 25 75+ 75+ 64 70 Belgium, 1998 Indonesia, 1990 4 6 55-64 55-64 13 39 9 11 65-74 65-74 33 61 27 22 75+ 75+ 67 76 United States, 1995 Chile, 1992 3 5 55-64 55-64 13 19 9 12 65-74 65-74 33 37 22 26 75+ 75+ 65 60 Japan, 1990 South Korea, 1995 3 5 55-64 55-64 14 33 8 11 65-74 65-74 39 65 24 28 75+ 75+ 75 88 Czech Republic, 1991 Zimbabwe, 1992 4 3 55-64 55-64 23 31 11 7 65-74 65-74 48 54 33 14 75+ 75+ 77 73 Source: U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 61
  • 65. migration and the migrants return to their native country when they Figure 6-6. retire (or return for other reasons). Percent Widowed in Belgium: 1998 ELDERLY MEN OUTNUMBER Percent ELDERLY WOMEN IN SOME 100 COUNTRIES While women outnumber men at 80 older ages in most countries, there are 18 countries in Asia and Africa where available data suggest that Female the reverse is true (Table 6-1). One 60 likely explanation involves the social status of women versus men in cer- tain cultures. The relatively low 40 Male numbers of women at older ages may reflect past levels of higher female than male mortality, which 20 could be related to discriminatory treatment accorded girls and women throughout their lifetime. High sex 0 ratios at older ages also may be sta- 20-24 30-34 40-44 50-54 60-64 70-74 80-84 90-94 100+ Age tistical artifacts. That is, women Source: Belgium National Institute of Statistics, 1998. (especially older women) may be undercounted to a greater extent than men in some national census- aged 65 and over were married in 3 percent of men and 31 percent of es, insofar as men are more likely to most study countries. Elderly women aged 55 to 64 were wid- interact with census enumerators women are much more likely to be owed in 1992. At ages 75 and over, and may neglect to provide informa- widowed. In 32 of the study coun- the respective figures were tion on all female household mem- tries, over half of elderly women are 14 percent of men and 73 percent bers. Furthermore, certain concen- widowed. of women. These data are typical of trated patterns of male labor the pattern seen in both developed migration may affect sex ratios to For both men and women, the pro- and developing countries. the extent that a significant portion portion married decreases with of migrants remains in the host older age and the proportion wid- The gender difference in marital sta- country after reaching age 65. owed increases. However, gender tus results from a combination of differences in survival and other fac- factors. The first is the aforemen- OLDER MEN ARE MARRIED; tors (see below) result in very differ- tioned sex difference in longevity; OLDER WOMEN ARE WIDOWED ent average ages of widowhood/ women simply live longer than Older men are more likely to be widowerhood. In the case of men. Secondly, women tend to married and older women are more Belgium in 1998, 82 percent of men marry men older than themselves likely to be widowed in most coun- aged 55 to 64 were married, com- which, combined with the sex dif- tries of the world (Figures 6-4 and pared with 74 percent of women in ference in life expectancy, increases 6-5). In all but six of the 51 study that age group. At ages 75 and the chance that a woman will find countries with data on martial sta- over, 65 percent of men were still herself without a spouse in her tus, over 70 percent of men aged married, compared with only older age. Furthermore, older wid- 65 and older were married. Even at 23 percent of women. The gender owed men have higher remarriage ages 75 and older, a majority of difference in proportions widowed is rates than older widowed women in men were married. In contrast, correspondingly pronounced (Figure many countries, often as a function only 30 to 40 percent of women 6-6). Data for Zimbabwe show that of cultural norms (Velkoff and 62 An Aging World: 2001 U.S. Census Bureau
  • 66. have never married. In some Figure 6-7. European countries, the larger Percent Divorced and Separated in the proportions of elderly women who United States: 1999 have never married may be attribut- Male able to World War II. Many of Female today’s elderly women were of prime marriage age soon after the 21.7 war, when there was a shortage of 20.8 potential spouses due to war 18.6 deaths. Higher-than-average pro- 17.6 18.0 17.8 portions of never-married elderly 15.8 15.1 are found in several Latin American and Caribbean nations, which could be a function of the prevalence of 9.3 consensual unions. While the cate- 8.8 gory “consensual union” is widely used in census tabulations in these countries, some people living (or who have lived) in a consensual union are likely to report them- selves to be never married. 35-39 40-44 45-54 55-64 65+ Age group FUTURE ELDERS MORE LIKELY TO BE DIVORCED Source: U.S. Census Bureau, 2000a. Percentages currently divorced among elderly populations tend to Kinsella, 1993; Cattell, 1997). The decreased. Some of the change is be low. However, this will change fact that women are likely to lose attributable to improved joint sur- in the near future in many countries their spouse has important econom- vival of husbands and wives (Myers, as younger cohorts with higher pro- ic consequences for individuals and 1992). In certain countries, some portions of divorced/separated peo- societies. A comparison of longitu- of the change can be explained by ple move into the ranks of the eld- dinal data from Germany and the the different marital experiences of erly (Gierveld, 2001). For instance, United States revealed that, birth cohorts. For instance, the in 1999 in the United States, around although the level of poverty is dif- diminishing effects of World War II 9 percent of the elderly were ferent in the two counties, most can be seen when the widowhood divorced or separated compared women in both nations experienced rates of older women in Russia are with nearly 15 percent of men and a decline in living standards upon compared for 1979 and 1994. In 19 percent of women aged 55 to widowhood, and many fell into 1979, 72 percent of older women in 64. The proportion divorced/sepa- poverty as a result of the loss of Russia were widowed while only rated is higher still for the age public/private pension support 58 percent were widowed in 1994. group 45 to 54 (Figure 6-7). The (Hungerford, 2001). By 1994, the cohorts that were changing marital composition of the most affected by the war were aged elderly population as these younger ELDERLY MORE LIKELY TO BE 75 and older. cohorts reach age 65 will affect the MARRIED THAN IN THE PAST nature and types of support servic- Over the last two or three decades, SMALL PROPORTIONS OF es that both families and govern- ELDERLY HAVE NEVER MARRIED the marital status of the elderly has ments may need to provide, espe- changed. In a majority of the study Relatively few elderly in most coun- cially with regard to the growing countries, the proportion of older tries have never married. In more number of elderly who lack direct men and women who are married than half of the study countries, familial support (Pezzin and Schone, has increased slightly and the pro- 5 percent or less of elderly men and 1999). portion who are widowed has 10 percent or less of elderly women U.S. Census Bureau An Aging World: 2001 63
  • 67. CHAPTER 7. Living Arrangements Living arrangements are affected by those living in the community, the countries of the world, it is not a host of factors including marital chances of institutionalization. surprising to find that the share of status, financial well-being, health older women living alone is higher Three major observations emerge status, and family size and struc- than that of men. For elderly men in from a cross-national comparison of ture, as well as cultural traditions developed countries, the propor- living arrangements of the elderly. such as kinship patterns, the value tions range from a low of 5 percent First, women in developed coun- placed upon living independently, in Japan to a high of 25 percent in tries are much more likely than men the availability of social services Sweden. Proportions of elderly to live alone as they age; older men and social support, and the physical women residing singly are univer- are likely to live in family settings. features of housing stock and local sally higher, reaching half or more Second, both elderly men and communities. On the individual in Denmark, Germany, and Sweden. women in developing countries level, living arrangements are Percentages in developing countries usually live with adult children. dynamic, representing both a result tend to be much lower, although Third, the use of nonfamily institu- of prior events and an antecedent the levels for men and women in tions for care of the frail elderly to other outcomes (Van Solinge, some countries (e.g., Argentina, varies widely around the world. 1994). On the societal level, pat- Cyprus) rival those of certain terns of living arrangements among MORE THAN HALF OF ELDERLY European nations. Again, older the elderly reflect other characteris- DANISH WOMEN LIVE ALONE women are more likely than older tics — demographic, economic, and men to live alone; St Lucia and Table 7-1 presents data from the cultural — which influence the cur- Taiwan are the only exceptions in 1990s on proportions of older peo- rent composition and robustness of Table 7-1. ple living alone in what are usually older citizens. In turn, living considered to be private (i.e., nonin- The gender gap for those living arrangements affect life satisfaction, stitutional) households. Since alone generally increases with age health, and most importantly for women outlive men in virtually all (Figure 7-1). However, for countries Table 7-1. Percent of Elderly Population Living Alone: Data From 1990 to 1999 Developed countries Male Female Developing countries Male Female Australia . . . . . . . . . . . . . . . . . . . . . 13.7 29.3 Argentina . . . . . . . . . . . . . . . . . . . . 11.2 21.1 Canada . . . . . . . . . . . . . . . . . . . . . . 14.1 33.7 Aruba . . . . . . . . . . . . . . . . . . . . . . . 12.5 15.4 Czech Republic . . . . . . . . . . . . . . . 19.0 47.5 Bolivia. . . . . . . . . . . . . . . . . . . . . . . 13.2 15.7 Denmark . . . . . . . . . . . . . . . . . . . . . 23.3 52.0 Cyprus . . . . . . . . . . . . . . . . . . . . . . 10.6 24.8 Finland. . . . . . . . . . . . . . . . . . . . . . . 19.5 46.5 Hong Kong . . . . . . . . . . . . . . . . . . 11.6 13.2 France . . . . . . . . . . . . . . . . . . . . . . . 15.3 40.2 Mexico . . . . . . . . . . . . . . . . . . . . . . 7.5 14.0 Germany . . . . . . . . . . . . . . . . . . . . . 16.9 50.8 Morocco (60+) . . . . . . . . . . . . . . . 11.3 44.7 Greece. . . . . . . . . . . . . . . . . . . . . . . 8.7 22.8 Philippines(60+) . . . . . . . . . . . . . . 4.4 6.4 Ireland . . . . . . . . . . . . . . . . . . . . . . . 18.9 27.7 South Korea(60+) . . . . . . . . . . . . 3.1 11.8 Japan . . . . . . . . . . . . . . . . . . . . . . . . 5.2 14.8 St. Lucia. . . . . . . . . . . . . . . . . . . . . 20.9 18.9 New Zealand . . . . . . . . . . . . . . . . . 17.8 38.0 Taiwan . . . . . . . . . . . . . . . . . . . . . . 13.0 7.4 Norway . . . . . . . . . . . . . . . . . . . . . . 21.3 44.7 Thailand(60+) . . . . . . . . . . . . . . . . 2.9 5.5 Portugal . . . . . . . . . . . . . . . . . . . . . . 9.4 23.9 Vietnam(60+) . . . . . . . . . . . . . . . . 2.5 8.1 Romania . . . . . . . . . . . . . . . . . . . . . 12.4 31.7 Sweden . . . . . . . . . . . . . . . . . . . . . . 25.1 49.9 United States . . . . . . . . . . . . . . . . . 15.1 36.8 Note: Data for Mexico are for seven cities, and refer to 1989. Data are for household populations aged 65 and over, unless otherwise noted. Data for Morocco refer to urban areas only. Sources: Compiled from data in United Nations Demographic Yearbooks (various dates) and national sources. U.S. Census Bureau An Aging World: 2001 65
  • 68. that disaggregate data at advanced ages, the gender difference tends to Figure 7-1. diminish at very old ages, presum- Percent of Elderly Living Alone in ably as a result of health and/or Germany and the United States Male economic factors that require insti- by Available Age Groups Female tutional caretaking, communal liv- Germany, 1999 United States, 1995 ing, or sharing of housing costs. Both numbers and proportions of 66 elderly living alone have risen 61 sharply in developed countries since the early 1960s, although 51 recent information suggests that 45 the rise in proportions might be leveling off in some nations. 36 Everywhere, however, the absolute 32 32 numbers are increasing. Figure 7-2 24 illustrates a trend common to most 20 developed countries, i.e., the 14 14 13 increase has been largely fueled by women. Data from the 1996 cen- sus of Canada show more than 700,000 elderly women living 65-69 70-74 75+ 65-74 75-84 85+ alone, a jump of more than 180,000 Age group since 1986. The number of elderly Source: National sources. women living alone grew at an average annual rate of 5.4 percent from 1961 to 1996, compared with Figure 7-2. a rate of 1.4 percent for the entire Elderly Living Alone in Canada: 1961 to 1996 Canadian population. One implica- tion of such change is that, in most Thousands developed countries, women must 800 anticipate a period of living alone at some point during their older years. “ELDERLY-ONLY” HOUSEHOLDS 600 ARE INCREASINGLY COMMON An earlier version of this report Female (Kinsella and Taeuber, 1993) point- 400 ed out that elderly people living alone were a significant factor in national household profiles in Europe. In several nations (e.g., 200 Male Belgium, Denmark, France, and the United Kingdom) in the 1980s, more than 11 percent of all national households consisted of a solitary 0 individual aged 65 or over. The 1961 1966 1971 1976 1981 1986 1991 1996 most common living arrangement Source: National sources. for the elderly in Europe was with 66 An Aging World: 2001 U.S. Census Bureau
  • 69. Table 7-2. Composition of European Households With Elderly Members: Early 1990s Percent elderly in households with: One person 65 Another person Another person Country years or over 65 years or under 65 years Two Three or (1 person over (2 person (2 person other more other household) household) household) persons persons Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.4 40.6 11.4 4.1 1.5 Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.7 29.4 14.2 15.0 23.7 Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.6 31.4 11.3 17.4 23.3 France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.1 39.9 11.9 9.7 6.5 Ireland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.2 25.3 13.1 16.4 18.9 Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.9 31.3 12.7 14.7 15.4 Netherlands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35.5 45.4 11.5 5.5 2.2 Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35.3 29.4 12.9 10.2 12.2 Portugal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.5 33.8 11.9 14.4 21.3 Finland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38.4 32.2 12.8 8.9 7.7 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.1 44.4 10.0 3.6 0.9 United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35.7 40.5 11.9 7.9 3.9 Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34.0 41.2 12.6 8.1 4.1 Source: Eurostat, 1996. another elderly person in a two- reimbursement payments which dis- national differences in elderly living person household. A related calcu- courage institutional living. arrangements in Europe are charac- lation revealed that three out of terized more by diversity than by every ten households in a 12-nation MAJORITY OF ELDERLY IN similarity. DEVELOPED COUNTRIES LIVE European aggregate contained at WITH OTHER PEOPLE TWO- OR THREE-GENERATION least one elderly person. Although high proportions of elder- HOUSEHOLD STILL THE Data from the 1991 census of Great DEVELOPING-COUNTRY NORM ly often live alone in developed Britain (for England, Wales, and countries, a majority of those aged In all developing regions of the Scotland) are even more striking; 65 and over still live with other world, with the possible exception 15 percent of all households in Great people. Data from 13 European of the Caribbean, the most common Britain consisted of a single pension- nations (Table 7-2) show that the living arrangement for elderly peo- er living alone, while another 10 per- proportion of elderly living with one ple (married or widowed) is with cent of households had two or more other elderly person only (in most children and/or grandchildren. pensioners with no other people cases a spouse) tends to be higher Between 72 and 79 percent of older present. Hence, 25 percent of all than the proportion living singly. (60 and over) respondents in 1984 households in the country consisted Between 10 and 14 percent of the World Health Organization surveys of pensioners only. Overall, one- elderly in the 13 nations live with in Malaysia, the Philippines, Fiji, and third of Britain’s households had at one other person who is less than South Korea lived with children least one resident pensioner. 65 years of age; many of these eld- (Andrews et al., 1986), and similar In developed countries, the growth erly are likely to be either men liv- results have been observed in coun- in “elderly-only” households may be ing with younger spouses, or wid- tries as diverse as India, Indonesia, due in part to changes in social and owed or divorced individuals living Cote d’Ivoire, Singapore, and (at economic policies. These include: with a child. Nations vary greatly in earlier times) six Latin American increases in benefits that allow older the proportion of elderly people liv- nations (Kinsella, 1990). More people to live independently of their ing in households of three or more recent data from four Asian nations children; programs that more easily people, from only 5 to 6 percent in (Figure 7-3) show a persisting pat- permit the conversion of housing Sweden and Denmark to 35 percent tern. While the levels may appear wealth into income; programs that or more in Ireland, Greece, Portugal, similar, the broad category of “living encourage the building of elder- and Spain. As earlier studies (Wall, with offspring” encompasses a friendly housing; and revisions in 1989; Pampel, 1992) have noted, plethora of specific family and U.S. Census Bureau An Aging World: 2001 67
  • 70. household types, differing not only among nations but among ethnic Figure 7-3. groups within nations. Such diver- Percent of People Aged 60 and Over Living sity points to the importance of cul- With Children in Four Asian Countries: tural and ideological as well as Circa 1996 Male demographic factors in the determi- Female nation of living arrangements of 90.6 older people (Albert and Cattell, 86.5 1994). 76.5 71.9 72.7 71.8 A growing concern in developing 67.4 69.4 countries is the extent to which the twin processes of modernization and urbanization will change tradi- tional family structures (Zhou, 2000). Data for most of the devel- oping world generally are insuffi- cient for documenting changes in living arrangements of the elderly. Although the case of Japan may not seem especially relevant to develop- ing nations, the extended family Philippines Singapore Thailand Vietnam structure common to the latter has (Red River Delta) historically been a feature of Japanese society as well. Time Sources: Anh et al., 1997; Chan, 1997; Knodel and Chayovan, 1997; and Natividad and Cruz, 1997. series data (Figure 7-4) show that the number and proportion of extended-family households in Figure 7-4. Japan have been declining, and that Living Arrangements of Japanese Elderly: 1960 to 1995 the proportions of elderly living alone or with spouse only have With kin Elderly Alone Institution been increasing. couple only These trends in Japan have led to 1960 the suggestion (Kamo, 1988) that the impact of industrialization has 1975 undermined the indigenous culture of Japan vis-a-vis the status of its 1985 elderly citizens, and set the stage for the eventual predominance of 1995 the nuclear family. Related to this is the notion of “intimacy at a dis- 0 20 40 60 80 100 Percent tance” (see, e.g., Stehouwer, 1968; Note: "With kin" includes very small numbers of elderly cohabitating with nonkin. Rowland, 1991). That is, as the Source: Japan Management and Coordination Agency, cited in Atoh, 1998. financial (and to some extent health) status of elderly people improves, a larger proportion are literature relates improvements in normative changes toward individ- able to afford to live alone and social security programs and gen- ualism and personal independence choose to do so in independent eral economic welfare to the ability (see, e.g., the discussion and refer- dwellings, while at the same time and desire of older persons to ences in Gierveld, 2001). This con- maintaining close familial contact choose independent living arrange- cept has found general currency in and exchange supports. A growing ments, presumably reflecting a variety of cultural settings, 68 An Aging World: 2001 U.S. Census Bureau
  • 71. internationally consistent data, and Figure 7-5. differences between countries Percent of Elderly in Residential Care: should be construed as orders of Early to Mid-1990s magnitude rather then as precise measurements. One attempt to col- Netherlands 8.8 late reasonably-comparative data on Sweden 8.7 residential care in the 1990s (OECD, Denmark 7.0 1996) suggests that usage rates for Australia 6.8 developed countries (Figure 7-5) Canada 6.8 range from 2 percent in Portugal to Germany 6.8 Luxembourg 6.8 9 percent in the Netherlands.1 New Zealand 6.7 There is substantial variation in the Norway 6.6 use of custodial institutions and in France 6.5 the mix of long-term care alterna- Belgium 6.4 tives and services (see, e.g., Ribbe Finland 6.4 et al., 1997; Mechanic and Japan 6.0 McAlpine, 2000). One study (Doty, United States 5.7 1992) suggests that Japan, United Kingdom 5.1 Australia, and North America have Ireland 5.0 made greater relative use of med- Austria 4.9 ical residential care, while an Italy 3.9 Spain 2.9 emphasis on nonmedical facilities Portugal 2.0 has been more apparent in Belgium, Turkey 0.2 Sweden, and Switzerland. Notes: Canada and Finland: figures represent the midpoint of an estimated range. In Eastern Europe and parts of the Japan and the Netherlands: some of the residential care is provided in hospitals. former Soviet Union, the combina- Sources: OECD, 1996; Jacobzone, 1999. tion of low fertility and the rapid increase in oldest-old population although a recent analysis of data lowered fertility, increased geo- might be expected to translate into from the Indonesian Family Life graphical mobility, and the rapid a growing use of institutional health Survey (Cameron, 2000) finds little advance in medical technology — care and maintenance services. To evidence that increases in the has led to a steep rise in numbers date, however, available information income of the elderly, or that of their of institutionalized elderly. The indicates that institutionalization of children, lead to a significant change highest rates of institutional use are older people is not common. In in traditional family structure. found in many of the world’s oldest Russia, for instance, the capacity of countries, and the absolute num- (number of existing places in) old NUMBERS OF bers of users are expanding even in people’s homes and nursing homes INSTITUTIONALIZED ELDERLY the face of increasing efforts to throughout the former USSR in the RISING enhance community-based services late 1980s was estimated to be A number of studies (e.g., Manton, and avoid or greatly reduce levels Stallard, and Liu, 1993; Weiner and of institutionalization. Illston, 1995; Leung, 2000) have 1 Although the percentage of elderly in insti- In spite of the intense media scruti- tutions at any given moment may be relatively documented the direct relationship low, on average around 5 percent in developed between population age-sex struc- ny of and controversy surrounding countries, an estimated 25 to 30 percent of institutional residence, the fact people who survive to age 65 can expect to ture, age-sex-specific rates of chron- spend some time in an institutional setting ic disease and disability, and the remains that relatively small propor- before they die (Sundstrom, 1994). Thus the tions of elderly populations reside longitudinal risk of experiencing institutional- need for long-term care. The con- ization is much higher than cross-sectional fluence of several macro trends in in institutions at any given time. rates might suggest. Considerable research interest currently is devoted to untangling the developed countries — older popu- Cross-national comparisons of insti- dynamics of institutional use, including transi- lation age structures, higher inci- tutionalized populations are prob- tions to and from such facilities and the under- lying health conditions that drive the transi- dence of noncommunicable disease, lematic due to the absence of tions. U.S. Census Bureau An Aging World: 2001 69
  • 72. between 320,000 (Muzafarov and Kurleutov 1994) and 380,000 Figure 7-6. (Bezrukov, 1993); the higher esti- Percent of Elderly in Institutions in mate represents less than 1.5 per- Austria and New Zealand: 1991 Austria cent of the USSR population aged New Zealand 65 and over as of 1988. While the average Russian view of institution- Male Female 37.7 alization may be extremely negative (Powell, 1991), there does appear to be an unmet need for institutional services. Lengthy waiting lists for institutional admission have been the norm for many years, and offi- 22.2 20.8 cial time series data for Russia show a steady rise in the number of nursing/old people’s homes, from 14.9 around 700 in 1985 to more than 10.7 900 in 1996. At the same time, the 8.9 9.5 number of places in such institu- 5.7 5.9 5.1 tions has remained fairly constant, 4.2 3.1 2.0 2.4 2.3 2.4 suggesting a downsizing of the 1.0 1.6 0.9 1.3 average facility. Older people living 65-69 70-74 75-79 80-84 85+ 65-69 70-74 75-79 80-84 85+ alone, and especially never-married Age group elderly men, are said to be at partic- Source: OECD, 1996. ularly high risk of institutionaliza- tion. In rural areas of the country, new policies aimed at alleviating octogenarians. More than half of district hospitals frequently serve as current and anticipated problems. all Norwegians aged 85 and over long-term residences for the elderly, Long-term care provision and/or reside in institutions at a given for social as well as health reasons homes for the aged have become point in time, as do one-third or (Bezrukov, 1993). increasingly accepted and common more of this age group in Australia Rates of institutionalization usually in countries — especially in and New Zealand. In fact, a major- are very low or negligible in the Southeast Asia — where sustained ity of people entering institutions developing world. In official rheto- fertility declines have led to rapid or other types of collective ric, at least, the Western model of population aging and reduced the dwellings have reached very institutional care for older people numbers of potential family care- advanced age. Those who enter at often is rejected as culturally inap- givers (Phillips, 2000; Bartlett and less-advanced ages tend to be propriate (Gibson, 1992). Outside Wu, 2000). either single or widowed and child- of Europe and North America, social less, i.e., people who are unlikely traditions and official decrees of fil- ELDERLY WOMEN to have young family members to PREDOMINATE IN ial and familial responsibility have rely upon for support (Soldo, 1987). INSTITUTIONAL POPULATIONS obviated, at least until recently, Institutional use is strongly associ- Women and the oldest old, therefore, debate about living arrangements ated with increasing age regardless are disproportionately represented of the elderly. Lately, however, a of national setting (Figure 7-6). In among the institutionalized elderly number of countries have recog- most developed countries, fewer (Figure 7-7). In the United States in nized that even if the family retains than 2 percent of the young old 1997, three-fourths of all nursing much of its support function for the (aged 65 to 69) are in institutions. home residents were women, and elderly, demographic and socioeco- This level rises fairly slowly until slightly more than half of all nursing nomic changes will inevitably pro- age 80, but many nations experi- home residents were aged 85 or duce strains. Consequently, many ence a sharp increase in institu- older. People in institutions at one developing nations have adopted tionalization rates among point in time, however, do not 70 An Aging World: 2001 U.S. Census Bureau
  • 73. discharge a high proportion of users Figure 7-7. back into the community, while oth- Percent of U.S. Elderly in Nursing Homes by Age: ers systems have relatively limited 1973-74, 1985, 1995, and 1997 rehabilitative services (Ribbe et al., 1997). Male Percent NATURE OF INSTITUTIONAL 30 USE HAS CHANGED Policies toward and practices of 25 institutionalization of older people in developed countries have changed over the past half-century. 20 In the United States and elsewhere, 85+ institutionalization in the early 15 1900s was generally associated with poverty and/or inability to work. The elderly often were 10 housed with younger “welfare popu- lations” and were supported largely 5 75-84 by local agencies. By the middle of the twentieth century, hospital- 65-74 based care for the elderly had 0 become more common, at least in 1973-74 1985 1995 1997 the United States, with financial and operational control likely to come Female from state governments. Beginning Percent 30 in the 1950s, social policy encour- 85+ aged a shift away from hospital use toward nursing-home use. The 25 United States experienced a rapid expansion of nursing home capacity (OECD, 1996), with emphasis on 20 providing for chronic disease and physical disability needs. Federal 15 funding assumed greater promi- nence, as did private sources of funding. More recently, the private 10 long-term care insurance industry 75-84 has grown rapidly, while new forms 5 of home and community-based 65-74 services (e.g., assisted living) have emerged. With U.S. long-term care 0 costs doubling each decade since 1973-74 1985 1995 1997 1970 (reaching an annual level of Source: U.S. Centers for Disease Control, 1999. $106 billion in 1995; Stallard, 1998), the mix of institutional and home-based care has been shifting necessarily remain there and “age in multiple transitions. Some national rapidly toward the latter, especially place” indefinitely. Many older indi- nursing home systems (e.g., the for the oldest old (Cutler and Meara, viduals who enter an institution Netherlands) have well-developed 1999). eventually leave, and many make rehabilitative programs which U.S. Census Bureau An Aging World: 2001 71
  • 74. Countries in Europe also have been attributed largely to differences in mere fact of elderly coresidence active in changing long-term care the organization and financing of with a younger generation(s) tells policies and practices in response long-term care as well as differing us little about the quality of intra- to population aging. Heightened sectoral responsibilities for care. household relationships and life sat- spending on institutional care Nations clearly are struggling with isfaction. prompted Great Britain to revamp alternative methods of long-term Instead of focusing on living legislation in the 1990s, transfer- care financing and provision, and it arrangements per se, attention ring more fiscal control to local gov- is hoped that countries can learn might better be directed to under- ernments and tightening means- from one another via cross-national standing the complex set of mecha- tested provisions. Austria instituted research that proceeds from the nisms and interpersonal relation- a new federal act in 1993 aimed types of efforts now underway in ships that determine the timing and primarily at increasing options relat- Europe. content of support for older per- ed to personal care arrangements BEYOND LIVING sons. This perspective is summed and supporting individuals in their ARRANGEMENTS up well by the phrase “function own homes for as long as possible. rather than form,” meaning that the Germany, in 1995, unveiled a sys- Living arrangements of older per- mechanisms and characteristics of tem of universal long-term care sons clearly are an important com- an individual’s support network are insurance which features expanded ponent of life, but we should be much more salient to well-being benefits without major changes in careful not to infer too much from (and to policymaking) than are mere means-testing. Importantly, the effi- cross-sectional descriptive data on attributes of who lives with whom cacy of such changes will be moni- residence patterns. We need to be (Hermalin, 1999). Survey research tored and evaluated by research aware of how living arrangements and methodology increasingly are projects underway in each country change as a function of the growth focused on the full mapping of (Wolf, 1998). in elderly populations and their complex kin networks, household shifting health and kin-availability As noted above, developed coun- and kin microsimulation techniques, profiles (Palloni, 2000). And as tries vary enormously in their use and new data-record linkages that alluded to earlier, the well-being of and view of institutional residency allow analysts and policymakers to individuals is not necessarily reflect- for older citizens. A study (Ribbe et better understand the underlying ed in living arrangements. Living al., 1997) of nursing homes in ten dynamics of intergenerational trans- alone in old age has sometimes nations found no apparent relation fers and well-being in old age been interpreted as a lack of famil- between the level of population (Hagestad, 2000; Wolf, 2000). ial and social integration, when in aging and the number of nursing Chapters 8 and 11 look further at fact it may be indicative of good home beds. The surprising lack of family and social support for older health, economic well-being, and cross-national consistency is persons. social connectedness. Likewise, the 72 An Aging World: 2001 U.S. Census Bureau
  • 75. CHAPTER 8. Family and Social Support of Older People Shifts in population age structure and not all working-age people because the relatively large post- generally result in new service actually work or provide direct sup- World War II birth cohorts will still demands and economic require- port to elderly family members. be of working age through at least ments. With an increasingly older The statistics discussed in the first 2010. In several nations (notably age structure comes change in the part of this chapter may be seen as the United Kingdom, the United relative numbers of people who can rough guides to when we can States, and Russia), the elderly sup- provide support to those who need expect the particular age distribu- port ratio will not change signifi- it. In the early 1980s, Myers and tion of a country to affect the need cantly from 2000 to 2010. Some Nathanson (1982) identified three for distinct types of social services, developed nations, however, are prominent issues regarding popula- housing, and consumer products. aging at a much faster pace. tion and the family: 1) the extent to These data suggest some of the fac- Between 2000 and 2015, the elderly which changes in social norms and tors that will shape patterns of support ratio in Denmark is likely to responsibilities, driven by the secu- social relationships and societal increase 33 percent (from 24 to 32), lar processes of urbanization and expenditures in the coming and the increase in the Czech modernization, alter traditional decades, but tell us little about the Republic will likely be 36 percent familial modes of caring for older changing nature of the health and (22 to 30). Most notably, Japan’s people; 2) the possible social sup- economic resources of the aged in elderly support ratio is expected to port burden resulting from reduced the future. jump 63 percent (from 27 to 44) economic self-sufficiency of aged during the 15-year period. people and the likelihood of height- RAPID RISE IN ELDERLY SUPPORT RATIOS EXPECTED IN From 2015 to 2030, the elderly sup- ened chronic disease morbidity and DEVELOPED COUNTRIES port ratio will increase by more than functional impairment related to AFTER 2010 40 percent in several developed longer life expectancy; and 3) the Broad changes in a nation’s age nations as the large working-age ways in which countries develop structure are reflected in changing cohorts begin to retire. In 2030, funding priorities for public care societal support ratios. These ratios Japan’s elderly support ratio is pro- systems given competing demands typically indicate the number of jected to be 52 (Figure 8-1). Italy is for scarce resources. youth and/or elderly people per 100 likely to have an elderly support To gain a broad view of these people aged 20 to 64 years, primary ratio of 49 in 2030, and nearly all dynamics, demographic assess- ages for participation in the labor European countries will have elderly ments of intergenerational support force. A commonly used measure support ratios over 40. New often consider various ratios of one of potential social support needs is Zealand has the lowest projected age group to another. This chapter the elderly support ratio (sometimes ratio (30) among the developed considers societal support ratios, called the elderly dependency ratio), nations in this study, with other rela- parent support ratios, and changes defined here as the number of peo- tively low figures seen in the United in kin availability. As seen through- ple aged 65 and over per 100 peo- States and Eastern Europe. out this report, the elderly popula- ple aged 20 to 64 in a given popula- tion is diverse in terms of its ELDERLY SUPPORT RATIOS IN tion. In the coming decades, elderly MOST DEVELOPING COUNTRIES resources, needs, and abilities. The support ratios will rise in developed TO CHANGE SLOWLY stereotype of the elderly as a pre- countries as a result of both declin- dominately dependent group that ing fertility and increasing longevity. Elderly support ratios are much drains a nation’s economy has erod- The rise has been and will continue lower in developing than in devel- ed. Not all elderly require support to be modest in most countries oped countries, often with ten or U.S. Census Bureau An Aging World: 2001 73
  • 76. fewer elderly people per 100 people aged 20 to 64. Among the 30 Figure 8-1. developing countries in this study, Elderly Support Ratios: 1950 to 2030 Uruguay had the highest level (24) in 2000, followed by Argentina (19) and Israel (18). Many developing Developed countries Number of people aged 65 and over per 100 people aged 20 to 64 countries will experience little if any 60 change in their elderly support ratios from 2000 to 2015, because the high-fertility cohorts of the 50 Japan 1960s and 1970s will still be under age 65 in 2015. Thailand and 40 South Korea stand out as excep- tions as they are expected to expe- rience relatively large increases in 30 the ratio between 2000 and 2015. Belgium In Bangladesh, Kenya, Malawi, Morocco, and Uruguay, on the other 20 United States hand, the elderly support ratio is expected to remain stable between 10 2000 and 2015. In Jamaica and Pakistan the elderly support ratio is projected to decline by 2015, even 0 1950 1965 1980 1995 2010 2025 though the absolute numbers of elderly population are increasing. Developing countries In countries where fertility remains Number of people aged 65 and over per 100 people aged 20 to 64 high or has just recently begun to 60 decline significantly — as in much of Africa and South Asia — elderly 50 support ratios should change little during the entire period 2000 to 2030. Eastern and Southeastern 40 Asia and parts of Latin America, on South Korea the other hand, could witness signif- 30 icant change during that time. The elderly support ratio is projected to at least double between 2000 and 20 Argentina 2030 in 11 Asian and Latin Egypt American study countries, and to triple in South Korea. 10 YOUTH SUPPORT RATIOS TO 0 DECLINE 1950 1965 1980 1995 2010 2025 The working-age population also Sources: United Nations, 1999 and U.S. Census Bureau, 2000a. provides support for young people. 74 An Aging World: 2001 U.S. Census Bureau
  • 77. bulk of support to both young and Figure 8-2. old alike. Youth and Elderly Components of the Total Youth From 2000 to 2015, the total sup- Support Ratio: 2000, 2015, and 2030 Elderly port ratio (TSR) should remain rela- (Youth ratio: people aged 0 to 19 per 100 people aged 20 to 64; Elderly ratio: people aged 65 and over per 100 people aged 20 to 64) tively stable in most developed countries as declining numbers of children more than offset growing Australia, 2000 46 21 numbers of elderly (Figure 8-2). 2015 40 26 From 2015 to 2030, however, 2030 40 37 increasing numbers of elderly peo- ple will boost the TSR in all devel- France, 2000 43 27 oped nations even though the 2015 39 32 youth component may decline 2030 38 44 slightly. Among the study countries, the proportional gain in the TSR Uruguay, 2000 59 24 from 2015 to 2030 is projected to 2015 55 24 be greatest in Russia (30 percent). 2030 50 28 The United States is projected to have the highest TSR (87) among Philippines, 2000 98 7 the developed countries in the year 2015 76 9 2030, with 7 other developed coun- 2030 60 13 tries also projected to have TSRs in excess of 80. Source: U.S. Census Bureau, 2000a. For the foreseeable future in devel- oping countries, major fertility Children outnumber working-age DIVERGENT TOTAL SUPPORT reductions are likely to outweigh adults in many developing RATIO PATTERNS IN growing numbers of elderly people. countries. As a result, youth sup- DEVELOPED VERSUS At the same time, the working-age DEVELOPING COUNTRIES population is increasing. Hence, port ratios — defined here as peo- ple under age 20 per 100 adults The total support ratio (youth plus future TSRs for the vast majority of aged 20 to 64 years — for 2000 elderly in relation to the working- developing countries are projected were in excess of 100 in several age population) provides a gross to be lower than in 2000. Even developing countries (mainly in indication of the overall support though growth rates for the youth Africa). In the developed countries burden on working-age adults. The and elderly will be higher than in in this study, however, youth sup- level of the total support ratio over developed countries during the port ratios ranged from only 31 in time is pertinent to policymakers, next three decades, TSRs in devel- Italy to 49 in New Zealand. but knowing the balance of old ver- oping countries often will be lower sus young may be more important because of the massive numbers of In most countries of the world, working-age adults in their popula- because supporting the young is youth support ratios are projected tions. Such change may portend a probably less costly than support- to decline between 2000 and 2030. window of economic opportunity ing the elderly (especially as the In countries where the present level for developing countries. As the elderly population itself ages). With is high, the youth support ratio may ratio of working-age to total popula- the major exception of education, decline by half or more. In Kenya, tion rises, economies have relatively the costs of young people are borne for example, the 2000 level of 133 more productive units and therefore by families more than by govern- is projected to plummet to 61 in more opportunity to grow (other ment programs, although some 2030. factors being equal). European governments provide the U.S. Census Bureau An Aging World: 2001 75
  • 78. THE USEFULNESS OF ELDERLY SUPPORT RATIOS Figure 8-3. Implicit in the standard definition of Standard and Alternative Elderly Support Ratios: 1998 an elderly support ratio is the notion that all people over age 64 24 Total population 65 and over per 100 are in some sense dependent on the 31 Finland total population 20-64 population in the working ages (20 31 31 Total population 65 and over to 64) who provide indirect support per 100 economically active to the elderly through taxes and 26 population 20-64 contributions to social welfare pro- 33 Japan Noneconomically active grams. We know, of course, that 25 population 65 and over per 23 100 economically active elderly populations are extremely population 20-64 diverse in terms of resources, 27 39 Noneconomically active needs, and abilities, and that many Spain population 65 and over per 39 elderly are not dependent in either 100 economically active 39 population 20 and over a financial or a physical (health) 22 sense. Older people pay taxes, United 27 often have income and wealth that States 24 fuel economic growth, and provide 23 support to younger generations. 25 Likewise, substantial portions of the 34 Germany working-age population may not be 33 financial earners, for reasons of 32 unemployment, inability to work, Sources: International Labour Office Yearbook of Labour Statistics, 1999 and U.S. Census Bureau, 2000a. pursuit of education, choosing to be out of the labor force, and so forth. While it is empirically difficult to they are not economically depend- factors as (1) workers under age 20; include factors such as intrafamily ent. The fourth bar builds on the (2) trends in unemployment; (3) financial assistance and child care third bar by adding these economi- average retirement ages; and (4) activities into an aggregate measure cally active elderly to the ratio levels of pension receipt and institu- of social support, it is feasible to denominator of other economically tionalization among the elderly, take account of employment charac- active individuals, on the assump- and/or the prevalence of high-cost teristics in both the working-age tion that these working elderly con- disabilities. and elderly populations. In Figure tinue to contribute tax revenue to 8-3, the topmost bar for each coun- national coffers. RAPIDLY CHANGING AGE try represents the standard elderly STRUCTURE IS A CHALLENGE support ratio as defined above. The The alternative ratios in each coun- TO SUPPORT IN SOME try are higher than the standard eld- DEVELOPING COUNTRIES second bar includes only the eco- nomically active population aged 20 erly support ratio, except in Japan One of the more dramatic demo- to 64 in the denominator, thereby where the elderly have a relatively graphic developments in the last excluding people who choose not to high rate of labor force participation two decades has been the pace of work, unpaid household workers, (often as part-time workers). To the fertility decline in many developing nonworking students, and perhaps extent that policy and program countries. The common perception those individuals whose health sta- agencies use support ratio calcula- is that below-replacement fertility tus keeps them out of the labor tions, the effect of including versus levels are seen only in the industri- force. The third bar represents a excluding labor force participation alized nations of the Northern calculation similar to the second bar, rates appears considerable in most Hemisphere. As of 2000, however, but removes economically active countries. Data permitting, other the total fertility rate was below people aged 65 and over from the adjustments might be made to replacement level in 21 developing numerator on the assumption that these ratios to account for such countries, mostly in Latin America 76 An Aging World: 2001 U.S. Census Bureau
  • 79. Figure 8-4. Population by Single Years of Age for China: 2000 Age Male Female 85+ 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 15 12 9 6 3 0 3 6 9 12 15 Millions Source: U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 77
  • 80. and the Caribbean and parts of Asia (U.S. Census Bureau, 2000a), and is Figure 8-5. declining steeply in many other Youth and Elderly Support Ratios in developing countries. China: 1985 to 2050 The situation in the People’s Youth ratio: people aged 0 to 19 per 100 people aged 20 to 64; Elderly ratio: people aged 65 and over per 100 people aged 20 to 64 Republic of China illustrates the 100 potential effect that rapidly- declining fertility may have vis-a-vis population aging. In 1979, China 80 established an official one-child-per- family policy aimed at curbing Youth growth in the world’s most popu- 60 lous nation. While the policy was relaxed somewhat in subsequent years, China’s total fertility rate 40 declined to an estimated level of 1.8 children per woman in 2000. As a result, China will age sooner 20 and more quickly than most devel- Elderly oping countries. China’s age profile in 2000 con- 0 tained a large “bulge” consisting of 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 people aged 26 to 37 (Figure 8-4). Source: U.S. Census Bureau, 2000a. The oldest people in this age bulge will be entering their sixties just prior to the year 2025. This popu- Figure 8-6. lation momentum will produce a Parent Support Ratios in Four World rapid aging of the Chinese popula- 1950 Regions: 1950, 2000, and 2030 2000 tion in the third and fourth decades (People aged 80 and over per 100 people aged 50 to 64) 2030 of the twenty-first century. Recent analyses of 1995 sample census data from China suggest higher old- age mortality than had been previ- 37 ously estimated, resulting in lower numbers of projected elderly peo- ple. Nevertheless, the number of Chinese aged 65 years and over is 20 17 now projected to increase from 15 88 million in 2000 to 197 million in 10 9 7 7 2025, and to 341 million in 2050. 4 4 3 3 Short of a catastrophic rise in adult mortality or massive emigration of South America Eastern Asia Western Europe Western Africa an unprecedented scale, we can be reasonably certain that this growth Note: South America includes: Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, French Guiana, Guyana, Paraguay, Peru, Suriname, Uruguay, and Venezuela. will occur, because the elderly of Eastern Asia includes: China, North Korea, South Korea, Hong Kong, Japan, Macau, Mongolia, and Taiwan Western Europe includes: Austria, Belgium, France, Germany, Liechtenstein, Luxembourg, Monaco, the middle decades of the twenty- Netherlands, and Switzerland. Western Africa includes: Benin, Burkina Faso, Cape Verde, Cote d' Ivoire, Gambia, Ghana, Guinea, first century are already born. Guinea-Bissau, Liberia, Mali, Mauritania, Niger, Nigeria, St. Helena, Senegal, Sierra Leone, and Togo. Eventually, China’s projected youth Sources: United Nations, 1997 and U.S. Census Bureau, 2000a. and elderly support ratios are likely 78 An Aging World: 2001 U.S. Census Bureau
  • 81. to converge (Figure 8-5), and we SANDWICH GENERATION A most of the oldest old were aged may anticipate a social and eco- DEVELOPED-COUNTRY 20 to 35. Of course, people in the nomic fabric radically different from PHENOMENON numerator (80 and over) are not that of today. One aspect of the changing age necessarily in the same families as structure of families that has those in the denominator (50 to 64 MORE PEOPLE WILL FACE received recent attention is the so- years). Thus, the PSR is only a CARING FOR FRAIL RELATIVES called “sandwich generation,” that rough indication of need for family In the eighteenth and nineteenth is, people who find themselves car- support over time. centuries, low levels of life ing for elderly parents while still Relatively few people aged 50 to 64 expectancy meant that people on caring for/supporting their own in 1950 worried about caring for average lived a relatively short children or grandchildren and often people aged 80 or older. In the amount of time in a multigenera- participating in the labor force. In developed countries in this study, tional family (United Nations, developed countries especially, the PSR ranged from five in Japan 1990b). While most older individu- more people will face the concern and Hungary to eleven in Norway in als lived with family members, and expense of caring for their very 1950. In developing countries, the years spent in an extended-family old, frail relatives with multiple, PSR ranged from two in Bangladesh arrangement were limited because chronic disabilities and illnesses. to eleven in Tunisia and Uruguay. the average person died shortly The need for help is likely to come Increases in the PSR since 1950 after becoming a grandparent at the very time when the adult chil- imply that a relatively larger share (Hareven, 1996). Declining mortali- dren of the frail elderly are near or of middle-aged adults now may ty and increased longevity have have reached retirement age.1 In expect to provide care. increased the odds of joint survival developing countries, the adult chil- Additionally, life expectancy has of different generations within a dren may well have children of their increased for the disabled, the men- family. In developed countries, own living in the household. Some tally retarded, and the chronically joint survival has manifested itself of the adult children may bear ill. Today’s care for older people in the “beanpole family,” a vertical health limitations of their own. may be more physically and psy- extension of family structure char- Those frail elderly without children chologically demanding than in the acterized by an increase in the may face institutionalization at ear- past, especially with regard to the number of living generations within lier ages than will people with sur- increased numbers of people with a lineage and a decrease in the viving adult children. cognitive diseases. As advances in number of members within each One measure of the pressure the medical technology affect the ability generation (Bengston, Rosenthal, sandwich generation may experi- to extend life, it is at least plausible and Burton, 1995). As mortality ence by caring for elderly parents is to expect the duration of chronic ill- rates continue to improve, more the parent support ratio (PSR), ness and the consequent need for and more people in their fifties and defined here as the number of peo- help to increase further, even if the sixties are likely to have surviving ple aged 80 and over per 100 peo- average age at onset of disability parents, aunts, and uncles. More ple aged 50 to 64, which in a gen- rises. children will know their grandpar- eral sense relates the oldest old to ents and even their great- In all countries examined except their offspring who were born when grandparents, especially their great- Bangladesh, Jamaica, Morocco, and grandmothers. There is no Pakistan, the PSR is projected to be historical precedent for a majority 1 It should be noted that the idea of a “sand- higher in 2030 than in 2000. The of middle-aged and young-old wich generation” needs further empirical eluci- ratio has and will evolve very differ- dation. Some researchers question the extent adults having living parents. to which middle generations actually provide ently within and among world Menken (1985) has estimated that care for younger and older generations. One regions, however (Figure 8-6). In study of 12 European Union countries (reported one in three women 50 years old in Hagestad, 2000) reports that only 4 percent South America, Eastern Asia, and had living mothers in the United of men and 10 percent of women aged 45-54 have overlapping responsibilities for children Western Europe, PSRs more than States in 1940, whereas by 1980 and for older persons who require care. The doubled between 1950 and 2000. importance of the “sandwich” concept in a the proportion had doubled to two PSRs will continue to rise between given society is likely to be determined, in part, in three. by the nature of formal institutions that provide 2015 and 2030. In Western Africa, assistance to elderly individuals and to families in general. most countries experienced little U.S. Census Bureau An Aging World: 2001 79
  • 82. change in the PSR from 1950 to older men who were receiving FAMILY ABILITY TO CARE FOR 2000, and the aggregate level will assistance with an instrumental ELDERLY MEMBERS MAY BE remain low in 2030 even though activity of daily living2 had their CHANGING absolute numbers of oldest-old peo- wife as the caregiver. However, Living with other people reduces ple in some nations are growing only 33 percent of older women the likelihood of using formal med- rapidly. The most pronounced relied on their husband as the care- ical care and increases the use of changes have occurred in the indus- giver, while 58 percent were aided informal care, at least in the U.S. trialized world. In 2000, the PSR by a daughter (Beland and context (Cafferata, 1988). Since was 20 or higher in 12 such nations Zunzunegui, 1996). most physical, emotional, and eco- (and also in Israel and Uruguay). nomic care to older individuals is Whether in the role of spouse or The difference in parent support provided by family members, the daughter, the fact remains that ratios between developed and demography of population aging is women provide the bulk of informal developing countries reflects differ- increasingly concerned with under- and/or long-term care for elderly ent trends in fertility. In 2030, standing and modeling kin availabil- people worldwide. While joint sur- those aged 50 to 64 (the potential ity. Kin availability refers to the vival increases the number of elder- support givers) were born between number of family members who will ly couples, the average woman 1966 and 1980. In most develop- potentially be available to elderly eventually outlives her husband and ing countries, fertility was still high individuals if and when various may have to rely on other family during this period, or just begin- forms of care are needed. A study members for personal care.3 Most ning to decline. Hence, this age by Tomassini and Wolf (2000) exam- studies (see, for example, Jenson group will be fairly large, resulting ined the effects of persistent low and Jacobzone, 2000, and the com- in low parent support ratios. In fertility in Italy on shrinking kin net- pilation of research in Blieszner and developed countries, fertility was works for the period 1994-2050; Bedford, 1996) have indicated that fairly low during this period, pro- throughout the simulation period, these other family members are ducing small birth cohorts that will about 15-20 percent of Italian women. Therefore, a variant of the result in higher parent support women aged 25 to 45 are the only parent support ratio may be useful, ratios in 2030. living offspring of their surviving namely, the ratio of people aged 80 mothers and thus are potentially SPOUSE MAY BE MOST LIKELY and over to women aged 50 to 64. fully responsible for their mothers’ TO PROVIDE CARE FOR Changes in this parent support ratio care. While reduced fertility and ELDERLY for females (PSRF) are similar to smaller families obviously imply A clear cross-national picture of those in the PSR, but the PSRF levels fewer potential caregivers, this caregiving for the elderly has yet to are much higher. In 2000, the PSRF effect is offset to some extent by emerge. In the 1980s, the stereo- was 54 in Norway and Sweden, the increased longevity. Modeling is typical view of caregiving was that highest level among the 52 study further complicated by the fact that of children caring for their aged nations. Most developed countries while demographic forces impose parent(s). More specifically, it gen- had PSRF levels in the 30s and 40s, constraints on family, household, erally was thought that adult while many developing nations had and kin structures, these structures daughters and daughters-in-law pro- PSRFs of 15 or less in 2000. also are determined by social and vided most of the personal care and Projections for the year 2030 sug- cultural factors that are difficult to help with household tasks, trans- gest that in Japan there will be 100 measure (Myers, 1992; Wolf, 1994; portation, and shopping for the eld- people aged 80 and over per 100 Van Imhoff, 1999). erly (United Nations, 1985). women aged 50 to 64, the highest Although this may still be the case, level among the 52 nations. Research is now addressing increases in joint survival mean whether the high rates of divorce that, for many older people in both observed in some nations will result developing as well as developed 2 Instrumental activities of daily living in a lack of kin support for people include preparing meals, shopping for personal countries, the main person who items, managing money, using the telephone, in older age, and whether “blended” provides care is their spouse and doing light housework. families and other forms of social 3 Although, in countries with relatively high (Shuman, 1994). One survey in levels of income, market developments increas- arrangements will, in the future, Spain found that 74 percent of ingly allow older individuals or their children to provide the types of care and purchase care services directly if desired. 80 An Aging World: 2001 U.S. Census Bureau
  • 83. support that are common today Figure 8-7. (Wachter, 1998; Murphy, 2001). Childlessness Among U.S. Women 40 to 44 Years The consensus to date foresees a Old: 1976 to 1998 declining biological kinship support (Percent) network for elderly people in devel- oped and many developing coun- 19 tries. Childlessness is another trait that will affect the nature of future 16 caregiving. Data over time for the 15 United States in Figure 8-7 show the increasing likelihood of being child- 11 less among women aged 40 to 44; 10 10 nearly one out of five such women in 1998 had no children. Trends in this characteristic could be an important determinant of eventual care arrangements as current and future cohorts of middle-aged women reach older age. 1976 1980 1984 1988 1992 1998 The issue of kin availability has Source: U.S. Census Bureau, June Current Population Surveys, 1976 to 1998. become especially important in the context of East and Southeast Asian countries, driven in large part by the rapid declines in fertility that Figure 8-8. have greatly reduced the average Percent of Women at Age 65 in South Korea family size of young-adult cohorts. With No Surviving Son: 1975 to 2025 The complex interplay of demo- Married graphic and cultural factors is illus- Widowed trated by the case of the Republic of Korea. There, two-thirds of the elderly are economically dependent on their adult children (Korea 30 29 Institute for Health and Social 26 Affairs, 1991), and cultural norms 24 dictate that sons provide economic support for elderly women who 17 17 have lost their spouses. Lee and 15 15 Palloni (1992) have shown that 13 declining fertility means an increase 11 10 10 in the proportion of Korean women with no surviving son (Figure 8-8). At the same time, increased male longevity means that the proportion of elderly widows also will decline. 1975 1985 1995 2005 2015 2025 Thus from the elderly woman’s Source: Lee and Palloni 1992. point of view, family status may not U.S. Census Bureau An Aging World: 2001 81
  • 84. deteriorate significantly in the coming years. From society’s per- Figure 8-9. spective, however, the demand for Proportion of Elderly People Receiving Home support of elderly women is likely Help Services: Early-to-Mid 1990s (In percent) to increase. The momentum of rapid population aging means that the fraction of the overall popula- Finland 24 tion that is elderly women (especial- Denmark 17 ly sonless and childless widows) Norway 14 will increase among successive Sweden 13 cohorts. Given the strong trend United Kingdom 13 toward nuclearization of family Netherlands 8 structure in the Republic of Korea, France 7 and the traditional absence of state Australia 7 involvement in socioeconomic sup- Belgium 6 port, the future standard of living United States 4 for a growing number of elderly Austria 3 widows is tenuous. A similar 3 Ireland prospect looms in Taiwan and Japan Germany 2 (Hermalin, Ofstedal, and Chi, 1992; Japan 2 Jordan, 1995). Simulations of kin Canada 2 availability in rural China (Jiang, Spain 2 1994) are more optimistic, suggest- ing that, in spite of relatively low Portugal 1 fertility, improvements in mortality New Zealand 1 will ease the future burden on the Italy 1 family support system. Only a very Note: Data for France, Australia, and Italy refer to 1985, 1988 and 1988, respectively. small percentage of rural house- Source: OECD, 1996. holds will have to support two or more elderly parents, and relatively now promote policies to maintain countries (Figure 8-9). Such few elderly will be childless. At the and support frail elderly people in services reached nearly one-fourth same time, simulations using fami- their own homes and communities of all elderly in Finland in 1990, up ly-status life table models devel- for as long as possible. Given the slightly from the level of 22 percent oped by Zeng, Vaupel, and changing nature of the family (in its in 1980. The available data sug- Zhenglian (1997; 1998) suggest many perturbations) and patterns of gest that countries with more that the family household structure kin availability, the development extensive provision of home help and living arrangements of Chinese and use of home help services services are those that have had a elders may change markedly during would appear to be a reasonable prolonged process of population the first half of this century; by step toward reducing the need for aging and now have relatively high- 2050, the percentage of Chinese institutionalization. To date, how- er proportions of oldest-old resi- elderly living alone could be 11 and ever, the use of home help appears dents (OECD, 1996). Structural pro- 12 times larger in rural and urban to be widespread only in grammatic factors also are areas, respectively, than in 1990. Scandinavian countries and the important, insofar as government HOME HELP SERVICES ARE United Kingdom. Comparative data support or subsidization of home MOST PREVALENT IN assembled by the Organization for help will almost certainly result in SCANDINAVIA Economic Co-Operation and greater use. In the United States, The previous chapter alluded to a Development from the early-to-mid- the use of home health care servic- change in social and governmental 1990s show that the proportion of es has grown substantially since the thinking about the desirability of elderly people receiving home help late 1980s, largely as a result of institutionalization. Some nations exceeds 10 percent in only five changes in medicare policy that 82 An Aging World: 2001 U.S. Census Bureau
  • 85. have made home health benefits production as well as paid employ- Grandparents in some developed available to more beneficiaries for ment (Hashimoto, 1991; Apt, 1992). countries often provide day care for longer periods of time. This in turn children so the grandchildren’s par- An important component of many has stimulated the home health ents can work or go to school. In older people’s lives is their role as care industry; the number of home the United States in 1995, 29 per- the giver of care. Older people pro- health agencies more than tripled cent of preschool children whose vide care for a variety of people between 1980 and 1994 parent(s) worked or were in school (spouses, older parents, siblings, (Freedman, 1999). were cared for by a grandparent children, and grandchildren) and do (Smith, 2000), typically the grand- ELDERLY PROVIDE AS WELL AS so for many reasons (illness of a mother. Because of the lack of ade- RECEIVE SUPPORT spouse or sibling, increased number quate day care in many Eastern of single-parent families, increased Many elderly receive financial help European countries and nations of female labor force participation, from adult children, but support is the former Soviet Union, the care orphaned grandchildren). Often the not a one-way street. In countries that grandmothers (babushkas) pro- care provided by older family mem- with well-established pension and vide for grandchildren may be inte- bers is essential to the well-being of social security programs, many gral to family functioning. a family. older adults give support (including In many Asian countries, where financial help, shelter, childcare, and THE IMPORTANCE OF coresidency is the norm, propor- the wisdom of experience) to their GRANDPARENTS tions of grandparents providing adult children and grandchildren. In some countries, nontrivial pro- care for grandchildren are substan- In the North American context, portions of older women and men tial. In the Philippines, Thailand, studies suggest that elderly parents are providing care to their grand- and Taiwan, approximately 40 per- are more likely to provide financial children. This care ranges from cent of the population aged 50 and help than to receive it (Soldo and occasional babysitting to being a older lived in a household with a Hill, 1993; Rosenthal, Martin- custodial grandparent. Survey data minor grandchild (under 18 years of Matthews, and Matthews, 1996). for the United States from the mid- age). In these same countries, The elderly in developing countries 1990s (Fuller-Thomson and Minkler, approximately half or more of those appear less likely than in developed 2001) indicate that 9 percent of all aged 50 and older who had a cores- countries to provide financial help; Americans with grandchildren under ident grandchild aged 10 or data from the Malaysian Family Life age 5 were providing extensive younger provided care for the child Survey indicate that the main direc- caregiving.4 In 1997, 3.9 million (Hermalin, Roan, and Perez, 1998). tion of monetary transfers between children (5.5 percent of all children As in the United States, grandmoth- noncoresident parents and children under age 18) lived in a household ers are more likely than grandfa- is from the latter to the former maintained by their grandparents thers in Asian countries to provide (Lillard and Willis, 1997). Ongoing (Casper and Bryson, 1998). Since care for their grandchildren (Chan, research in Asia is beginning to 1990, the number of children living 1997; Uhlenberg, 1996). reveal the complexity of familial in households headed by grandpar- exchange, not just among parents Many grandparents find themselves ents has increased, especially for and children but among wider fami- in the position of going beyond pro- children in households with only ly and social networks as well viding occasional care to becoming grandparents and grandchildren. (Agree, Biddlecom, and Valente, the sole providers of care for their Trends in several factors (e.g., 1999). Beyond the financial realm, grandchildren. One reason for this divorce, HIV/AIDS, drug abuse, and it seems clear that older persons in situation is the migration of the child abuse) may have contributed developing countries make substan- middle generation to urban areas to to the increase in these types of tial contributions to family well- work. Past research has found that families. being, in ways ranging from social- this is not unusual in Afro- ization to housekeeping and child Caribbean countries (Sennott-Miller, 4 Extensive caregiving in this context meant care. Such activities free younger 1989). These “skip-generation” providing at least 30 hours of child care in an adult women for employment in average week and/or caring for grandchildren families are found in all regions of for at least 90 nights in 1 year. unpaid family help in agricultural U.S. Census Bureau An Aging World: 2001 83
  • 86. the world and may be quite preva- of the epidemic are particularly Africa is estimated to be 12.1 mil- lent. One study in rural Zimbabwe devastating in Sub-Saharan Africa, lion (UNAIDS/WHO 2000). For found that 35 percent of house- where it is estimated that in 1999 many of these children, grandpar- holds were skip-generation house- 8.6 percent of the population aged ents have become the main care- holds (Hashimoto, 1991). 15 to 49 was infected with an HIV giver (Levine, Michaels, and Back, virus that causes AIDS. High rates 1996). One study (Ryder et al., THE HIV/AIDS EPIDEMIC IS of adult infection and AIDS deaths 1994) in the city of Kinshasa found CHANGING GRANDPARENTS’ leave many children in need of that the principal guardian for ROLES care. The cumulative number of 35 percent of AIDS orphans was a The AIDS epidemic has affected the AIDS orphans5 in Sub-Saharan grandparent. number of grandparents who are caring for grandchildren in many 5 AIDS orphans are defined as HIV-negative countries of the world. The effects children who lost their mother or both parents to AIDS when the children were under age 15. 84 An Aging World: 2001 U.S. Census Bureau
  • 87. CHAPTER 9. Educational Attainment and Literacy Educational attainment is linked to aware of the benefits and disadvan- implications for intergenerational many aspects of a person’s well- tages of certain types of behaviors solidarity (Choi, 1992). being. Research has shown that associated with personal health. Educational attainment of the eld- higher levels of education usually Education also is related to joint erly varies substantially among the translate into better health status, survival of spouses, to living countries in this report. The latest higher incomes, and consequently arrangements, and to changing data for the United States show higher standards of living (Guralnik value systems which have et al., 1993; Preston and Taubman, 1994; Smith and Kington, 1997; Liu, Figure 9-1. Hermalin and Chuang, 1998). Percent of People With Completed Male, 25-44 People with higher educational lev- Secondary Education or More Female, 25-44 els tend to have lower mortality Male, 65+ rates and better overall health than Female, 65+ their less-educated counterparts (Elo and Preston, 1996; Zimmer et al., 86.9 89.0 1998), as well as better cognitive United States, 1998 67.6 functioning in older age (Stern and 66.6 Carstensen, 2000). Part of the rea- 21.6 son for this finding is that more- 23.9 Brazil, 1991 educated people tend to have higher 6.6 4.9 incomes throughout their lifetime, 17.2 which means they can afford better 12.8 health care than people with lower Bolivia, 1992 4.7 levels of education. Higher work- 2.3 ing-life income also translates into 20.4 13.4 higher levels of retirement savings China, 1990 2.8 and income. Hence, people with 0.5 higher educational levels may be 94.6 less dependent on their family for 92.1 Romania, 1992 45.5 financial assistance in later years. 31.9 Education significantly affects how 87.3 91.9 effectively people utilize health Russia, 1994 31.8 care. In the United States, for 21.6 example, where educational levels 64.0 of the elderly are relatively high, 61.3 Singapore, 1995 many older people, especially those 15.2 5.4 aged 85 and older, have trouble 79.6 understanding basic medical 82.5 Sweden, 1998 instructions. Even something as 44.0 simple as taking medicine correctly 38.0 may be a problem. Education fur- Note: Data for Sweden 65+ refer to ages 65-74. ther affects health because well- Sources: U.S. Census Bureau, 2000a and national sources. educated people may be more U.S. Census Bureau An Aging World: 2001 85
  • 88. that two-thirds of all people aged 65 and older had completed at least Figure 9-2. secondary education.1 Comparable Percent of People With Completed Male, 35-44 completion levels in other devel- University Level of Education: 1998 Female, 35-44 oped countries are somewhat lower. Male, 55-64 Less than a third of the elderly in Developed countries Female, 55-64 Russia, for example, had finished 19 Australia 18 secondary-level education (Figure 11 9-1). Levels of education of the eld- 9 19 erly are much lower in developing Germany 14 15 countries. In Brazil, Bolivia, and 5 China, less than 5 percent of the 12 Italy 11 elderly population had a completed 7 3 secondary education. 29 Netherlands 23 FUTURE ELDERLY WILL HAVE 22 12 MORE EDUCATION 9 11 During the twentieth century, educa- Poland 11 9 tional attainment has increased 6 markedly in most countries of the Portugal 7 5 world. This improvement is clearly 4 reflected in the data on educational 14 Sweden 13 attainment by age. In some devel- 12 11 oped countries, younger cohorts are 27 more than twice as likely as the eld- United States 26 26 erly to have completed secondary 18 education. In developing countries, the difference between younger and Developing countries older cohorts is even more striking. 9 Women aged 25 to 44 in Bolivia were Argentina 11 5 more than five times as likely as 4 women aged 65 and older to have a 8 9 completed secondary education. Brazil 6 4 The educational attainment of the 4 2 elderly has risen during the last China 5 2 several decades in many countries, 11 and will continue to increase in the 3 India 5 future. For example, around 1 27 percent of the elderly in the 3 Indonesia 1 United States had completed at 1 0 least secondary education in 1970; 33 by 1998 the percentage had Jordan 18 15 2 1 Educational attainment in this report refers 26 in theory to completion of a particular educa- 11 South Korea 14 tional level. Data have been derived from pri- mary national tabulations as well as from fig- 2 ures reported to international organizations. 9 While large differences in educational attain- 4 Turkey 5 ment exist, at least some of the variation is like- ly due to different concepts, definitions, and 1 methods of data collection. We have attempted to make the data on educational attainment in Note: Percent for females aged 55-64 in Indonesia is zero. Source: Organization for Economic Co-Operation and Development, 2001. this report as comparable as possible across countries. 86 An Aging World: 2001 U.S. Census Bureau
  • 89. older were literate in 1991, and just Figure 9-3. 67 percent of the age group 75 and Estimated Literacy Rates for Population Aged 60 and over. Proportions literate among Over, by Sex, in Five Developing Regions: 1980 and 1995 older Greek women were even (Percent literate) 1980 lower — approximately two-thirds 1995 of women aged 65 and older and 26 M 36 slightly over half of women aged 75 Arab States 6 and older. F 11 40 In developing countries, literacy M 64 may be uncommon among older Eastern Asia/Oceania 9 populations. Many of today’s elder- F 27 ly lived much of their lives prior to 68 M 75 the rapid increase in educational Latin America/ 58 attainment that occurred in the sec- Caribbean F 68 ond half of the twentieth century. 21 Consequently, many older people, M 34 Sub-Saharan Africa and again particularly women, have 9 F low levels of literacy. While cohort 14 33 changes ensure that the future edu- M Southern Asia 41 cation profile of the elderly will 8 improve, it is important to remem- F 13 ber that in many countries, a major- Source: United Nations Educational Scientific and Cultural Organization, 1995. ity of today’s elderly are illiterate (Hugo, 1992). This fact needs to be explicitly recognized and consid- jumped to 67 percent. As younger, 5 to 25 percent in the developed ered when developing programs to more-educated cohorts continue to countries and from 1 to 12 percent assist older populations. age, their attainment levels will be in developing countries. Small pro- reflected in the educational status portions of women aged 55 to 64 Figure 9-3 presents estimated litera- of tomorrow’s elderly. in developed countries have a uni- cy rates for the population aged 60 versity education, and proportions and over, by sex, in five developing UNIVERSITY EDUCATION NOT in most developing countries are regions for 1980 and 1995.2 In all YET THE NORM AT ANY AGE smaller still. five regions, older men are more lit- Although educational attainment erate than older women. In three of has been improving throughout the LITERACY RATE OF MANY the five regions, less than half of ELDERLY POPULATIONS STILL twentieth century, university-level older men and less than 15 percent LOW education still is not widespread. of older women were literate in Relatively few people complete this In many developed countries, litera- 1995. Among developing regions, level of education and the propor- cy data no longer are collected Latin America and the Caribbean tion is usually lowest among the because education, at least at the has the highest aggregate literacy elderly. In many developed coun- primary level, is so widespread that levels for older populations; three- tries, less than a third of people literacy is considered to be univer- quarters of men and two-thirds of aged 35 to 44 have a university sal. However, this is not always the women aged 60 and over were liter- education (OECD, 1998a; 2001). case for the elderly, particularly for ate, similar to the levels noted The proportion of the “near elderly” older women and the oldest old. above in Greece. (i.e., aged 55 to 64) with this level Data from some countries that still of education is even lower (Figure collect literacy information show 2 These estimates were produced by the Division of Statistics of UNESCO using the most 9-2). For the set of countries that substantial proportions of the recently available national data. For details included in this graph, among men elderly may be unable to read and about the methodology used to produce these estimates see UNESCO, 1995, Methodology aged 55 to 64, the proportion with write. In Greece, for example, only Used in the 1994 Estimates and Projections of a university education ranged from 77 percent of people aged 65 and Adult Illiteracy, Statistical Issue STE-18, Division of Statistics, Paris. U.S. Census Bureau An Aging World: 2001 87
  • 90. Countries vary greatly in rates of lit- eracy among the elderly. In Chile, Figure 9-4. over 80 percent of the elderly were Percent Literate in Two Age Groups: literate in 1992, while in Uganda Latest Available Year Male less than a fifth of the elderly were Female literate (Figure 9-4). In most coun- Bolivia, 1992 tries, older men are much more 91.9 25-44 likely to be literate than older 76.6 53.8 women. In Brunei, older men were 65+ 28.3 four times as likely to be literate as Botswana, 1993 were older women. Although very 70.2 small proportions of the elderly 25-44 75.2 population are literate in many 12.9 developing countries, the rates rise 65+ 11.3 rapidly for younger cohorts. For all Brunei, 1991 the countries in Figure 9-4, with the 96.0 exception of women in Uganda, 25-44 91.1 well over half of the population 45.1 65+ 10.6 aged 25 to 44 were literate. And among this younger age group, the Chile, 1992 difference in literacy by sex tends 96.6 25-44 to dissipate. 96.9 81.3 65+ 80.9 GENDER DIFFERENTIAL IN EDUCATIONAL LEVELS AMONG Iran, 1991 THE ELDERLY IS OFTEN 79.4 SUBSTANTIAL 25-44 56.1 31.3 In nearly all countries, older men 65+ 14.0 have higher average levels of educa- Jordan, 1994 tion than do older women. Just as 96.6 overall levels of education vary wide- 25-44 86.9 ly among countries, so does the gen- 51.0 65+ der difference in education at older 16.9 ages. The gender gap3 in education- Thailand, 1990 al attainment of the elderly is larger 97.3 in many developed countries than in 25-44 95.4 developing countries, where the 76.3 65+ 53.2 level of education for the elderly is so low for both men and women Uganda, 1991 that the difference between them is 73.0 25-44 small. In many developed countries, 44.0 26.7 where overall attainment levels are 65+ 7.2 3 The gender gap is defined as the absolute Zimbabwe, 1992 difference in percentage points between the 89.5 educational level of men and women. For 25-44 instance, in Romania in 1992, 45.5 percent of 74.5 men and 31.9 percent of women aged 65 and 41.0 65+ older had completed secondary or higher edu- 23.3 cation. The gender gap for the population aged 65 and older with these levels of schooling Sources: United Nations Educational Scientific and Cultural Organization, 1995 and country sources. would be the difference between the two levels (13.6 points). 88 An Aging World: 2001 U.S. Census Bureau
  • 91. the disadvantages that today’s Figure 9-5. older women may face because of Gender Gap in Literacy Rates for Selected their lower levels of education rela- Age Groups: Latest Available Year tive to men should begin to abate when these younger cohorts reach Bolivia, 1992 the ranks of the elderly. 25-29 9.1 35-39 18.6 GENDER GAP IS INVERSELY 55-59 32.9 RELATED TO AGE AT OLDER 60-64 31.6 AGES 65+ 25.5 Examining the gender gap by age in Brunei, 1991 developing countries reveals the 25-29 0.9 differential rate of improvement in 35-39 6.4 educational attainment. Figure 9-5 55-59 48.0 presents the gender gap in literacy 60-64 43.7 rates for five age groups in several 65+ 34.5 countries with data from the 1990s. Iran, 1991 A somewhat counter-intuitive pic- 25-29 18.9 ture emerges for the three older age 35-39 26.5 groups, namely, that the gender 55-59 23.4 gap decreases as age increases. In 60-64 19.7 other words, there is a larger 65+ 17.3 absolute difference between male Jordan, 1994 and female literacy rates at ages 55 25-29 3.9 to 59 than among people aged 65 35-39 13.9 47.3 and over. The increase in the gen- 55-59 der gap for younger-old age groups 60-64 44.8 reflects historical patterns of educa- 65+ 34.1 Uganda, 1991 tional promotion. When countries 23.9 with low overall levels of education 25-29 32.5 and limited resources began to 35-39 55-59 35.4 improve the educational attainment 60-64 27.3 of their populations, the initial 65+ 19.5 focus was more on educating males Yemen, 1994 than females. In most developing 25-29 48.5 countries, people aged 55 and over 35-39 37.8 were of school age when formal 55-59 28.7 education was not widespread. 60-64 21.6 Although educational attainment 65+ 19.4 was improving, it was improving more for men than for women. Note: The gender gap is defined as the absolute difference in percentage points between the literacy rate for men and women. Thus, for these older age groups, Source: United Nations Educational Scientific and Cultural Organization, 1995 and country sources. the gender gap is less among the elderly than among people aged 55 to 64. For some countries in Figure higher, the difference between older older cohorts. In various coun- 9-5, the gender gap at ages 25 to men and women is larger. tries, younger women complete 29 and 35 to 39 is smaller than at secondary education at higher As suggested by the data on litera- the older ages, indicating a more rates than do men, and in some cy discussed earlier, gender differ- equal inclusion of both sexes in nations the gender difference in ences in educational attainment educational programs in recent university-level attainment at are much smaller for younger than years. younger ages is negligible. Thus, U.S. Census Bureau An Aging World: 2001 89
  • 92. EDUCATIONAL DISADVANTAGE IS COMMON IN RURAL AREAS Figure 9-6. The quality and quantity of rural Percent Literate by Age, Sex, and Urban and Rural Residence, Yemen: 1994 educational facilities in most nations tend to be inferior to those Percent in urban areas. Consequently, liter- 100 acy levels and educational attain- ment are lower in rural areas, par- ticularly in developing countries. 80 Male urban Data for Yemen (Figure 9-6) illus- trate the common pattern wherein the rural disadvantage in literacy is 60 evident for both sexes (i.e., rural Male rural males have lower rates than urban males and rural females have lower 40 rates than urban females). It Female rural appears that differences by gender are greater than differences by 20 urban/rural residence. Urban Female urban women have lower literacy rates than rural men, except at the very 0 youngest ages. The sexes also dif- 10 15 20 25 30 35 40 45 50 55 60 65 70 75+ fer in size of the rural/urban gap by Age age. Rural males consistently have Source: Population and Housing Census of Yemen, 1994. lower literacy rates than urban Table 9-1. Earnings Ratios of Selected Age Groups in 15 Countries by Level of Educational Attainment: 1995 55-64 years/ 45-54 years/25-29 years 45-54 years Country Less than upper Upper Nonuniversity secondary secondary tertiary University Overall Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.30 1.26 1.35 2.06 0.85 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.05 1.46 1.37 1.96 0.86 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.21 1.23 1.29 1.60 0.92 Finland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.43 1.36 1.69 2.11 0.90 France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.18 1.47 1.45 1.95 1.07 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.97 1.28 1.10 1.76 0.97 Ireland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.24 1.59 1.59 2.25 (NA) Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.23 1.44 1.64 1.99 0.86 Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.13 1.61 1.33 1.65 0.84 New Zealand . . . . . . . . . . . . . . . . . . . . . . . 1.25 1.39 1.16 1.93 0.95 Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.10 1.26 1.88 1.67 (NA) Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.38 1.26 1.67 1.70 0.90 Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . 1.06 1.25 1.52 1.80 0.97 United Kingdom . . . . . . . . . . . . . . . . . . . . . 0.93 1.09 1.41 1.50 0.81 United States . . . . . . . . . . . . . . . . . . . . . . . 1.29 1.28 1.39 1.67 0.89 NA Not available. Note: Ratios reflect gross annual earnings before taxes. Data for Finland and Ireland refer to 1994 and 1993 respectively. Source: Organization for Economic Co-Operation and Development (Employment Outlook 1998). 90 An Aging World: 2001 U.S. Census Bureau
  • 93. males in Yemen, and the gap reading skills generally worsen with labor force entrants (25 to 29 years) between the two areas is fairly even age; a recent study also demon- generally rises with educational across the age spectrum. For strated that functional health attainment level, but the right-hand women, on the other hand, the gap literacy is markedly lower at older column of Table 9-1 indicates an between rural and urban literacy ages, even after controlling for vari- overall decline in earnings of people levels is much wider at the younger ables such as cognitive dysfunction, 55 to 64 relative to people 45 to than the older ages, suggesting that physical functioning and visual acu- 54. Because educational attainment in recent years urban women have ity (Baker et al., 2000). This study may be very different by age been afforded greater relative raises questions about the effects of cohort, this table may confound access to education than have their life-long education, the efficacy of pure age effects with returns to rural counterparts. tests that measure cognitive ability, education. Further analysis of avail- and other age-related changes that able data suggests that, when aver- EDUCATION AFFECTS OTHER may affect testing procedures. aged over all countries, the earn- DIMENSIONS OF LIFE ings premium of peak-earning Table 9-1 shows ratios of average As mentioned earlier, education is workers relative to recent entrants earnings for workers in three age related to health behavior and rises considerably with higher edu- groups. The earnings ratio of per- income accumulation throughout cational attainment (OECD, 1998 sons in the so-called peak earning the life course. Studies in the Employment Outlook). years (45 to 54) to those of recent United States have shown that U.S. Census Bureau An Aging World: 2001 91
  • 94. CHAPTER 10. Labor Force Participation and Retirement Rapid growth of elderly populations Figure 10-1 presents data on formal with age. On the other hand, the may put pressure on a nation’s economic activity for older women work status of older workers differs financial resources. This concern is and men in three countries, chosen dramatically among the countries. based, at least partially, on the to represent different levels of eco- In Rwanda, more than three- assumption that the elderly do not nomic development. Some of the quarters of all women aged 60 to contribute to the economy. However, patterns mentioned above are 64 are economically active, and many older people do work, and apparent in these data; older even at ages 70 and older, a sub- examining the labor force participa- women have lower participation stantial number remain active in the tion and characteristics of older rates than older men, and participa- labor force. In contrast, although a workers gives a clearer picture of tion rates for both sexes decrease nontrivial proportion of women their contribution. Information on older workers also is useful in plan- ning economic development and the Figure 10-1. financing of retirement. Economically Active Older Population by Age for Rwanda, Peru, and New Zealand: Late 1990s Some characteristics of older work- (Percent) ers seem not to vary among coun- Male Female tries. In all countries, the elderly Rwanda, 1996 account for a small proportion of 86.9 60-64 the overall labor force. Their share 77.8 of the total labor force in the study 81.5 65-69 73.4 countries ranges from less than 66.0 1 percent to 7 percent. A second 70-74 51.7 commonality is that labor force par- 45.6 75+ ticipation declines as people near 32.5 retirement age. A third is that par- Peru, 1999 ticipation rates are higher for older 72.5 60-64 men than for older women. 38.2 56.3 Other characteristics of older work- 65-69 23.4 ers show interesting differences 70-74 46.2 across countries. The rate of 26.2 29.0 participation of older workers varies 75+ 11.1 substantially, and generally is lower New Zealand, 1999 in developed than in developing countries. Only 2 percent of men 57.5 60-64 32.5 aged 65 and over participate in the 20.4 labor force in some developed 65-69 10.0 countries, whereas in certain devel- 8.8 70-74 oping countries well over half of 2.9 2.8 elderly men are economically active. 75+ 1.1 The occupational concentration of older workers also varies widely Source: International Labour Office (various issues of the Yearbook of Labour Statistics). among countries. U.S. Census Bureau An Aging World: 2001 93
  • 95. aged 60 to 64 are economically active in New Zealand, participation Figure 10-2a. rates decrease dramatically with Labor Force Participation Rates for Men Aged 55 age so that for women aged 70 to to 59 in Developed Countries 74, less than 3 percent are still (Percent) Early 1970s active. Cross-national differences in Late 1990s levels of labor force activity are 88.4 associated with societal wealth; Australia 72.5 countries with high GNP (gross Austria 83.7 national product) tend to have 63.7 much lower labor force participa- 82.3 Belgium tion rates of the elderly and near 52.4 elderly than do low-income coun- 86.5 Bulgaria 58.0 tries (Clark, York, and Anker, 1997). Canada 84.9 In richer countries, the elderly or 72.2 near elderly can afford to retire 85.0 Czech Republic because of pension schemes or 77.1 social security systems. These pro- 81.8 France grams are often lacking in poorer 70.4 countries. Germany 86.8 76.5 84.4 Hungary 45.9 75.0 Italy 54.1 94.2 Japan 94.7 79.3 Luxembourg 53.4 92.1 New Zealand 81.2 Poland 90.9 55.2 88.4 Sweden 84.4 United States 86.8 78.4 Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). 94 An Aging World: 2001 U.S. Census Bureau
  • 96. TIME TREND IN LABOR FORCE PARTICIPATION DIFFERS BY Figure 10-2b. GENDER Labor Force Participation Rates for Men Aged 60 to 64 in Developed Countries The trend in most developed coun- (Percent) Early 1970s tries has been for labor force Late 1990s participation rates for older men to decline in recent decades. Figure Australia 75.6 46.7 10-2 shows male labor force partici- 44.9 pation rates for three older age Austria 16.7 groups in 16 developed countries; Belgium 79.3 in all of these countries, participa- 18.6 tion rates declined between the Bulgaria 33.6 early 1970s and the late 1990s. 11.1 These declines are particularly pro- 74.1 Canada nounced for men aged 60 to 64. 46.6 33.3 In ten of the sixteen countries in Czech Republic 27.5 the early 1970s, well over half of 54.6 men aged 60 to 64 were still active. France 16.4 Germany 68.8 30.3 43.7 Hungary 10.6 40.6 Italy 31.7 85.8 Japan 74.1 Luxembourg 45.5 15.5 New Zealand 69.2 57.4 Poland 83.0 33.4 75.7 Sweden 55.5 United States 73.0 54.8 Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). U.S. Census Bureau An Aging World: 2001 95
  • 97. In the remaining six countries activ- ity rates ranged from 33 percent to Figure 10-2c. 46 percent. By the late 1990s, only Labor Force Participation Rates for Men Japan, New Zealand, Sweden, and Aged 65+ in Developed Countries the United States had male partici- (Percent) Early 1970s pation rates over 50 percent. Rates Late 1990s also have fallen for the 65-and-over Australia 22.2 age group.1 In the early 1970s, 9.6 only two countries in Figure 10-2 8.0 Austria had participation rates lower than 4.6 10 percent for elderly men; by the 6.8 Belgium late 1990s, most of the countries 2.8 had rates less than 10 percent. But 10.3 Bulgaria 4.5 as discussed later in this chapter, Canada 23.6 the trend in declining participation 9.9 rates for older men has stopped or 14.6 Czech Republic even reversed in a number of devel- 7.2 oped countries. 10.7 France 2.3 Germany 16.0 1 In the United States, much of the decline in 4.5 labor force participation rates for elderly men 16.7 occurred earlier in the twentieth century. Hungary According to Costa (1998), 70 percent of the 3.8 decline in participation rates between 1880 and 13.4 1990 for men aged 65 and older occurred Italy 6.3 before 1960. Japan 54.5 35.5 Luxembourg 10.1 1.9 New Zealand 21.3 10.4 Poland 56.4 15.3 Sweden 15.2 United States 24.8 16.9 Note: Later data for Belgium 65+ refer to ages 65-69 and for Hungary to ages 65-74; data for Sweden 65+ are not reported by the ILO. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). 96 An Aging World: 2001 U.S. Census Bureau
  • 98. The trend for older women in these Figure 10-3a. developed countries differs from Labor Force Participation Rates for Women the male pattern. In many coun- Aged 55 to 59 in Developed Countries tries, female participation rates (Percent) Early 1970s have increased for almost all adult Late 1990s age groups up to age 60, whereas rates for elderly women have 28.3 Australia declined (Figure 10-3). In some 44.6 35.8 cases, the increase among women Austria 25.4 aged 55 to 59 has been quite 20.0 marked. In New Zealand, for exam- Belgium 27.6 ple, 60 percent of women aged 55 Bulgaria 26.1 to 59 were economically active in 10.7 1998, up from 28 percent in 1971. 38.7 Canada 50.6 36.5 Czech Republic 33.2 42.1 France 51.7 34.5 Germany 55.3 29.2 Hungary 16.6 16.9 Italy 22.7 53.8 Japan 58.7 18.6 Luxembourg 24.6 27.5 New Zealand 60.1 Poland 68.1 35.0 41.1 Sweden 79.0 47.4 United States 61.8 Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). U.S. Census Bureau An Aging World: 2001 97
  • 99. While female participation was increasing at younger ages, nearly Figure 10-3b. all developed countries experienced Labor Force Participation Rates for Women a decrease in elderly female labor Aged 60 to 64 in Developed Countries force participation between the (Percent) Early 1970s early 1970s and the late 1990s. Late 1990s Very small proportions of elderly 16.0 women currently are economically Australia 18.3 active in developed nations; among 13.2 Austria the 22 developed countries in 8.7 Appendix A, Table 10, only Japan, Belgium 7.6 Poland, and the United States have 6.7 elderly female participation rates 8.2 Bulgaria 4.7 above 4 percent.2 29.1 Canada 26.0 2 Rates for Norway and Ukraine in Appendix 18.2 Czech Republic Table 10 are about 9 percent, but these refer to 12.9 only a portion of their elderly female popula- 27.9 tions, i.e., women aged 65 to 74. France 15.2 Germany 17.7 12.7 Hungary 17.1 5.5 Italy 9.9 8.1 Japan 43.3 39.8 Luxembourg 12.0 11.7 15.5 New Zealand 32.5 Poland 51.1 19.2 25.7 Sweden 46.5 36.1 United States 38.8 Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). 98 An Aging World: 2001 U.S. Census Bureau
  • 100. WHY THE BIG DECREASE FOR OLDER MEN? Figure 10-3c. Labor Force Participation Rates for Women Several reasons may account for the Aged 65+ in Developed Countries sharp decline in activity rates of (Percent) Early 1970s older men in developed countries. Late 1990s An increase in societal wealth is 4.2 most likely the main reason for the Australia 3.1 drop in participation rates. A sec- 3.2 ondary reason may be that new Austria 2.0 technologies have changed the 2.2 industrial and occupational organi- Belgium 0.8 zation of many economies, and 1.7 Bulgaria generated the need for a recently 2.1 8.3 trained labor force. New technolo- Canada 3.4 gies can make the skills of older 5.2 workers obsolete and these workers Czech Republic 2.7 may choose to retire rather than 5.0 learn new skills (Ahituv and Zeira, France 2.0 2000; Bartel and Sicherman, 1993). 5.7 Germany In countries with persistently high 1.6 levels of unemployment, there may 5.8 Hungary be formal and informal pressures 1.6 3.2 on older workers to leave the labor Italy 1.7 force to make room for younger Japan 19.7 workers. Perhaps most importantly, 14.9 the growth and proliferation of Luxembourg 4.0 financial incentives for early retire- 0.6 ment have enabled many older 3.5 New Zealand workers to afford to stop working. 3.9 33.0 In much of Eastern Europe and the Poland 8.5 former Soviet Union, older workers 3.2 are choosing early retirement over Sweden unemployment as new market 10.0 mechanisms prompt firms to fire United States 8.9 redundant workers (Commander Note: Later data for Belgium 65+ refer to ages 65-69 and for Hungary to ages 65-74; data for Sweden and Yemtsov, 1997). 65+ are not reported by the ILO. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). U.S. Census Bureau An Aging World: 2001 99
  • 101. ELDERLY IN DEVELOPING COUNTRIES HAVE HIGH Figure 10-4a. PARTICIPATION RATES Labor Force Participation Rates for Men Aged 55 The proportion of economically to 59 in Developing Countries (Percent) Early 1970s active elderly men is high in devel- Late 1990s oping countries compared with more-industrialized nations. Not 80.4 Argentina surprisingly, many elderly people in 82.8 predominantly rural agrarian soci- Chile 82.6 83.4 eties work of necessity, while Ethiopia “retirement” may be a luxury 93.3 reserved for urban elites. In nations Mexico 86.2 as diverse as Bangladesh, 86.8 Indonesia, Jamaica, Mexico, Peru 92.8 85.6 Pakistan, and Zimbabwe, more than Singapore 73.9 50 percent of all elderly men are 77.9 considered to be economically South Korea 85.4 active. Economic activity rates of 81.0 older and elderly women also are Tunisia 86.0 78.4 higher in developing than in devel- Turkey 88.0 oped countries. Some national data 60.3 may understate the true economic Uruguay 81.2 activity of women, particularly in 89.3 developing countries where much Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the of the work that women engage in Yearbook of Labour Statistics). is not counted or captured in cen- suses and labor force surveys, or is not considered to be “economic.” Figure 10-4b. Many of the activities that older Labor Force Participation Rates for Men Aged 60 women are involved in, such as to 64 in Developing Countries (Percent) Early 1970s subsistence agriculture or house- Late 1990s hold industries, often are not well 57.2 documented by conventional data Argentina 63.2 collection methods (Hedman, 72.1 Chile Perucci, and Sundström, 1996). 69.2 Ethiopia 89.4 Mexico 81.5 77.3 Peru 83.9 72.5 Singapore 55.6 48.6 67.9 South Korea 65.5 Tunisia 66.5 54.1 Turkey 83.0 54.0 Uruguay 58.9 59.3 Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). 100 An Aging World: 2001 U.S. Census Bureau
  • 102. Data on economic activity rates Figure 10-4c. over time for developing countries Labor Force Participation Rates for Men do not show as clear a trend for Aged 65+ in Developing Countries older workers as seen in developed (Percent) Early 1970s countries. Although many devel- Late 1990s oping countries have experienced a decrease in economic activity of Argentina 29.1 27.6 older male workers, in most such Chile 42.4 countries the decrease is much 27.4 smaller than in developed coun- Ethiopia tries (Figure 10-4). Akin to the pat- 65.7 tern in developed countries, many Mexico 67.1 52.4 developing countries have wit- Peru 61.5 nessed an increase in labor force 41.1 participation rates for women aged Singapore 31.9 55 to 64 (Figure 10-5). Unlike the 21.7 pattern in developed nations, sev- South Korea 35.1 40.2 eral developing countries also have Tunisia 38.0 experienced increases in participa- 34.0 tion for women aged 65 and older. Turkey 67.8 Because of the problems with sta- 33.6 tistics on female economic activity Uruguay 20.9 19.4 mentioned above, these changes could reflect “real” increases in Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the activity rates as well as improve- Yearbook of Labour Statistics). ments in data collection. Figure 10-5a. Labor Force Participation Rates for Women Aged 55 to 59 in Developing Countries (Percent) Early 1970s Late 1990s Argentina 16.2 35.4 Chile 15.5 32.4 Ethiopia 58.6 Mexico 15.4 32.4 Peru 16.1 47.5 Singapore 16.2 28.4 South Korea 39.1 51.2 Tunisia 11.3 12.2 Turkey 50.0 30.4 Uruguay 21.7 41.0 Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). U.S. Census Bureau An Aging World: 2001 101
  • 103. AGRICULTURE STILL IMPORTANT SOURCE OF Figure 10-5b. EMPLOYMENT FOR ELDERLY Labor Force Participation Rates for Women Just as labor force participation Aged 60 to 64 in Developing Countries rates of older workers vary among (Percent) Early 1970s countries, so do levels of concentra- Late 1990s tion in various occupations. Argentina 10.3 Economies in developed countries 22.6 have shifted from agriculture and Chile 11.1 21.0 heavy industries toward services Ethiopia and light industries, which is a shift 50.0 from physically demanding and Mexico 14.4 sometimes hazardous jobs to work 25.9 Peru 13.4 which requires less brawn and dif- 38.2 ferent technical skills. This shift 13.4 Singapore may benefit older workers insofar 14.9 as jobs requiring mental ability 26.9 South Korea rather than physical strength may 46.3 Tunisia 8.6 enable them to remain active 7.7 longer. Conversely, the shift could 47.6 Turkey be detrimental to older workers if 23.4 the new jobs require skills or train- Uruguay 12.2 ing which older workers may not 23.9 have or easily acquire. Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). Figure 10-5c. Labor Force Participation Rates for Women Aged 65+ in Developing Countries (Percent) Early 1970s Late 1990s Argentina 4.7 8.9 Chile 6.5 6.5 Ethiopia 27.5 Mexico 11.8 14.6 Peru 8.5 19.2 Singapore 6.5 5.2 South Korea 10.6 21.4 Tunisia 4.8 3.5 Turkey 35.1 13.3 Uruguay 3.6 6.7 Note: Data for Ethiopia in the early 1970s are not available. Sources: U.S. Census Bureau, 2000a and International Labour Office (various issues of the Yearbook of Labour Statistics). 102 An Aging World: 2001 U.S. Census Bureau
  • 104. Not surprisingly, agriculture is by Figure 10-6. far the most common occupation Percent of Elderly Workers in Agriculture for older and elderly workers in Male most developing countries (Figure Female 10-6). And despite the worldwide 88.3 trend away from employment in China, 1990 93.2 agriculture, this sector was still an Ecuador, 1990 57.0 important source of employment in 30.0 many developed countries during 30.7 Japan, 1995 the 1970s and 1980s. Even in the 34.7 21.2 1990s, a nontrivial proportion of New Zealand, 1991 15.3 economically active elderly in some Philippines, 1990 76.4 developed countries worked in the 32.2 agricultural sector. In 1995 in 75.7 Turkey, 1990 Japan, 31 percent of elderly men 95.6 11.4 and 35 percent of elderly women United States, 1998 3.2 were involved in agriculture. Aggregate data from the early Note: Data for New Zealand refer to ages 60 and over. Source: National sources. 1990s for 12 European Union nations showed that agriculture Table 10-1. Older Workers (55 and Over) per 100 Younger Workers (Under 55) in Selected Job Sectors: 1998 Goods-producing sector Service sector Agriculture, hunting and Manu- Personal Social Total forestry facturing Total services services OECD average 15 39 10 12 11 12 Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 28 6 7 7 7 Belgium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 21 5 7 8 7 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 27 10 10 9 10 Czech Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 12 9 10 9 13 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 34 11 12 15 13 Finland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 24 9 9 8 10 France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 20 6 8 10 8 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 30 14 15 16 16 Greece. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 68 11 11 12 9 Ireland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 45 7 10 10 13 Italy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 33 8 13 12 13 Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 131 8 13 (NA) (NA) Luxembourg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 19 6 7 5 10 Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 28 8 11 13 7 Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 21 8 7 6 8 New Zealand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 26 7 7 - 8 Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 36 16 15 9 18 Portugal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 112 10 15 18 12 Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 36 11 12 14 13 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 50 17 19 15 21 Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 39 18 17 20 18 United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 32 14 13 14 14 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 25 13 15 11 16 - Represents zero. NA Not available. Source: Excerpted from Organization for Economic Co-Operation and Development (Employment Outlook 2000). U.S. Census Bureau An Aging World: 2001 103
  • 105. continued to employ a dispropor- tionate share of older relative to Figure 10-7. younger workers (Eurostat, 1993a). Percent Distribution of Elderly Workers by More than 20 percent of economi- Occupation in the United States: 1998 Male Female cally active men and women aged 60 and older in these countries Managerial/ 33.8 worked in agriculture compared Professional/ 24.7 Technical with about 5 percent of people 14.8 Sales aged 14 to 59. More recent infor- 16.7 mation for 23 OECD countries 16.9 Clerical/Service shows that the ratio of older (55+) 47.8 to nonolder (54 and under) workers Production 23.2 is generally much higher in agricul- 7.6 ture, hunting, and forestry than in 11.4 Agriculture any other goods-producing or serv- 3.2 ice sector (Table 10-1). Source: Bureau of Labor Statistics, unpublished tabulations from the Current Population Survey, average annual data, 1998. NEARLY TWO-THIRDS OF ELDERLY FEMALE U.S. WORKERS IN SERVICE AND of retirement itself. During periods OLDER WOMEN MORE LIKELY SALES of economic contraction in highly THAN OLDER MEN TO WORK industrialized nations, governments PART TIME Figure 10-7, which presents the occupational distribution of elderly may actively encourage older work- Some older workers use part-time male and female workers in the ers to cease active employment at work as a gradual transition to United States, shows distinct differ- relatively young ages. On the other retirement (Walker, 1999). Part-time ences by gender. In 1998, almost hand, when the labor market is work is an option that may appeal half of elderly working women were tight, governments may look for to older workers by enabling them employed in clerical or service jobs methods to entice older workers to to remain active in the labor force compared with only 17 percent of remain in the labor force or re-enter while also pursuing leisure activities elderly men. A majority of working the labor force. (Quinn and Kozy, 1996). Data for elderly men held either manageri- working men aged 60-64 in nine In developed countries, retirement al/professional/technical positions developed countries show large dif- from the workforce was an event (34 percent) or production jobs ferences in the prevalence of part- that occurred almost exclusively at (23 percent). Corresponding figures time employment, ranging from a regulated age until the 1950s, for elderly women were 25 percent less than 8 percent in Italy and with little possibility of receiving a and 8 percent. Unlike the situation Germany to more than 35 percent pension prior to that age (Tracy, in some developed countries, only in Sweden and the Netherlands 1979). Since then, countries have 11 and 3 percent of active elderly (OECD, 2000). Available data from adopted a wide range of approach- U.S. men and women, respectively, developed countries suggest that es to providing old-age security, worked in agriculture. older working women are much and different potential routes have more likely than older men to be emerged for people making the BRIDGES TO RETIREMENT involved in part-time work (Figure transition from labor force participa- Just as the propensity to work at 10-8). In Australia, three-fourths of tion to retirement. Some of these older ages varies considerably from elderly women who were economi- different routes are working part country to country, so too do pat- cally active in 1999 worked part time, leaving career jobs for transi- terns of retirement and the concept time, compared to fewer than half tion jobs, or leaving the labor force of economically active elderly men. because of a disability. 104 An Aging World: 2001 U.S. Census Bureau
  • 106. data availability, the effects of the Figure 10-8. business cycle, and differences in Percent of Economically Active Elderly Population definitions across countries.3 In Working Part Time in Australia, countries with available data, Canada, and France Male unemployment rates for all age Female groups commonly were higher in 74.2 1999 than in 1980. Gender differ- ences in unemployment rates at older ages are not consistent; in some countries, men have higher 53.2 rates and in others the reverse is true. When unemployment rates 44.1 for ages 55 to 64 are disaggregat- 38.6 ed, rates for the age group 55 to 33.2 59 tend to be somewhat higher than for the age group 60 to 64, 20.7 perhaps because people in the older age group may opt to retire if possible rather than be unemployed. Although the unemployment rate may be lower for older than for Australia 1999 Canada 1991 France 1995 younger workers, older people who Note: Data for France refer to ages 60 and over. are unemployed tend to remain Source: National sources. unemployed longer than their younger counterparts. In several In the 15 European Union countries gradual retirement is still relatively OECD countries, well over half of as a whole, 41 percent of working uncommon in industrialized nations. unemployed people aged 55 and women aged 55-64 were in part- The strongest tendency toward part- over had been unemployed continu- time positions in 1998, compared time work was seen in Japan, ously for more than 1 year. In most with just 8 percent of working men Sweden, and the United States. OECD countries, the proportion of in that age group (Eurostat, 2000). long-term unemployed people aged UNEMPLOYMENT LOW AMONG 55 and older is much higher than The rate of part-time work for peo- THE ELDERLY among younger age groups. A simi- ple nearing retirement generally was The elderly typically have low lev- lar pattern is seen in some Eastern increasing with time in the late els of unemployment compared to European nations. In Bulgaria in 1980s/early 1990s (Eurostat, younger workers. In developed 1995, 74 percent of unemployed 1993b). Even though percentages countries, unemployment rates for men aged 50 to 59 and 78 percent of older workers who work part the elderly frequently are less than of unemployed women aged 50 to time may be substantial, a recent 5 percent (OECD Labour Force 54 had been without work for more OECD (2000) analysis notes that Statistics, 2000). However, people than 1 year (European Commission, these percentages often represent aged 55 to 64 often have unem- 1995). only a small fraction of the total ployment rates higher than or simi- older population, since many people lar to rates for people aged 25 to have retired by age 60. Looking at 3 And, in many developing countries, the 54 (Figure 10-9). Establishing a older male cohorts as they aged in lack of programs to provide monetary support time trend in unemployment rates during unemployment means that most people the 1990s, the study concludes that cannot “afford” to be unemployed. for older people is hindered by U.S. Census Bureau An Aging World: 2001 105
  • 107. Figure 10-9. 1980 Unemployment for Three Age Groups: 1980, 1990, and 1999 1990 (Percent) 1999 Canada Germany Male Female Male Female 15.5 12.8 8.7 8.1 7.6 7.2 7.2 7.0 7.3 6.8 6.9 6.7 6.2 6.5 5.7 5.8 5.9 5.1 5.1 4.9 4.4 3.6 3.8 2.7 2.9 1.8 1.5 2.0 2.2 0.8 0.6 0.4 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ United States Sweden Male Female Male Female 7.3 6.5 6.0 5.9 5.9 5.1 4.6 4.6 3.8 3.4 3.1 3.0 3.0 3.4 3.3 3.1 3.1 3.2 3.0 2.7 2.8 2.6 1.6 1.3 1.7 1.7 1.6 1.2 1.3 1.0 1.2 1.3 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ Japan France Male Female Male Female 12.6 10.7 9.0 8.7 8.7 7.6 6.7 6.3 6.2 5.9 6.0 4.8 4.4 3.7 3.7 3.4 3.3 2.9 2.8 2.2 2.0 2.1 1.9 1.5 1.4 1.4 1.5 1.2 1.4 0.5 0.5 0.5 0.6 0.8 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ 25-54 55-64 65+ Age group Age group Notes: Unemployment rates for people aged 65+ in Canada, Germany, Japan, and Sweden were reported to be zero in certain years. Latest data for Canada refer to 1998. 1980 data for Germany refer to West Germany. Source: Organization for Economic Co-Operation and Development, 2000 (Labor Force Statistics 1979-1999). 106 An Aging World: 2001 U.S. Census Bureau
  • 108. DISCOURAGEMENT REDUCES there is no work available or discouragement of older workers is NUMBER OF OLDER WORKERS because they do not know where to thought to be related to changes in The definition of discouraged work- look. Workers who become dis- occupational structure and the sub- er differs somewhat from country to couraged from actively seeking sequent need for a more-educated country, but the basic concept work are no longer considered part workforce which favor younger refers to people who are no longer of the economically active popula- over older workers. looking for work because they think tion. In some countries, Cross-national data on discouraged workers are fairly sparse. One com- parison of 13 countries for 1993 Figure 10-10. indicates that older workers make Older Share of Economically Active Population and up a disproportionate share of all Discouraged Worker Population: 1993 discouraged workers, except in (In percent) Economically Sweden. Illustrative data for six of Australia active people these countries (Figure 10-10) show 11 aged 55-64 as Male that while people aged 55 to 64 55 a percent of 7 all economically account for a small proportion of all Female active people economically active people, they 29 aged 25-64 Canada account for a much larger propor- Discouraged tion of all discouraged workers, 11 Male workers aged 50 especially in the United Kingdom 55-64 as a percent 9 where more than two-thirds of all Female of all discouraged 42 workers aged discouraged male workers were Denmark 25-64 aged 55 to 64. In countries with 13 data over time, discouraged older Male 44 11 workers were more numerous in Female 53 the early 1990s than in the early Mexico 1980s. Discouragement seems to 12 be more permanent among older Male 36 workers, as they are less likely to 8 Female re-enter the labor force than their 26 younger counterparts. Survey data United Kingdom for 1990 for Belgium and France 14 Male 69 show that more than half of men 12 aged 55 to 59 who had lost their Female 53 jobs in the 3 years preceding the United States surveys were no longer in the labor 12 force (OECD, 1995). The correspon- Males 26 ding figure for all workers was clos- 11 Females 24 er to one-quarter. Because of the difficulties older people face in Note: Data for Norway include persons aged 64-74; data for the United States include people aged 65 and over. obtaining a new job, discourage- Sources: Organization for Economic Co-Operation and Development, Employment Outlook 1995 and International Labour Office Yearbook of Labour Statistics, 1994. ment often becomes a transition from unemployment to retirement. U.S. Census Bureau An Aging World: 2001 107
  • 109. INVALIDITY AND DISABILITY PROGRAMS MAY BE AVENUES Figure 10-11. TO RETIREMENT Percent of Retired Men Aged 55 to 64 Who Left Last Another path to retirement for older Job Due to Own Illness or Disability: 1995-96 workers has been disability pro- grams. In Europe during the last Switzerland 28.9 three decades, economic recessions Finland 25.0 and high unemployment led some Germany 22.9 governments (e.g., Germany, the United Kingdom 22.8 Netherlands, Sweden) to encourage Spain 18.3 retirement by means of public measures such as disability Luxembourg 16.6 schemes and long-term sickness Netherlands 15.6 benefits. In many countries in the Ireland 15.1 late 1980s and early 1990s, disabil- Denmark 9.5 ity pensioners made up the largest Belgium 7.7 proportion of all early pensioners (OECD, 1992). Data for 1990 France 7.3 showed that the proportion of older Sweden 7.0 people receiving an invalidity bene- Italy 5.2 fit could be very large, e.g., nearly Greece 4.1 one-third of all people aged 60 to Austria 2.6 64 in Finland and Sweden, and nearly half of the same age group Portugal 2.1 in Norway (OECD, 1995). While Source: Excerpted from Organization for Economic Co-Operation and Development numerous nations have modified or (Employment Outlook, 2000). revamped their disability/invalidity programs during the last decade, it appears likely that the varying fully entitled to public pension awarded. In spite of the lowering national provisions of such pro- benefits. These reductions were of statutory retirement ages, the grams have an impact on retirement propelled by a combination of fac- actual average age of retirement is patterns. Comparative data for 16 tors including general economic lower than the statutory age in a European countries in the mid- conditions, changes in welfare phi- large majority of industrialized 1990s (Figure 10-11) show that the losophy, and private pension countries. Of the 24 countries in percentage of older retired men trends. The proliferation of early Table 10-2, the actual age exceeds who retired due to their own illness retirement schemes has increased the standard age only in Greece, or disability ranged from 2 percent the number and usually the propor- Japan, and Turkey, and also in in Portugal to 29 percent in tion of older workers who avail Iceland for men and in Italy for Switzerland. themselves of such programs (Tracy women. In several countries (e.g., and Adams, 1989). Austria, Belgium, and Finland), the ACTUAL RETIREMENT AGE average man retires 6 years or more OFTEN LOWER THAN One important issue for policymak- before the standard retirement age. STATUTORY AGE ers and pension funds is the rela- Differences are often greater for tionship between the standard Over several decades, many indus- women, approaching 10 years in (statutory) retirement age and “actu- trialized nations lowered the stan- Luxembourg and the Netherlands. al” retirement age, the average age dard age at which people become at which retirement benefits are 108 An Aging World: 2001 U.S. Census Bureau
  • 110. Table 10-2. income than did earlier cohorts of Standard and Actual Retirement Age in 24 Countries: elderly. 1995 Some nations have raised (or are Male Female Country considering an increase in) statutory Standard Actual Standard Actual retirement age as one means of off- Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 61.8 60 57.2 setting the fiscal pressures of popu- Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 58.6 60 56.5 lation aging,4 in addition to foster- Belgium . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 57.6 60 54.1 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 62.3 65 58.8 ing policies that encourage labor Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . 67 62.7 67 59.4 force participation at older ages. Finland . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 59.0 65 58.9 The effect of such actions is not yet France . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 59.2 60 58.3 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . 65 60.5 65 58.4 certain. Data for nine OECD coun- Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 62.3 57 60.3 tries from 1975 to 1990 reveal a Iceland . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 69.5 67 66.0 Ireland . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 63.4 66 60.1 general downward trend in actual Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 60.6 57 57.2 retirement age during the period Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 66.5 58 63.7 1975-1990, with an apparent level- Luxembourg . . . . . . . . . . . . . . . . . . . . . . . 65 58.4 65 55.4 Netherlands. . . . . . . . . . . . . . . . . . . . . . . . 65 58.8 65 55.3 ing off in the latter part of the peri- New Zealand. . . . . . . . . . . . . . . . . . . . . . . 62 62.0 62 58.6 od. Gendell’s (1998) analysis of Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 63.8 67 62.0 Portugal . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 63.6 62.5 60.8 Germany, Japan, Sweden, and the Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 61.4 65 58.9 United States generally supports Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 63.3 65 62.1 Switzerland . . . . . . . . . . . . . . . . . . . . . . . . 65 64.6 62 60.6 this picture (Figure 10-12). Turkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 63.6 55 66.6 However, several studies have United Kingdom . . . . . . . . . . . . . . . . . . . . 65 62.7 60 59.7 argued that the early retirement United States . . . . . . . . . . . . . . . . . . . . . . 65 63.6 65 61.6 trend in the United States has Note: The standard age of retirement (also called the statutory age) refers to the age stopped (Smeeding and Quinn, of eligibility for full public pension benefits. The actual age reflects the estimated average age of transition to inactivity among older workers. 1997; Burkhauser and Quinn, 1997; Source: Organization for Economic Co-Operation and Development, 1998 (Ageing Quinn, 1997). Furthermore, a recent Working Paper 1.4). OECD (2000) analysis of employ- ment rates notes that the rates for men aged 55 to 59 and 60 to 64 TREND IN EARLY RETIREMENT health of older workers were two have increased slightly in the late MAY BE CHANGING other factors said to have increased 1990s in both the United States and The downward shift in the statutory early retirement. More recent the Netherlands, and have stopped age at retirement during the 1970s research, however, discounts the declining in Canada, Germany, and 1980s in developed countries importance of these factors (Levine Finland, Japan, Sweden and the was accompanied by an increase in and Mitchell, 1993; Blondal and United Kingdom. The OECD study the number of public early retire- Scarpetta, 1998; Fronstin, 1999) suggests that this change is related ment programs and a correspon- and points instead to changes in to the secular economic upturn in ding increase in the number of social security/private pension pro- the latter 1990s. retirees leaving the labor force prior visions as well as to improved eco- to the statutory age. Some coun- nomic status of older workers and 4 In the United States, for example, the tries promoted early retirement as a increases in wealth overall. As Social Security system was revised in 1983 to means of offsetting persistently Ruggles (1992) has noted in the establish higher statutory retirement ages for people born after 1937 (i.e., who reach age 65 high levels of unemployment. In context of the United States, com- after the year 2002). An individual’s retirement Denmark, for example, a voluntary parisons of today’s elderly with the age is linked to year of birth; beginning in the year 2003, the “normal” retirement age of 65 early retirement scheme was con- elderly in previous decades suggest will edge higher in small increments until reach- great increases in economic status. ing 67 years in the year 2025 (Robertson, structed to encourage older work- 1992). Germany’s 1992 Pension Act also pro- ers to leave the labor market People entering the ranks of the vides for a progressive increase in pensionable elderly have higher educational age beginning at the turn of the century. (Petersen, 1991). Mandatory retire- ment practices and worsening attainment, higher-paid employment histories, and higher average U.S. Census Bureau An Aging World: 2001 109
  • 111. Figure 10-12. Average Age at Labor Force Exit in Four Countries: Late 1960s to Early 1990s Germany Sweden Age Age 70 70 65 Male 65 Male Female Female 60 60 55 55 0 0 1965 1970 1975 1980 1985 1990 1995 1965 1970 1975 1980 1985 1990 1995 Japan United States Age Age 70 70 Male Female 65 65 Male Female 60 60 55 55 0 0 1965 1970 1975 1980 1985 1990 1995 1965 1970 1975 1980 1985 1990 1995 Source: Gendell, 1998. 110 An Aging World: 2001 U.S. Census Bureau
  • 112. ADULTS SPENDING GREATER years before entry into the labor number of years in retirement had PORTION OF LIFE IN market (primarily spent in school); jumped to 12. Unlike the trend for RETIREMENT years not in work due to unemploy- men, the average number of years Gains in life expectancy during the ment and/or economic inactivity; in employment for women has been twentieth century have intersected years in the labor force; and years increasing, reflecting the temporal with declining retirement ages to in retirement. Figure 10-13 shows changes in female labor force par- produce an increase in the propor- that in 1960, men on average could ticipation described earlier. At the tion of an individual’s life spent in expect to spend 46 years in the same time, the amount of time retirement. The OECD, using an labor force and a little more than women live after reaching retire- average of unweighted data for 15 1 year in retirement. By 1995, the ment age increased greatly, from member countries, has decomposed number of years in the labor force 9 years in 1960 to more than the life course into four states: had decreased to 37 while the 21 years in 1995. Figure 10-13. Decomposition of the Life Course, OECD Average: 1960 to 1995 Male Female Age Age 90 90 Life 80 80 expectancy Life expectancy Years in retirement 70 70 Years in retirement Age of 60 retirement 60 Age of retirement 50 50 Years in employment Years in employment 40 40 30 30 Years not in work 20 Years not in work 20 Age of Age of entry in entry in the labor the labor 10 Years before the labor market market 10 Years before the labor market market 0 0 1960 1970 1980 1990 1995 1960 1970 1980 1990 1995 Note: Based on an unweighted average of data for 15 member countries, using average life expectancies and labor force patterns as they existed for the years shown. These graphs are illustrative of overall trends, and should not be construed as representing the experience of any particular age cohort. Source: Organization for Economic Co-Operation and Development, 1998b. U.S. Census Bureau An Aging World: 2001 111
  • 113. PUBLIC PENSION SYSTEM eventual benefit payment rate for benefits may be initially obtained at PROVISIONS SOMETIMES people who retire after age 60. In age 62, but the benefit payment INDUCE EARLY RETIREMENT the United States, social security rate is less than if a worker retires Research has begun to consider national differences among labor force participation at older ages as Figure 10-14. a consequence (intended or unin- Implicit Tax Rate on Work in France and tended) of retirement provisions the United States: Circa 1995 and/or tax policy. In some coun- Percent tries, retirement benefit payments 90 are increased for people who post- pone their retirement beyond the France allowable early retirement age. In 60 other countries, there is no future benefit to be gained by postponing retirement. One synthesis of vari- 30 ous studies in industrialized nations (Gruber and Wise, 1999) looked at the “implicit tax on work,” a con- 0 cept which contrasts the longer United States stream of future benefit payments that a worker would receive by -30 retiring at an early age versus the shorter stream of future payments that a worker might receive by -60 delaying his/her retirement. In 56 58 60 62 64 66 68 70 Age France, for example, social security Source: Gruber and Wise, 1999. Reproduced with permission from graphs in "Social Security and benefits are first available at age Retirement Around the World," Research Highlights in the Demography and Economics of Aging, 60, and there is no increase in the Issue No. 2, June 1998. 112 An Aging World: 2001 U.S. Census Bureau
  • 114. later, e.g., at age 65. Figure 10-14 States. The age-specific retirement surprisingly corresponds to the compares the implicit tax rates on age in France shows a steep jump large implicit tax rise at that age. work in France and the United in retirement at age 60, which not The same study also considered a “tax force to retire,” defined as the total of the annual tax rates on Figure 10-15. work between the ages of 55 and Tax Rates and Unused Capacity in 11 Developed Countries: Circa 1995 69. Plotting this variable against a measure of unused productive Unused productive capacity (55-65), in percent capacity (simply, the percentage of 80 people aged 55 to 65 who were not 70 Belgium working) reveals a strong cross- Italy national relationship between the France 60 Netherlands two (Figure 10-15). This finding United Kingdom suggests that the financial structure Germany of national social security systems 50 Spain Canada may reward early retirement, and 40 United States that attempts to encourage Sweden increased labor force participation 30 at older ages may be largely contin- Japan gent upon policy changes in these 20 systems. This study also highlights the potential power of focusing on 10 system design features, and stands as a powerful example of the 0 importance of cross-national 0 1 2 3 4 5 6 7 8 9 10 research on aging-related issues. Tax force to retire (55-69) Source: Gruber and Wise, 1999. Reproduced with permission from graphs in "Social Security and Retirement Around the World," Research Highlights in the Demography and Economics of Aging, Issue No. 2, June 1998. U.S. Census Bureau An Aging World: 2001 113
  • 115. CHAPTER 11. Pensions and Income Security Public pensions have become the retirement age. This example is tional transfers to provide adequate financial lifeline of the elderly in based on average ages of retire- retirement income for older citi- many societies. While some ment for employees in 1995 esti- zens. The number of countries European public pension systems mated by the Organization for with an old age/disability/survivors date back to the end of the nine- Economic Co-Operation and program increased from 33 in 1940 teenth century, current systems are Development (OECD) and popula- to 167 in 1999 (Figure 11-2). The the result of changes instituted tion age/sex structures for 2000 World Bank (1994) has estimated largely after World War II. The most and 2030 estimated and projected that formal public programs provide obvious and, to governments, most by the U.S. Census Bureau. The coverage for approximately 30 per- worrisome consequence of project- numerator of the ratio comprises all cent of the world’s older (aged 60 ed population aging will be an people at or over the average age and over) population, with some 40 increase in budgetary outlays in the of retirement in each country, and percent of the world’s working-age form of old-age pension payments, the denominator all people between population making contributions especially in those countries in the age of 20 and the average toward that support. which public pensions are predomi- retirement age, assuming no nately financed on a pay-as-you-go change in the average age of retire- LABOR FORCE PENSION COVERAGE VARIES FROM basis. Increases in migration also ment between 2000 and 2030. The UNIVERSAL TO NIL are prompting governmental con- ratio increases notably over time in cern about the “exporting” of cash all cases, and more than doubles Mandatory old-age pension plans benefits to retirees in other coun- for men in the Netherlands. now cover more than 90 percent of tries (Bolderson and Gains, 1994). the labor force in most developed Many nations, both developed and PUBLIC OLD-AGE SECURITY countries. Governments are respon- SYSTEMS PROLIFERATING sible for mandating, financing, man- developing, are now reconsidering their existing old-age security sys- Since the Second World War, public aging, and insuring public pensions. tems, often with an eye toward pension plans have played an Public pension plans usually offer introducing or strengthening private increasingly important role in pro- defined benefits that are not tied to pension schemes. viding retirement income to older individual contributions, but rather, people. Old-age pension schemes are financed by payroll taxes. This DEMOGRAPHIC CHANGE have become social institutions in arrangement is commonly referred ALONE MAY DOUBLE to as a “pay-as-you-go” system inso- many if not most countries through- RETIREE/WORKER RATIO out the world. The goal of most far as current revenues (taxes on The potential effect of demographic public old-age pension schemes is working adults) are used to finance change on future retired popula- to provide all qualifying individuals the pension payments of people tion/worker ratios, holding other with an income stream during their who are retired from the labor force factors constant, may be approxi- later years, income which is: 1) (Mortensen, 1992). mated in various ways. The most continuous; 2) adequate; 3) con- Most pay-as-you-go systems in commonly used indicator, as dis- stant, in terms of purchasing industrialized countries initially cussed in Chapter 8, is an elderly power; and 4) capable of maintain- promised generous benefits. support ratio which contrasts one ing the socioeconomic position of These pension programs, at their population segment (people aged the retired in relation to that of the inception, were based on a small 65 and over) to another (people active population (Nektarios, 1982). number of pensioners relative to a aged 20 to 64). One variation on The major impetus for development large number of contributors this theme, shown in Figure 11-1 of public pension systems, particu- (workers). As systems matured, for 10 developed countries, allows larly in industrialized countries, was ratios of pensioners to contributors for national differences in average the inability of private intergenera- grew and in some countries U.S. Census Bureau An Aging World: 2001 115
  • 116. became unsustainable, particularly during periods of economic stag- Figure 11-1. nation. One result of such Ratio of Retirement-Age to Working-Age Population in Ten Countries: 2000 and 2030 changes was the development of (Average retirement age shown in parentheses) private pension systems to com- 2000 plement public pension systems Male 2030 (Fox, 1994). Other measures taken 23 or considered have included Australia (61.8) 43 increasing worker contribution 24 rates, restructuring or reducing Denmark (62.7) 44 benefits, and raising the standard 40 age of retirement (ISSA, 1993; Finland (59.0) 75 Holtzmann and Stiglitz, 2001). 34 France (59.2) 57 In developing countries, public pen- sion systems typically cover a much 35 Italy (60.6) 63 smaller fraction of workers than in 21 industrialized nations (Figure 11-3). Japan (66.5) 41 Even economically vibrant societies such as Hong Kong and Thailand 30 Netherlands (58.8) 63 offer no publicly supported, compre- hensive retirement pension scheme 30 Spain (61.4) 54 (Bartlett and Phillips, 1995; Domingo, 1995). In many cases, 26 United Kingdom (62.7) 45 coverage in developing countries is restricted to certain categories of 20 United States (63.6) 37 workers such as civil servants, mili- tary personnel, and employees in Female the formal economic sector. Rural, 40 predominantly agricultural workers Australia (57.2) 72 have little or no pension coverage in 42 much of the developing world, Denmark (59.4) 69 although some governments have 46 taken steps to address this situa- Finland (58.9) 81 tion. Each state in India, for exam- 48 ple, has implemented an old age France (58.3) 80 pension scheme for destitute people 58 with no source of income and no Italy (57.2) 105 family support (Kumar, 1998). While 35 pension amounts are minimal and Japan (63.7) 65 coverage far from universal, the for- 50 mal institution of such a system Netherlands (55.3) 96 affords a nation a foundation upon 46 which to expand future coverage. Spain (58.9) 85 44 United Kingdom (59.7) 71 31 United States (61.6) 52 Note: Ratios represent the number of persons at or above average retirement age per 100 persons between age 20 and the average retirement age in 1995. Each national average is shown in parentheses. Sources: OECD, 1998 (Aging Working Paper 1.4) and U.S. Census Bureau, 2000a. 116 An Aging World: 2001 U.S. Census Bureau
  • 117. Informal (usually family) systems Figure 11-2. provide the bulk of social support Number of Countries With Public Old-Age/Disability/ for older individuals in many coun- Survivors Program: 1940 to 1999 tries, particularly in Africa and South Asia. As economies expand and nations urbanize, informal sup- 167 port systems such as extended fam- ily care and mutual aid societies have tended to weaken. A major 135 challenge for governments in devel- 123 oping nations is to effect the expan- sion of formal-system coverage 97 (especially in rural areas) while maintaining support for extant informal mechanisms. 58 HOW GENEROUS ARE PUBLIC 44 PENSIONS? 33 The “value” of pensions can be con- strued and measured in different ways, depending on how many and 1940 1949 1958 1969 1979 1989 1999 which people in a given household rely on pension income, the taxable Source: U.S. Social Security Administration, 1999. status of such income, the type of job a retiree was engaged in, the level of pension income in a given Figure 11-3. society vis-a-vis other benefits such Percent of Labor Force Covered by Public Old-Age as universal health care, and so Pension Program: Circa 1995 forth. The concept of “replacement rate” is often used as a measure of Nigeria 1 how much of a person’s pre- Zambia 10 retirement income is supplied by her/his pension. MacKellar and Morocco 21 McGreevey (1999) note that, in South Korea 30 industrialized countries, the aver- Jordan 40 age pension rose from 14 percent Brazil 50 of the average wage in 1930 to 55 Poland 66 percent in 1980. A comparison of Russia 72 gross income replacement of social Germany 94 security and other compulsory retirement pension programs in Italy 96 12 European nations circa 1990 Source: World Bank, 1998. (International Benefits Information U.S. Census Bureau An Aging World: 2001 117
  • 118. Service, 1993) revealed that replace- ment rates ranged from 46 percent Figure 11-4. to 102 percent, based on average Expected Old-Age Public Pension Replacement annual pay for a manufacturing Rate in 26 Countries: 1995 worker with dependent spouse. (Percent) For reasons mentioned above, there is no single replacement rate in any Australia 41 national retirement program, and Austria 80 cross-national comparisons there- Belgium 68 fore are difficult. For comparative Canada 52 purposes, however, the OECD has Czech Republic 53 constructed, for 1995, a synthetic Denmark 56 indicator of the expected gross Finland 60 replacement rate (as a percent of France 65 earnings) for a 55-year-old individ- Germany 55 Greece 120 ual who retires at the standard age Hungary 55 of entitlement to a public pension. Iceland 93 This indicator takes account of two Ireland 40 earnings levels (average and two- Italy 80 thirds of average) and two types of Japan 52 households (single earner and Luxembourg 93 worker with a dependent spouse). Netherlands 46 Pensions in some countries can be New Zealand 61 expected to replace a large percent- Norway 60 age of earnings, and even to match Poland 54 or exceed the latter in Greece and Portugal 83 Spain. At the other end of the spec- Spain 100 trum are Australia and Ireland, Sweden 74 Switzerland 49 where the public-pension replace- United Kingdom 50 ment rates are on the order of United States 56 40 percent. For the majority of countries examined, the expected replacement rates are about one- Note: Synthetic indicator based on different earning levels and household types; see text and source for more detail. half to two-thirds of pre-retirement Source: OECD, 1998 (Aging Working Paper 1.4). income (Figure 11-4). 118 An Aging World: 2001 U.S. Census Bureau
  • 119. PUBLIC PENSIONS ABSORB Figure 11-5. ONE-SEVENTH OF GDP IN SOME Public Pension Expenditure as a Percent of Gross COUNTRIES Domestic Product in 25 Countries: Circa 1996 The cost of public pensions general- ly is greatest among industrial Italy, 1995 15.0 nations, most of which have pay-as- Uruguay, 1996 15.0 France, 1995 13.3 you-go systems. Pension expendi- Germany, 1995 12.0 ture had, on average, come to Croatia, 1997 11.6 exceed 9 percent of gross domestic United Kingdom, 1995 10.2 product (GDP) in OECD nations in Hungary, 1996 9.7 United States, 1995 7.2 the early 1990s, and represented 8 Japan, 1995 6.6 percent of GDP in Eastern Europe. Cyprus, 1996 6.4 Between 1960 and 1990, one- Israel, 1996 5.9 quarter of the increase in total pub- Russia, 1996 5.7 Kazakhstan, 1997 5.0 lic expenditure in OECD countries Brazil, 1996 4.7 was growth in pension expenditure; Australia, 1995 4.6 on average, the latter grew twice as Turkey, 1995 3.7 fast as did GDP By 1996, public . China, 1996 2.7 Sri Lanka, 1996 pension spending in Italy and 2.4 Algeria, 1997 2.1 Uruguay had reached 15 percent of Vietnam, 1998 1.6 GDP (Palacios and Pallares-Miralles, Ecuador, 1997 1.0 2000). Expenditure levels typically Uganda, 1997 0.8 Togo, 1997 are much lower in most developing 0.6 Mexico, 1996 0.4 countries (Figure 11-5), where rela- Cote d'Ivoire, 1997 0.3 tively younger populations and Source: Palacios and Pallares-Miralles, 2000. smaller pension programs do not yet place large demands on GDP . ADMINISTRATIVE COSTS OF PUBLIC PENSION SYSTEMS Figure 11-6. HIGH IN SOME DEVELOPING Net Administrative Fees as a Percent of COUNTRIES Total Contributions in Eight Latin American The cost of administering a public Individual-Account Systems: 1999 pension scheme is an important fac- tor in the scheme’s overall efficacy. Argentina 23.0 In many developing countries, Mexico 22.1 administrative costs as a percent of Peru 19.0 total old-age benefits have been El Salvador 17.6 high (e.g., 10-15 percent in Brazil Chile and Turkey) relative to the devel- 15.6 oped world — administrative costs Uruguay 14.3 as a percent of old-age benefits are Colombia 14.1 less than 2 percent in most OECD Bolivia 5.5 countries. In many developed countries, the cost/benefit ratio Source: James, Smalhout and Vittas, 2001. declined in the 1970s and early U.S. Census Bureau An Aging World: 2001 119
  • 120. 1980s (Estrin, 1988) as a result of: 1) government austerity programs Figure 11-7. that helped contain administrative Percent of Pensioners With Income From costs; 2) increases in total benefit Employer-Provided Pensions: Late 1990s expenditures, reflecting not only Male the maturation and/or expansion of Female 20 Australia programs but also the impact of 7 inflation; and 3) greater use of com- 54 Canada puters for the processing of bene- 31 fits, with corresponding gains in 21 Germany efficiency. 9 The World Bank (1994) compiled 76 Netherlands 23 information on administrative costs 66 per participant in publicly managed United Kingdom 32 pension plans as a percent of per 48 capita income in the early 1990s, United States 26 demonstrating that such costs were considerably higher in lower-income Source: Whitehouse, 2000. countries. For example, costs per participant as a percent of national per capita income were 8 percent in Tanzania, 7 percent in Burundi, and high income workers in developed employers to provide pension bene- 2.3 percent in Chile, compared with countries. About one-third of the fits for employees above a certain one-tenth of 1 percent or less in labor force in OECD countries was income level. OECD estimates for Switzerland and the United States. enrolled in occupational pension the early 1990s, compiled from var- These data illustrate the importance plans circa 1990, but a much ious sources, show occupational of an educated labor force, commu- smaller proportion is covered in scheme coverage in 19 industrial- nications infrastructure, and other most developing countries and tran- ized countries ranging from 5 per- advanced technological input to the sitional economies, where employ- cent in Italy and Greece to 90 per- pension production function. More er-sponsored schemes tend to cover cent in Sweden and the Netherlands recently, James, Smalhout, and only public-sector workers (World (OECD, 1998, Ageing Working Paper Vittas (2001) estimated the net Bank, 1994). 2.2). administrative fee in individual account systems as a percent of a Cross-national estimates of occupa- Of course, not all workers who are person’s total contribution in eight tional-scheme coverage among all covered by occupational plans are Latin American countries (Figure 11- workers vary widely in the pub- in fact enrolled in them. Further, 6). These fees were in double dig- lished literature, due in part to dif- the percent of older people actually its in seven of the eight nations as ferent temporal references as well receiving benefits from employer- of 1999, and exceeded 20 percent as different definitions of what an provided plans is likely to be lower in Argentina and Mexico. occupational scheme is.1 Most still, because many retirees either occupational plans are employer- did not have access to such plans ROUGHLY ONE-THIRD OF specific, but in some nations (e.g., during their working years, or did OECD WORKERS COVERED BY Denmark and the Netherlands) not participate for enough years to OCCUPATIONAL PENSION PLANS plans are organized on an industry- become vested. Whitehouse (2000) While public pension systems are wide basis, with compulsory partici- has compiled data on the percent- more widespread than occupational pation a result of collective bargain- age of pensioners with income from pension plans, the latter are grow- ing. Switzerland requires all employer-provided pensions in ing in coverage. Occupational eight developed countries in the pension plans tend to be a more 1 For a useful discussion of occupational late 1990s (Figure 11-7). In coun- pension schemes within the context of broader important source of retirement retirement system reforms, see OECD, 1998, tries where a gender breakdown is income than public pensions for Ageing Working Paper 3.4. 120 An Aging World: 2001 U.S. Census Bureau
  • 121. Box 11-1. Chile is the Developing-Country Model for Pension Privatization Chile first enacted a public pension scheme in 1911, administered retirement fund (additional payroll and expanded its program following the European deductions are made for life insurance and fund social insurance model financed on a pay-as-you-go expenses). Workers themselves select from many basis. Between 1960 and 1980, the ratio of pension- competing investment companies, are free to switch ers to contributing workers increased from 9 per 100 their accounts, and have several options for with- to 45 per 100, due to rapidly changing demographics drawal and annuities upon retirement. To reduce mis- and increasing tax evasion on the part of employees management risks, the government assumes a major and employers (Williamson, 1992). These changes, supervisory and regulatory role (Schulz, 1993). occurring in the context of a stagnant economy, By most accounts, the Chilean experiment during its resulted in a situation where the pension system was initial decade was a success, with real annual returns no longer able to meet current obligations. Faced on contributions averaging in excess of 12 percent with an increasingly bleak future scenario, the Chilean during the 1980s. From 1995 to 1998, however, government in 1980 abandoned its public system in annual rates of return were much lower and in favor of a compulsory savings plan administered by 2 years were negative, before rebounding in 1999 private-sector companies. (Figure 11-8). Overall, the long-term (19-year) aver- Since 1981, all wage and salary earners are required to age real return exceeds 11 percent. Observers have contribute 10 percent of their earnings to a privately pointed out several drawbacks to the system, such as high administrative costs, workers’ loss of freedom vis- Figure 11-8. a-vis one-tenth of their earn- Real Rate of Return of Chile's Private ings, and the fact that even- Pension System: 1981 to 1999 tual income replacement rates are not guaranteed, i.e., Percent 30 are reliant on investment earnings that may suffer in Average, 1981-1999: times of economic stagnation 25 11.2% (Gillion and Bonilla, 1992). Nevertheless, many countries 20 in Latin America, Eastern Europe, and Asia have adopt- 15 ed or are seriously consider- ing aspects of the Chilean 10 system, or are experimenting with variations on the theme 5 (Kritzer, 2000; Fox and Palmer, 2001). Consideration of increased privatization of 0 social security systems is now commonplace in much -5 of the developed world as 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 well, and has become a hotly Source: Reported in Palacios and Pallares-Miralles, 2000. debated political topic. U.S. Census Bureau An Aging World: 2001 121
  • 122. available, the data show that men are much more likely to receive Figure 11-9. such benefits, as would be expect- Private Pension Fund Assets as a Percent ed from past gender patterns of of GDP: 1970 and 1998 labor force participation. 1970 1998 Company-based pension programs 14 Canada 48 in the developing world are found 19 most frequently in former British Denmark 22 colonies and in countries with large 29 multinational subsidiaries. Most Netherlands 86 such programs are subject to less 38 regulation and lower funding Switzerland 75 requirements than their counter- 21 parts in industrialized countries, United Kingdom 84 although both Indonesia and South 29 Africa have developed comprehen- United States 86 sive and well-regulated private pen- sion systems. Coverage of private- Note: Later Swiss figure refers to 1996. Sources: World Bank, 1994 and OECD, 2000. sector workers is increasing in a number of other large developing nations such as Brazil, India, and Table 11-1. Mexico (World Bank, 1994). Payroll Tax Rates for Provident Fund Schemes: Early-to-Mid-1990s PRIVATE PENSION FUND (Percentage of wages) ASSETS A MAJOR SOURCE OF LONG-TERM CAPITAL Country Employees Employer Private pension fund assets are Africa The Gambia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 10 sizeable in many developed coun- Ghana. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 12.5 tries. In 1998, such assets were Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 equivalent to more than 80 percent Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 6 Swaziland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 of GDP in the Netherlands, the Tanzania. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 10 United Kingdom, and the United Uganda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 10 Zambia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 States (Figure 11-9). Most occupa- tional-plan funds have been invest- Asia Fiji . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7 ed in private-sector assets, are India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 10 internationally diversified, and have Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 Kiribati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 earned higher returns than publicly Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 11 managed funds (Davis, 1993; Nepal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 10 1998). Singapore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-30 10 Solomon Island . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8 Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 12 These funds have grown consider- Western Samoa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5 ably over the last three decades. Latin America The average annual growth rate of Argentina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 0 pension funds for OECD countries Chile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 0 Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 9 as a whole between 1990 and 1996 Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 0 was 11 percent (OECD, 1998b). Source: World Bank, 1994. 122 An Aging World: 2001 U.S. Census Bureau
  • 123. This increase is expected to contin- Where provident-fund coverage is nonmonetary indicators of welfare, ue for at least the short term, extensive, such funds may in effect and concluded that in all 12 because many aging nations with be the public pension system. nations, the nonelderly were better relatively underdeveloped pre- off than the elderly (Tsakloglou, The performance of provident funds funded systems have considerable 1996). Data for France, however, globally has been erratic. In some room for growth. The looming poli- suggest that extreme poverty East Asian countries (notably cy question is whether population (below the level of guaranteed mini- Singapore, which has the world’s aging will depress rates of return mum income) is much less common largest provident fund), funds typi- on private pension funds. As the among the elderly than among cally have earned positive annual large post-World War II cohorts younger households (David and investment returns. In other move into retirement, they are Starzec, 1993). nations, inflation and poor econom- expected to divest some of their ic growth have lessened the value The OECD (2000) has concluded financial assets accumulated during of fund contributions; in Sri Lanka, that there has been a stable or their working years. This prospect for example, the real annual rate of improving economic picture for highlights the importance of the return for the Employee Provident older people, both in absolute nature of pension fund manage- Fund in the 1980s and early 1990s terms and relative to the nonelderly ment, changes in government regu- often was negative (International population. Poverty rates for older lations, and the varying sociocultu- Labour Office, 1993). Such perform- people have declined in most ral impetuses for retirement savings ance has led several countries to nations, as has the share of older (OECD, 2000; National Research abandon provident schemes in people among the poor. In the Council, 2001). favor of defined-benefit pension United States, the overall situation PROVIDENT FUNDS plans (Palacios and Pallares-Miralles, of elderly people improved dramati- PARAMOUNT IN SOME 2000). cally during the last third of the DEVELOPING COUNTRIES twentieth century. Studies of real ARE LIVING STANDARDS OF median household income (adjusted A provident fund is a form of com- THE ELDERLY CHANGING IN pulsory defined-contribution pro- for household size) have demon- DEVELOPED COUNTRIES? gram wherein regular contributions strated much larger gains for elder- Given the maturation of public pen- ly people relative to the general are withheld from employee wages sions systems, increases in the level population (Radner, 1995; McNeil, and invested for later repayment. of female labor force participation, 1998). And, poverty among the Payouts typically are in the form of and the development of private elderly has declined. One-third of a lump sum upon retirement, but pension schemes, one might expect all U.S. elderly were below the may also be made earlier in times that older citizens in industrialized poverty line in 1960; by the mid- of special need. Except in some nations are better off, economically, 1990s, the level had declined to Latin American countries, employ- than previous generations of elderly 10 percent, lower than among chil- ers match or exceed the employee people. And, there is a growing dren under the age of 18 (Friedland contribution. Although provident perception in some countries that and Summer, 1999). funds can cover private-sector the elderly as a whole are faring workers, they are managed publicly. Data from the Luxembourg Income better than other population sub- Malaysia, in 1951, was the first groups. However, the complexity of Study reveal considerable inter- nation to establish a wide scale measuring economic well-being country variation in poverty rates provident fund, and other Asian often precludes a definitive assess- among elderly citizens. An analysis nations (e.g., India, Singapore, and ment of these issues, and there is of standardized information from Sri Lanka) have had provident funds considerable concern about the will- nine nations (Smeeding and for more than 40 years. By the ingness and ability of households to Saunders, 1998) suggests that mid-1990s more than 20 nations adequately save for retirement Canada, Germany, and Hungary pro- had developed such schemes (Table needs (see, e.g., MacKellar, 2000). vide their elderly the best overall 11-1). None of these countries had One study of 12 European Union protection from poverty relative to a public pay-as-you-go system at countries in the late 1980s com- the other six countries (Figure the time its provident fund was pared survey data on consumption 11-10). This analysis also highlights established (World Bank, 1994). expenditure, income, and the fact that overall figures may U.S. Census Bureau An Aging World: 2001 123
  • 124. mask large differences among (Jiminez-Martin, Labeaga, and the Health and Retirement Survey in population subgroups, as seen in Granado, 1999). In order to capture the United States (see Burkhauser the data for elderly women living the complexity of such transitions and Gertler, 1995 for a comprehen- alone. The economic vulnerability and understand their significance sive overview of this study, and of single elderly women also has for policy planning, several nations National Research Council, 2001 for been noted in a 14-country study of have mounted (or are planning to future recommendations). data from the European Community initiate) longitudinal studies akin to Household Panel (Heinrich, 2000). One obviously important compo- nent of elderly living standards is Figure 11-10. health care and its costs. As the Percent of Elderly Living in Households With Less latter have escalated in the 1990s, a Than 50 Percent of Adjusted National Median Disposable Income All elderly growing body of research has Elderly women focused on identifying the costs of living alone specific illnesses and on projecting 29 Australia, 1990 health expenditures (see, e.g., 62 Cutler and Meara, 1999; Mayhew, 7 Canada, 1991 1999; and OECD, 2000). Other 16 major thrusts of current research in 16 Finland, 1991 the economics of aging seek: to 35 8 more fully and accurately measure Germany, 1989 13 levels of household wealth and 9 assets; to better assess differences Hungary, 1995 14 in these variables within popula- 18 tions; and to understand transitions Japan, 1992 N/A in income and poverty status, 11 Poland, 1992 particularly as they relate to chang- 19 ing health status at older ages. 25 Taiwan, 1991 Data from the European Community 74 Household Panel Survey is begin- 23 United States, 1994 43 ning to shed light on the interplay of health status and retirement deci- Note: Percent for elderly women living alone in Japan is not reported. Source: Smeeding and Saunders, 1998. sions of older European couples 124 An Aging World: 2001 U.S. Census Bureau
  • 125. APPENDIX A. Detailed Tables U.S. Census Bureau An Aging World: 2001 125
  • 126. Table 1. Total Population, Percent Elderly, and Percent Oldest Old: 1975, 2000, 2015, and 2030 (In thousands) 1975 2000 Country Percent of Percent of 80+ as a Percent of Percent of 80+ as a Total population population percent of Total population population percent of population 65+ 80+ 65+ population 65+ 80+ 65+ United States . . . . . . . . . . . . . . . 220,165 10.5 2.1 20.4 275,563 12.6 3.3 26.5 Western Europe Austria . . . . . . . . . . . . . . . . . . . . . . . 7,579 14.9 2.3 15.5 8,131 15.4 3.4 22.2 Belgium . . . . . . . . . . . . . . . . . . . . . . 9,796 13.9 2.3 16.4 10,242 16.8 3.5 20.8 Denmark . . . . . . . . . . . . . . . . . . . . . 5,060 13.4 2.4 18.0 5,336 14.9 4.0 26.7 France . . . . . . . . . . . . . . . . . . . . . . . 52,699 13.5 2.5 18.3 59,330 16.0 3.7 23.3 Germany . . . . . . . . . . . . . . . . . . . . . 78,679 14.8 2.2 14.6 82,797 16.2 3.5 21.6 Greece. . . . . . . . . . . . . . . . . . . . . . . 9,047 12.2 2.1 17.1 10,602 17.3 3.5 20.2 Italy. . . . . . . . . . . . . . . . . . . . . . . . . . 55,441 12.0 1.9 16.0 57,634 18.1 4.0 22.2 Luxembourg . . . . . . . . . . . . . . . . . . 362 13.0 2.2 17.0 437 14.0 3.0 21.2 Norway . . . . . . . . . . . . . . . . . . . . . . 4,007 13.7 2.5 18.2 4,481 15.2 4.4 28.6 Sweden . . . . . . . . . . . . . . . . . . . . . . 8,193 15.1 2.7 17.8 8,873 17.3 5.0 29.2 United Kingdom . . . . . . . . . . . . . . . 56,226 14.0 2.4 17.0 59,508 15.7 4.0 25.5 Eastern Europe Bulgaria . . . . . . . . . . . . . . . . . . . . . . 8,722 10.9 1.4 12.8 7,797 16.5 2.2 13.2 Czech Republic . . . . . . . . . . . . . . . 9,997 12.9 1.7 13.5 10,272 13.9 2.4 17.1 Hungary. . . . . . . . . . . . . . . . . . . . . . 10,532 12.6 1.7 13.3 10,139 14.6 2.5 17.4 Poland . . . . . . . . . . . . . . . . . . . . . . . 34,022 9.5 1.2 12.4 38,646 12.3 2.1 16.8 Russia . . . . . . . . . . . . . . . . . . . . . . . 134,233 8.9 1.2 14.0 146,001 12.6 2.0 15.9 Ukraine . . . . . . . . . . . . . . . . . . . . . . 49,016 10.5 1.6 15.0 49,153 13.9 2.2 16.0 North America/Oceania Australia . . . . . . . . . . . . . . . . . . . . . 13,900 8.7 1.5 17.4 19,165 12.4 3.0 24.0 Canada . . . . . . . . . . . . . . . . . . . . . . 23,209 8.4 1.6 19.3 31,278 12.7 3.1 24.8 New Zealand . . . . . . . . . . . . . . . . . 3,083 8.7 1.4 16.4 3,820 11.5 2.9 25.0 Asia Bangladesh. . . . . . . . . . . . . . . . . . . 76,582 3.6 0.3 8.4 129,194 3.3 0.5 15.0 China . . . . . . . . . . . . . . . . . . . . . . . . 927,808 4.4 0.6 12.5 1,261,832 7.0 0.9 13.1 India . . . . . . . . . . . . . . . . . . . . . . . . . 620,701 3.8 0.3 8.1 1,014,004 4.6 0.6 13.1 Indonesia. . . . . . . . . . . . . . . . . . . . . 135,666 3.2 0.3 8.6 224,784 4.5 0.4 10.0 Israel . . . . . . . . . . . . . . . . . . . . . . . . 3,455 7.8 1.0 12.3 5,842 9.9 2.4 23.9 Japan . . . . . . . . . . . . . . . . . . . . . . . . 111,524 7.9 1.1 13.5 126,550 17.0 3.7 21.7 Malaysia . . . . . . . . . . . . . . . . . . . . . 12,258 3.7 0.5 13.3 21,793 4.1 0.5 13.5 Pakistan. . . . . . . . . . . . . . . . . . . . . . 74,734 3.0 0.3 10.9 141,554 4.1 0.5 13.3 Philippines. . . . . . . . . . . . . . . . . . . . 43,010 2.7 0.4 13.4 81,160 3.6 0.5 13.6 Singapore . . . . . . . . . . . . . . . . . . . . 2,263 4.1 0.4 9.7 4,152 6.8 1.5 21.3 South Korea . . . . . . . . . . . . . . . . . . 35,281 3.6 0.4 10.1 47,471 7.0 1.0 13.9 Sri Lanka. . . . . . . . . . . . . . . . . . . . . 13,603 4.1 0.5 13.2 19,239 6.5 1.0 15.6 Thailand. . . . . . . . . . . . . . . . . . . . . . 41,359 3.0 0.3 10.9 61,231 6.4 0.9 13.9 Turkey . . . . . . . . . . . . . . . . . . . . . . . 40,025 4.5 0.4 7.9 65,667 6.0 0.9 15.2 Latin America/Caribbean Argentina. . . . . . . . . . . . . . . . . . . . . 26,049 7.6 0.9 12.1 36,955 10.4 2.2 21.7 Brazil . . . . . . . . . . . . . . . . . . . . . . . . 108,167 3.9 0.5 12.5 172,860 5.3 0.8 15.3 Chile. . . . . . . . . . . . . . . . . . . . . . . . . 10,337 5.3 0.8 14.5 15,154 7.2 1.2 16.8 Colombia . . . . . . . . . . . . . . . . . . . . . 25,381 3.6 0.4 11.8 39,686 4.7 0.6 12.2 Costa Rica . . . . . . . . . . . . . . . . . . . 1,968 3.4 0.5 13.6 3,711 5.2 0.9 17.9 Guatemala . . . . . . . . . . . . . . . . . . . 6,018 2.8 0.4 13.1 12,640 3.6 0.5 13.5 Jamaica . . . . . . . . . . . . . . . . . . . . . . 2,013 5.8 0.8 14.5 2,653 6.8 1.5 21.6 Mexico . . . . . . . . . . . . . . . . . . . . . . . 59,099 4.0 0.7 17.9 100,350 4.3 0.6 14.9 Peru . . . . . . . . . . . . . . . . . . . . . . . . . 15,161 3.5 0.3 9.1 27,013 4.7 0.7 15.0 Uruguay. . . . . . . . . . . . . . . . . . . . . . 2,829 9.6 1.6 16.9 3,334 12.9 2.7 21.2 Africa Egypt . . . . . . . . . . . . . . . . . . . . . . . . 38,841 4.2 0.4 9.7 68,360 3.8 0.4 10.2 Kenya. . . . . . . . . . . . . . . . . . . . . . . . 13,741 3.7 0.5 12.8 30,340 2.7 0.4 13.0 Liberia . . . . . . . . . . . . . . . . . . . . . . . 1,609 3.7 0.9 23.3 3,164 3.4 0.6 16.8 Malawi . . . . . . . . . . . . . . . . . . . . . . . 5,244 2.2 0.2 8.0 10,386 2.8 0.3 9.9 Morocco. . . . . . . . . . . . . . . . . . . . . . 17,305 3.7 0.5 14.2 30,122 4.6 0.7 14.2 Tunisia . . . . . . . . . . . . . . . . . . . . . . . 5,668 3.5 0.5 13.1 9,593 6.0 0.8 13.3 Zimbabwe . . . . . . . . . . . . . . . . . . . . 6,143 2.6 0.3 9.9 11,343 3.5 0.5 14.7 126 An Aging World: 2001 U.S. Census Bureau
  • 127. Table 1. Total Population, Percent Elderly, and Percent Oldest Old: 1975, 2000, 2015, and 2030—Con. (In thousands) 2015 2030 Country Percent of Percent of 80+ as a Percent of Percent of 80+ as a Total population population percent of Total population population percent of population 65+ 80+ 65+ population 65+ 80+ 65+ United States . . . . . . . . . . . . . . . . . . . 312,524 14.7 3.8 25.8 351,326 20.0 5.3 26.4 Western Europe Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . 8,316 18.8 4.9 26.2 8,278 25.2 7.0 27.9 Belgium . . . . . . . . . . . . . . . . . . . . . . . . . . 10,336 19.4 5.7 29.3 10,175 25.4 7.3 28.8 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . 5,521 18.9 4.4 23.6 5,649 23.0 7.1 30.8 France . . . . . . . . . . . . . . . . . . . . . . . . . . . 61,545 18.8 5.8 30.9 61,926 24.0 7.5 31.2 Germany . . . . . . . . . . . . . . . . . . . . . . . . . 85,192 20.2 5.4 26.6 84,939 25.8 7.2 28.1 Greece . . . . . . . . . . . . . . . . . . . . . . . . . . . 10,735 20.6 6.3 30.5 10,316 25.4 7.8 30.8 Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56,631 22.2 6.8 30.5 52,868 28.1 9.0 32.1 Luxembourg . . . . . . . . . . . . . . . . . . . . . . 519 15.3 4.1 27.1 580 19.8 5.2 26.2 Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . 4,767 17.4 4.6 26.3 5,018 22.0 6.6 30.0 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . 8,900 21.4 5.7 26.8 8,868 25.1 8.6 34.3 United Kingdom . . . . . . . . . . . . . . . . . . . 61,047 18.4 4.9 26.8 61,481 23.5 7.0 29.7 Eastern Europe Bulgaria . . . . . . . . . . . . . . . . . . . . . . . . . . 6,663 20.2 4.6 23.0 5,668 25.9 7.2 27.8 Czech Republic . . . . . . . . . . . . . . . . . . . 10,048 18.8 4.2 22.3 9,409 24.7 7.4 30.0 Hungary . . . . . . . . . . . . . . . . . . . . . . . . . . 9,666 17.6 4.3 24.2 9,034 22.5 6.3 27.9 Poland . . . . . . . . . . . . . . . . . . . . . . . . . . . 38,668 15.0 3.8 25.1 37,377 22.2 5.5 24.8 Russia . . . . . . . . . . . . . . . . . . . . . . . . . . . 141,073 13.8 3.1 22.7 132,859 20.5 4.1 20.0 Ukraine. . . . . . . . . . . . . . . . . . . . . . . . . . . 45,294 15.0 3.2 21.1 42,273 19.7 4.2 21.5 North America/Oceania Australia . . . . . . . . . . . . . . . . . . . . . . . . . . 21,697 15.8 4.1 25.9 23,497 21.1 6.0 28.5 Canada. . . . . . . . . . . . . . . . . . . . . . . . . . . 35,653 16.1 4.3 26.8 39,128 22.9 6.2 26.9 New Zealand. . . . . . . . . . . . . . . . . . . . . . 4,396 13.7 3.5 25.7 4,768 17.8 5.0 28.2 Asia Bangladesh . . . . . . . . . . . . . . . . . . . . . . . 160,486 4.4 0.6 12.5 184,478 7.2 1.0 13.5 China . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1,397,414 9.5 1.7 18.0 1,483,121 16.0 2.9 18.3 India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1,241,572 5.9 0.9 14.5 1,437,103 9.0 1.4 15.7 Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . 275,152 6.2 1.1 16.9 312,592 10.9 1.7 15.6 Israel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6,992 11.1 3.0 26.6 7,873 14.9 3.9 26.5 Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125,843 24.9 7.0 28.2 116,740 28.3 11.1 39.3 Malaysia. . . . . . . . . . . . . . . . . . . . . . . . . . 28,414 5.9 0.8 14.3 35,306 9.4 1.6 16.9 Pakistan . . . . . . . . . . . . . . . . . . . . . . . . . . 185,715 4.5 0.7 15.0 226,251 6.5 0.9 14.4 Philippines . . . . . . . . . . . . . . . . . . . . . . . . 106,098 4.9 0.7 14.6 129,448 7.7 1.2 15.9 Singapore . . . . . . . . . . . . . . . . . . . . . . . . 6,646 8.7 2.1 24.1 9,047 14.8 3.0 20.4 South Korea . . . . . . . . . . . . . . . . . . . . . . 52,239 11.3 2.2 19.3 53,763 19.5 4.2 21.3 Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . 21,527 9.5 1.7 17.7 22,937 15.2 3.1 20.2 Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . 68,139 9.8 1.8 18.0 71,311 16.4 3.1 19.3 Turkey. . . . . . . . . . . . . . . . . . . . . . . . . . . . 76,685 7.9 1.6 19.9 84,195 12.9 2.4 18.7 Latin America/Caribbean Argentina . . . . . . . . . . . . . . . . . . . . . . . . . 42,916 11.8 3.1 26.0 47,229 14.7 4.0 27.3 Brazil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192,313 8.1 1.5 18.7 203,489 13.2 2.7 20.6 Chile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17,405 10.7 2.1 19.3 18,915 16.4 3.7 22.3 Colombia . . . . . . . . . . . . . . . . . . . . . . . . . 49,189 6.5 1.0 15.6 57,666 11.5 1.8 15.9 Costa Rica. . . . . . . . . . . . . . . . . . . . . . . . 4,583 7.3 1.4 19.2 5,272 12.8 2.4 18.9 Guatemala . . . . . . . . . . . . . . . . . . . . . . . . 18,105 4.1 0.7 17.8 24,038 5.6 1.0 17.6 Jamaica . . . . . . . . . . . . . . . . . . . . . . . . . . 2,992 7.4 1.8 24.1 3,353 12.5 2.3 18.7 Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . 121,712 6.3 1.0 16.6 139,125 10.2 1.9 18.7 Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33,551 6.4 1.2 18.3 39,253 9.9 1.9 19.0 Uruguay . . . . . . . . . . . . . . . . . . . . . . . . . . 3,730 13.5 3.8 28.2 4,109 15.5 4.4 28.1 Africa Egypt . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85,219 5.1 0.6 11.9 99,583 8.0 1.1 14.2 Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33,612 3.8 0.6 16.6 34,836 5.2 1.1 20.7 Liberia. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4,655 4.0 0.8 21.1 6,745 4.2 1.0 24.9 Malawi . . . . . . . . . . . . . . . . . . . . . . . . . . . 12,017 3.1 0.4 12.7 12,817 3.2 0.6 17.5 Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . 37,832 5.5 1.0 17.4 44,664 9.1 1.4 15.2 Tunisia . . . . . . . . . . . . . . . . . . . . . . . . . . . 11,174 7.6 1.5 19.8 12,322 12.7 2.3 17.7 Zimbabwe . . . . . . . . . . . . . . . . . . . . . . . . 10,548 5.0 1.0 20.2 9,086 6.4 1.8 27.7 Source: United Nations, 1999 and U.S. Census Bureau, 2000a. U.S. Census Bureau An Aging World: 2001 127
  • 128. Table 2. Population by Age: 2000 and 2030 (In thousands) 2000 Country 0 to 24 25 to 54 55 to 64 65 to 69 70 to 74 75 to 79 80 years All ages years years years years years years and over United States . . . . . . . . . . . . . . . 275,563 97,064 119,662 24,001 9,436 8,753 7,422 9,225 Western Europe Austria . . . . . . . . . . . . . . . . . . . . . . . 8,131 2,325 3,629 926 347 332 294 278 Belgium . . . . . . . . . . . . . . . . . . . . . . 10,242 3,033 4,432 1,052 521 462 383 358 Denmark . . . . . . . . . . . . . . . . . . . . . 5,336 1,594 2,340 610 219 194 167 212 France . . . . . . . . . . . . . . . . . . . . . . . 59,330 18,852 25,513 5,471 2,711 2,466 2,100 2,216 Germany . . . . . . . . . . . . . . . . . . . . . 82,797 22,309 36,224 10,813 4,104 3,592 2,846 2,911 Greece. . . . . . . . . . . . . . . . . . . . . . . 10,602 3,088 4,480 1,195 605 521 341 371 Italy. . . . . . . . . . . . . . . . . . . . . . . . . . 57,634 14,873 25,640 6,696 3,093 2,766 2,253 2,313 Luxembourg . . . . . . . . . . . . . . . . . . 437 133 199 44 19 17 12 13 Norway . . . . . . . . . . . . . . . . . . . . . . 4,481 1,435 1,937 426 167 164 156 196 Sweden . . . . . . . . . . . . . . . . . . . . . . 8,873 2,655 3,674 1,010 380 363 343 447 United Kingdom . . . . . . . . . . . . . . . 59,508 18,549 25,496 6,138 2,585 2,347 2,018 2,373 Eastern Europe Bulgaria . . . . . . . . . . . . . . . . . . . . . . 7,797 2,354 3,255 902 453 382 282 169 Czech Republic . . . . . . . . . . . . . . . 10,272 3,250 4,505 1,093 448 409 323 244 Hungary. . . . . . . . . . . . . . . . . . . . . . 10,139 3,207 4,336 1,113 479 420 326 257 Poland . . . . . . . . . . . . . . . . . . . . . . . 38,646 13,915 16,676 3,319 1,616 1,372 953 794 Russia . . . . . . . . . . . . . . . . . . . . . . . 146,001 49,232 64,197 14,160 5,996 6,182 3,299 2,936 Ukraine . . . . . . . . . . . . . . . . . . . . . . 49,153 16,052 20,607 5,647 2,080 2,294 1,377 1,096 North America/Oceania Australia . . . . . . . . . . . . . . . . . . . . . 19,165 6,629 8,427 1,727 668 633 509 573 Canada . . . . . . . . . . . . . . . . . . . . . . 31,278 10,154 14,322 2,838 1,147 1,012 822 984 New Zealand . . . . . . . . . . . . . . . . . 3,820 1,417 1,637 324 124 116 91 110 Asia Bangladesh. . . . . . . . . . . . . . . . . . . 129,194 76,298 42,947 5,645 1,744 1,206 710 643 China . . . . . . . . . . . . . . . . . . . . . . . . 1,261,832 515,155 572,082 86,822 34,926 25,426 15,908 11,513 India . . . . . . . . . . . . . . . . . . . . . . . . . 1,014,004 536,947 373,956 56,037 18,477 13,785 8,627 6,175 Indonesia. . . . . . . . . . . . . . . . . . . . . 224,784 113,419 88,231 13,080 4,616 2,872 1,559 1,006 Israel . . . . . . . . . . . . . . . . . . . . . . . . 5,842 2,617 2,257 391 168 152 120 138 Japan . . . . . . . . . . . . . . . . . . . . . . . . 126,550 34,782 53,858 16,385 7,031 5,812 4,012 4,670 Malaysia . . . . . . . . . . . . . . . . . . . . . 21,793 11,583 8,175 1,152 353 260 151 119 Pakistan. . . . . . . . . . . . . . . . . . . . . . 141,554 86,109 43,165 6,485 2,317 1,637 1,071 770 Philippines. . . . . . . . . . . . . . . . . . . . 81,160 46,410 28,087 3,717 1,220 820 504 401 Singapore . . . . . . . . . . . . . . . . . . . . 4,152 1,333 2,281 253 98 76 49 61 South Korea . . . . . . . . . . . . . . . . . . 47,471 18,091 22,191 3,875 1,365 895 594 460 Sri Lanka. . . . . . . . . . . . . . . . . . . . . 19,239 8,759 7,933 1,295 455 358 244 195 Thailand. . . . . . . . . . . . . . . . . . . . . . 61,231 25,879 27,045 4,386 1,591 1,086 699 545 Turkey . . . . . . . . . . . . . . . . . . . . . . . 65,667 32,182 25,619 3,935 1,514 1,099 721 597 Latin America/Caribbean Argentina. . . . . . . . . . . . . . . . . . . . . 36,955 16,326 13,856 2,946 1,209 1,026 761 831 Brazil . . . . . . . . . . . . . . . . . . . . . . . . 172,860 84,691 68,842 10,136 3,501 2,594 1,693 1,403 Chile. . . . . . . . . . . . . . . . . . . . . . . . . 15,154 6,733 6,194 1,133 391 310 210 184 Colombia . . . . . . . . . . . . . . . . . . . . . 39,686 19,897 15,867 2,071 742 546 338 225 Costa Rica . . . . . . . . . . . . . . . . . . . 3,711 1,888 1,438 193 70 53 35 34 Guatemala . . . . . . . . . . . . . . . . . . . 12,640 7,922 3,765 496 184 132 79 62 Jamaica . . . . . . . . . . . . . . . . . . . . . . 2,653 1,312 1,027 135 56 48 36 39 Mexico . . . . . . . . . . . . . . . . . . . . . . . 100,350 54,699 36,241 5,092 1,722 1,197 757 642 Peru . . . . . . . . . . . . . . . . . . . . . . . . . 27,013 14,735 9,606 1,409 500 351 223 189 Uruguay. . . . . . . . . . . . . . . . . . . . . . 3,334 1,348 1,257 298 138 118 84 91 Africa Egypt . . . . . . . . . . . . . . . . . . . . . . . . 68,360 37,706 24,572 3,509 1,162 748 401 261 Kenya. . . . . . . . . . . . . . . . . . . . . . . . 30,340 20,064 8,423 1,021 340 237 146 108 Liberia . . . . . . . . . . . . . . . . . . . . . . . 3,164 1,985 929 141 43 29 19 18 Malawi . . . . . . . . . . . . . . . . . . . . . . . 10,386 6,955 2,757 385 126 85 49 28 Morocco. . . . . . . . . . . . . . . . . . . . . . 30,122 16,826 10,472 1,434 538 393 262 197 Tunisia . . . . . . . . . . . . . . . . . . . . . . . 9,593 4,834 3,638 542 223 173 106 77 Zimbabwe . . . . . . . . . . . . . . . . . . . . 11,343 7,281 3,238 422 156 112 74 59 128 An Aging World: 2001 U.S. Census Bureau
  • 129. Table 2. Population by Age: 2000 and 2030—Con. (In thousands) 2030 Country 0 to 24 25 to 54 55 to 64 65 to 69 70 to 74 75 to 79 80 years All ages years years years years years years and over United States . . . . . . . . . . . . . . . 351,326 115,218 128,484 37,305 19,844 17,878 14,029 18,569 Western Europe Austria . . . . . . . . . . . . . . . . . . . . . . . 8,278 1,916 3,034 1,244 636 500 367 582 Belgium . . . . . . . . . . . . . . . . . . . . . . 10,175 2,553 3,683 1,357 709 628 501 744 Denmark . . . . . . . . . . . . . . . . . . . . . 5,649 1,515 2,076 760 357 295 246 400 France . . . . . . . . . . . . . . . . . . . . . . . 61,926 16,405 22,599 8,073 3,775 3,435 3,005 4,635 Germany . . . . . . . . . . . . . . . . . . . . . 84,939 20,074 31,104 11,886 6,502 5,192 4,036 6,145 Greece. . . . . . . . . . . . . . . . . . . . . . . 10,316 2,287 3,814 1,594 692 620 503 807 Italy. . . . . . . . . . . . . . . . . . . . . . . . . . 52,868 10,165 18,788 9,033 4,115 3,307 2,685 4,775 Luxembourg . . . . . . . . . . . . . . . . . . 580 164 229 72 35 29 22 30 Norway . . . . . . . . . . . . . . . . . . . . . . 5,018 1,405 1,856 653 295 261 217 331 Sweden . . . . . . . . . . . . . . . . . . . . . . 8,868 2,210 3,267 1,167 555 482 424 763 United Kingdom . . . . . . . . . . . . . . . 61,481 16,077 22,663 8,296 4,215 3,336 2,598 4,296 Eastern Europe Bulgaria . . . . . . . . . . . . . . . . . . . . . . 5,668 1,096 2,240 866 381 362 315 408 Czech Republic . . . . . . . . . . . . . . . 9,409 1,933 3,718 1,436 583 532 512 696 Hungary. . . . . . . . . . . . . . . . . . . . . . 9,034 2,012 3,662 1,330 485 506 472 566 Poland . . . . . . . . . . . . . . . . . . . . . . . 37,377 9,260 15,184 4,642 2,087 2,267 1,882 2,056 Russia . . . . . . . . . . . . . . . . . . . . . . . 132,859 35,650 53,589 16,428 8,288 7,776 5,681 5,446 Ukraine . . . . . . . . . . . . . . . . . . . . . . 42,273 11,383 17,099 5,479 2,553 2,292 1,683 1,783 North America/Oceania Australia . . . . . . . . . . . . . . . . . . . . . 23,497 6,643 8,941 2,960 1,356 1,212 975 1,410 Canada . . . . . . . . . . . . . . . . . . . . . . 39,128 10,368 14,987 4,800 2,581 2,249 1,728 2,414 New Zealand . . . . . . . . . . . . . . . . . 4,768 1,400 1,914 607 233 209 166 238 Asia Bangladesh. . . . . . . . . . . . . . . . . . . 184,478 71,167 84,043 16,060 5,248 3,821 2,356 1,783 China . . . . . . . . . . . . . . . . . . . . . . . . 1,483,121 437,787 588,812 219,501 84,958 59,230 49,367 43,466 India . . . . . . . . . . . . . . . . . . . . . . . . . 1,437,103 558,161 614,683 135,423 49,013 35,886 23,744 20,194 Indonesia. . . . . . . . . . . . . . . . . . . . . 312,592 112,472 132,916 33,067 12,612 9,740 6,450 5,335 Israel . . . . . . . . . . . . . . . . . . . . . . . . 7,873 2,724 3,154 825 327 288 246 310 Japan . . . . . . . . . . . . . . . . . . . . . . . . 116,740 25,589 40,441 17,661 7,094 6,391 6,562 13,002 Malaysia . . . . . . . . . . . . . . . . . . . . . 35,306 15,017 13,891 3,063 1,236 919 617 563 Pakistan. . . . . . . . . . . . . . . . . . . . . . 226,251 95,929 98,625 17,008 5,826 4,186 2,568 2,109 Philippines. . . . . . . . . . . . . . . . . . . . 129,448 55,474 53,369 10,581 3,738 2,814 1,877 1,596 Singapore . . . . . . . . . . . . . . . . . . . . 9,047 2,345 4,158 1,204 468 360 239 274 South Korea . . . . . . . . . . . . . . . . . . 53,763 14,515 20,967 7,819 3,470 2,959 1,803 2,231 Sri Lanka. . . . . . . . . . . . . . . . . . . . . 22,937 7,102 9,580 2,771 1,142 945 694 703 Thailand. . . . . . . . . . . . . . . . . . . . . . 71,311 21,219 28,987 9,441 3,927 3,189 2,303 2,245 Turkey . . . . . . . . . . . . . . . . . . . . . . . 84,195 26,295 36,793 10,231 3,940 2,899 2,002 2,036 Latin America/Caribbean Argentina. . . . . . . . . . . . . . . . . . . . . 47,229 16,082 19,374 4,834 1,963 1,705 1,376 1,895 Brazil . . . . . . . . . . . . . . . . . . . . . . . . 203,489 66,334 87,458 22,898 9,091 7,080 5,099 5,530 Chile. . . . . . . . . . . . . . . . . . . . . . . . . 18,915 5,863 7,817 2,133 1,006 823 582 691 Colombia . . . . . . . . . . . . . . . . . . . . . 57,666 21,940 23,069 6,034 2,481 1,872 1,217 1,053 Costa Rica . . . . . . . . . . . . . . . . . . . 5,272 1,796 2,248 554 235 188 124 127 Guatemala . . . . . . . . . . . . . . . . . . . 24,038 12,128 9,105 1,455 493 368 252 238 Jamaica . . . . . . . . . . . . . . . . . . . . . . 3,353 1,062 1,447 425 158 110 73 78 Mexico . . . . . . . . . . . . . . . . . . . . . . . 139,125 52,128 58,225 14,646 5,165 3,700 2,621 2,639 Peru . . . . . . . . . . . . . . . . . . . . . . . . . 39,253 14,952 16,694 3,735 1,394 1,031 711 736 Uruguay. . . . . . . . . . . . . . . . . . . . . . 4,109 1,444 1,593 434 183 156 119 179 Africa Egypt . . . . . . . . . . . . . . . . . . . . . . . . 99,583 38,878 43,516 9,212 3,144 2,261 1,437 1,136 Kenya. . . . . . . . . . . .